tdehaeze/dehaeze21_desig_compl_filte
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Reworked all sections / add many figures

Browse Source 
Sent for first internal review
This commit is contained in:
Thomas Dehaeze 2021-08-31 14:19:41 +02:00
parent f57207d529
commit c6d715734f
32 changed files with 1620 additions and 805 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

567
journal/dehaeze21_desig_compl_filte.org
View File

@ -1,4 +1,4 @@
#+TITLE: A new method of designing complementary filters for sensor fusion using $\mathcal{H}_\infty$ synthesis
#+TITLE: A new method of designing complementary filters for sensor fusion using the $\mathcal{H}_\infty$ synthesis
:DRAWER:
#+LATEX_CLASS: elsarticle
#+LATEX_CLASS_OPTIONS: [preprint, sort&compress]
@ -10,14 +10,14 @@
#+LATEX_HEADER_EXTRA: \author[a1,a2]{Thomas Dehaeze\corref{cor1}}
#+LATEX_HEADER_EXTRA: \author[a3,a4]{Mohit Verma}
#+LATEX_HEADER_EXTRA: \author[a2,a4]{Christophe Collette}
#+LATEX_HEADER_EXTRA: \cortext[cor1]{Corresponding author. Email Address: dehaeze.thomas@gmail.com}
#+LATEX_HEADER_EXTRA: \cortext[cor1]{Corresponding author. Email Address: thomas.dehaeze@esrf.fr}
#+LATEX_HEADER_EXTRA: \address[a1]{European Synchrotron Radiation Facility, Grenoble, France}
#+LATEX_HEADER_EXTRA: \address[a2]{University of Li\`{e}ge, Department of Aerospace and Mechanical Engineering, 4000 Li\`{e}ge, Belgium.}
#+LATEX_HEADER_EXTRA: \address[a3]{CSIR --- Structural Engineering Research Centre, Taramani, Chennai --- 600113, India.}
#+LATEX_HEADER_EXTRA: \address[a4]{Universit\'{e} Libre de Bruxelles, Precision Mechatronics Laboratory, BEAMS Department, 1050 Brussels, Belgium.}
#+LATEX_HEADER: \usepackage{subcaption}
#+LATEX_HEADER: \usepackage{caption,subcaption}
#+LATEX_HEADER: \usepackage{amsfonts}
#+LATEX_HEADER: \usepackage{siunitx}
#+LATEX_HEADER_EXTRA: \usepackage{tabularx}
@ -26,6 +26,7 @@
#+LATEX_HEADER_EXTRA: \usepackage[hyperref]{xcolor}
#+LATEX_HEADER_EXTRA: \usepackage[top=2cm, bottom=2cm, left=2cm, right=2cm]{geometry}
#+LATEX_HEADER_EXTRA: \hypersetup{colorlinks=true}
#+LATEX_HEADER_EXTRA: \setlength{\parskip}{1em}
:END:
* Build :noexport:
@ -63,19 +64,18 @@
* Abstract and Keywords :ignore:
#+begin_frontmatter
#+begin_abstract
Sensor have limited bandwidth and are accurate only in a certain frequency band.
In many applications, the signals of different sensor are fused together in order to either enhance the stability or improve the operational bandwidth of the system.
The sensor signals can be fused using complementary filters.
The tuning of complementary filters is a complex task and is the subject of this paper.
The filters needs to meet design specifications while satisfying the complementary property.
This paper presents a framework to shape the norm of complementary filters using the $\mathcal{H}_\infty$ norm minimization.
The design specifications are imposed as constraints in the optimization problem by appropriate selection of weighting functions.
The proposed method is quite general and easily extendable to cases where more than two sensors are fused.
Finally, the proposed method is applied to the design of complementary filter design for active vibration isolation of the Laser Interferometer Gravitation-wave Observatory (LIGO).
In order to obtain a better estimate of a quantity being measured, several sensors having different characteristics can be merged with a technique called "sensor fusion".
The obtained "super sensor" can combine the benefits of the individual sensors provided that the complementary filters used in the fusion are well designed.
Indeed, properties of the super sensor are linked to the magnitude of the complementary filters.
Properly shaping the magnitude of complementary filters is a difficult and time-consuming task.
In this study, we address this issue and propose a new method for designing complementary filters.
This method uses weighting functions to specify the wanted shape of the complementary filter that are then easily obtained using the standard $\mathcal{H}_\infty$ synthesis.
The proper choice of the weighting functions is discussed, and the effectiveness and simplicity of the design method is highlighted using several examples.
Such synthesis method is further extended for the shaping of more than two complementary filters.
#+end_abstract
#+begin_keyword
Sensor fusion \sep{} Optimal filters \sep{} $\mathcal{H}_\infty$ synthesis \sep{} Vibration isolation \sep{} Precision
Sensor fusion \sep{} Complementary filters \sep{} $\mathcal{H}_\infty$ synthesis \sep{} Vibration isolation \sep{} Motion control
#+end_keyword
#+end_frontmatter
@ -83,95 +83,70 @@ Sensor fusion \sep{} Optimal filters \sep{} $\mathcal{H}_\infty$ synthesis \sep{
<<sec:introduction>>
** Introduction to Sensor Fusion :ignore:
# Basic explanations of sensor fusion
# "Fusing" several sensor
- cite:anderson53_instr_approac_system_steer_comput earliest application of complementary filters (A simple RC circuit was used to physically realize the complementary filters)
- cite:bendat57_optim_filter_indep_measur_two roots of sensor fusion
Measuring a physical quantity using sensors is always subject to several limitations.
First, the accuracy of the measurement will be affected by several noise sources, such as the electrical noise of the conditioning electronics being used.
Second, the frequency range in which the measurement is relevant is bounded by the bandwidth of the sensor.
One way to overcome these limitations is to combine several sensors using a technique called "sensor fusion"\nbsp{}cite:bendat57_optim_filter_indep_measur_two.
Fortunately, a wide variety of sensors exist, each with different characteristics.
By carefully choosing the fused sensors, a so called "super sensor" is obtain that combines benefits of individual sensors and yields a better estimate of the measured physical quantity. \par
** Advantages of Sensor Fusion :ignore:
# Sensor Fusion can have many advantages / can be applied for various purposes
Sensor fusion can have many advantages.
In some situations, it is used to increase the bandwidth of the sensor cite:shaw90_bandw_enhan_posit_measur_using_measur_accel,zimmermann92_high_bandw_orien_measur_contr,min15_compl_filter_desig_angle_estim.
For instance by increasing the high frequency bandwidth of a position sensor using an accelerometer.
Decrease the noise: cite:hua05_low_ligo,hua04_polyp_fir_compl_filter_contr_system,plummer06_optim_compl_filter_their_applic_motion_measur
[[cite:robert12_introd_random_signal_applied_kalman][chapter 8]]
Increased robustness (sensor measuring different quantities): cite:collette15_sensor_fusion_method_high_perfor,yong16_high_speed_vertic_posit_stage
\par
In some situations, sensor fusion is used to increase the bandwidth of the measurement\nbsp{}cite:shaw90_bandw_enhan_posit_measur_using_measur_accel,zimmermann92_high_bandw_orien_measur_contr,min15_compl_filter_desig_angle_estim.
For instance, in\nbsp{}cite:shaw90_bandw_enhan_posit_measur_using_measur_accel, the bandwidth of a position sensor is increased by fusing it with an accelerometer providing the high frequency motion information.
For other applications, sensor fusion is used to obtain a estimate of the measured quantity with lower noise\nbsp{}cite:hua05_low_ligo,hua04_polyp_fir_compl_filter_contr_system,plummer06_optim_compl_filter_their_applic_motion_measur,robert12_introd_random_signal_applied_kalman.
More recently, the fusion of sensors measuring different physical quantities has been proposed to obtained interesting properties for control\nbsp{}cite:collette15_sensor_fusion_method_high_perfor,yong16_high_speed_vertic_posit_stage.
In\nbsp{}cite:collette15_sensor_fusion_method_high_perfor, an inertial sensor used for active vibration isolation is fused with a sensor collocated with the actuator to improve the stability margins of the feedback controller. \par
** Applications :ignore:
# The applications of sensor fusion are numerous
The applications of sensor fusion are numerous.
It is widely used for attitude estimation of unmanned aerial vehicle
cite:baerveldt97_low_cost_low_weigh_attit,pascoal99_navig_system_desig_using_time,corke04_inert_visual_sensin_system_small_auton_helic,batista10_optim_posit_veloc_navig_filter_auton_vehic,jensen13_basic_uas,min15_compl_filter_desig_angle_estim
Motion control
cite:shaw90_bandw_enhan_posit_measur_using_measur_accel,zimmermann92_high_bandw_orien_measur_contr
Tjepkema et al. cite:tjepkema12_sensor_fusion_activ_vibrat_isolat_precis_equip used sensor fusion to isolate precision equipment from the ground motion.
Gravitational wave observer cite:heijningen18_low:
LIGO cite:hua05_low_ligo,hua04_polyp_fir_compl_filter_contr_system
VIRGO cite:lucia18_low_frequen_optim_perfor_advan
Practical applications of sensor fusion are numerous.
It is widely used for the attitude estimation of several autonomous vehicles such as unmanned aerial vehicle\nbsp{}cite:baerveldt97_low_cost_low_weigh_attit,corke04_inert_visual_sensin_system_small_auton_helic,jensen13_basic_uas and underwater vehicles\nbsp{}cite:pascoal99_navig_system_desig_using_time,batista10_optim_posit_veloc_navig_filter_auton_vehic.
Naturally, it is of great benefits for high performance positioning control as shown in\nbsp{}cite:shaw90_bandw_enhan_posit_measur_using_measur_accel,zimmermann92_high_bandw_orien_measur_contr,min15_compl_filter_desig_angle_estim,yong16_high_speed_vertic_posit_stage.
Sensor fusion was also shown to be a key technology to improve the performances of active vibration isolation systems\nbsp{}cite:tjepkema12_sensor_fusion_activ_vibrat_isolat_precis_equip.
This is particularly apparent for the isolation stages of gravitational wave observer\nbsp{}cite:collette15_sensor_fusion_method_high_perfor,heijningen18_low such as the ones used at the LIGO\nbsp{}cite:hua05_low_ligo,hua04_polyp_fir_compl_filter_contr_system and at the VIRGO\nbsp{}cite:lucia18_low_frequen_optim_perfor_advan. \par
** Kalman Filtering / Complementary filters :ignore:
There are mainly two ways to perform sensor fusion: using complementary filters or using Kalman filtering cite:brown72_integ_navig_system_kalman_filter.
Kalman filtering cite:odry18_kalman_filter_mobil_robot_attit_estim
Relations between CF and Kalman: cite:becker15_compl_filter_desig_three_frequen_bands
Advantages of complementary filtering over Kalman filtering for sensor fusion:
- Less computation cite:higgins75_compar_compl_kalman_filter
- For Kalman filtering, we are forced to make assumption about the probabilistic character of the sensor noises cite:robert12_introd_random_signal_applied_kalman
- More intuitive frequency domain technique
There are mainly two ways to perform sensor fusion: either using a set of complementary filters\nbsp{}cite:anderson53_instr_approac_system_steer_comput or using Kalman filtering\nbsp{}cite:brown72_integ_navig_system_kalman_filter,odry18_kalman_filter_mobil_robot_attit_estim.
For sensor fusion applications, both methods are sharing many relationships\nbsp{}cite:brown72_integ_navig_system_kalman_filter,higgins75_compar_compl_kalman_filter,robert12_introd_random_signal_applied_kalman,becker15_compl_filter_desig_three_frequen_bands.
However, for Kalman filtering, we are forced to make assumption about the probabilistic character of the sensor noises\nbsp{}cite:robert12_introd_random_signal_applied_kalman whereas it is not the case for complementary filters.
Furthermore, the main advantages of complementary filters over Kalman filtering for sensor fusion are their very general applicability, their low computational cost\nbsp{}cite:higgins75_compar_compl_kalman_filter, and the fact that they are very intuitive as their effects can be easily interpreted in the frequency domain. \par
** Design Methods of Complementary filters :ignore:
In some cases, complementary filters are implemented in an analog way such as in cite:yong16_high_speed_vertic_posit_stage,moore19_capac_instr_sensor_fusion_high_bandw_nanop, but most of the time it is implemented numerically which allows much more complex
A set of filters is said to be complementary if the sum of their transfer functions is equal to one at all frequencies.
For the earliest used of complementary filtering for sensor fusion, analog circuits were used to physically realize the filters\nbsp{}cite:anderson53_instr_approac_system_steer_comput.
Analog complementary filters are still used today\nbsp{}cite:yong16_high_speed_vertic_posit_stage,moore19_capac_instr_sensor_fusion_high_bandw_nanop, but most of the time they are now implemented numerically as it allows for much more flexibility. \par
Multiple design methods have been used for complementary filters
Several design methods have been developed over the years to optimize complementary filters.
The easiest way to design complementary filters is to use analytical formulas.
Depending on the application, the formulas used are of first order\nbsp{}cite:corke04_inert_visual_sensin_system_small_auton_helic,yeh05_model_contr_hydraul_actuat_two,yong16_high_speed_vertic_posit_stage, second order\nbsp{}cite:baerveldt97_low_cost_low_weigh_attit,stoten01_fusion_kinet_data_using_compos_filter,jensen13_basic_uas or even higher orders\nbsp{}cite:shaw90_bandw_enhan_posit_measur_using_measur_accel,zimmermann92_high_bandw_orien_measur_contr,stoten01_fusion_kinet_data_using_compos_filter,collette15_sensor_fusion_method_high_perfor,matichard15_seism_isolat_advan_ligo. \par
- Analytical methods:
- first order: cite:corke04_inert_visual_sensin_system_small_auton_helic,yong16_high_speed_vertic_posit_stage
- second order: cite:baerveldt97_low_cost_low_weigh_attit, cite:stoten01_fusion_kinet_data_using_compos_filter, cite:jensen13_basic_uas
- higher order: cite:shaw90_bandw_enhan_posit_measur_using_measur_accel, cite:zimmermann92_high_bandw_orien_measur_contr, cite:collette15_sensor_fusion_method_high_perfor, cite:matichard15_seism_isolat_advan_ligo
- cite:pascoal99_navig_system_desig_using_time use LMI to generate complementary filters (convex optimization techniques), specific for navigation systems
- cite:hua05_low_ligo,hua04_polyp_fir_compl_filter_contr_system: FIR + convex optimization
- Similar to feedback system:
- cite:plummer06_optim_compl_filter_their_applic_motion_measur use H-Infinity to optimize complementary filters (flatten the super sensor noise spectral density)
- cite:jensen13_basic_uas design of complementary filters with classical control theory, PID
- 3 complementary filters: cite:becker15_compl_filter_desig_three_frequen_bands
As the characteristics of the "super sensor" depends on the design of the complementary filters\nbsp{}cite:dehaeze19_compl_filter_shapin_using_synth, several optimization techniques were developed over the years.
Some are based on the finding the optimal parameters of analytical formulas\nbsp{}cite:jensen13_basic_uas,min15_compl_filter_desig_angle_estim,becker15_compl_filter_desig_three_frequen_bands, while other are using convex optimization tools\nbsp{}cite:hua04_polyp_fir_compl_filter_contr_system,hua05_low_ligo such as linear matrix inequalities\nbsp{}cite:pascoal99_navig_system_desig_using_time.
As shown in\nbsp{}cite:plummer06_optim_compl_filter_their_applic_motion_measur, the design of complementary filters can also be linked to the standard mixed-sensitivity control problem.
Therefore, all the powerful tools developed for the classical control theory can also be used for the design of complementary filters.
For instance, in\nbsp{}cite:jensen13_basic_uas the two gains of a Proportional Integral (PI) controller are optimized to minimize the noise of the super sensor. \par
** Problematic / gap in the research :ignore:
- Robustness problems: cite:zimmermann92_high_bandw_orien_measur_contr,plummer06_optim_compl_filter_their_applic_motion_measur change of phase near the merging frequency
- Trial and error
- Although many design methods of complementary filters have been proposed in the literature, no simple method that allows to shape the norm of the complementary filters is available.
The common objective to all these complementary filters design methods is to obtain a super sensor that has desired characteristics, usually in terms of noise and dynamics.
Moreover, as reported in\nbsp{}cite:zimmermann92_high_bandw_orien_measur_contr,plummer06_optim_compl_filter_their_applic_motion_measur, phase shifts and magnitude bumps of the "super sensors" dynamics can be observed if either the complementary filters are poorly designed or if the sensors are not well calibrated.
Therefore, the robustness of the fusion is also of concerned when designing the complementary filters.
Although many design methods of complementary filters have been proposed in the literature, no simple method that allows to specify the wanted super sensor characteristic while ensuring good fusion robustness has been proposed. \par
** Describe the paper itself / the problem which is addressed :ignore:
Most of the requirements => shape of the complementary filters
=> propose a way to shape complementary filters.
Fortunately, both the robustness of the fusion and the super sensor characteristics can be linked to the magnitude of the complementary filters\nbsp{}cite:dehaeze19_compl_filter_shapin_using_synth.
Based on that, this paper introduces a new way to design complementary filters using the $\mathcal{H}_\infty$ synthesis which allows to shape the magnitude of the complementary filters in an easy and intuitive way. \par
** Introduce Each part of the paper :ignore:
Section ref:sec:requirements
Section ref:sec:hinf_method
Section ref:sec:application_ligo
Section ref:sec:discussion
Section\nbsp{}ref:sec:requirements introduces the sensor fusion architecture and demonstrates how typical requirements can be linked to the complementary filters magnitudes.
In section\nbsp{}ref:sec:hinf_method, the shaping of complementary filters is written as an $\mathcal{H}_\infty$ optimization problem using weighting functions, and the simplicity of the proposed method is illustrated with an example.
The synthesis method is further validated in Section\nbsp{}ref:sec:application_ligo by designing complex complementary filters.
Section\nbsp{}ref:sec:discussion compares the proposed synthesis method with the classical mixed-sensitivity synthesis, and extends it to the shaping of more than two complementary filters.
* Sensor Fusion and Complementary Filters Requirements
<<sec:requirements>>
@ -184,64 +159,61 @@ These requirements are discussed in this section.
** Sensor Fusion Architecture
<<sec:sensor_fusion>>
A general sensor fusion architecture using complementary filters is shown in Figure ref:fig:sensor_fusion_overview where several sensors (here two) are measuring the same physical quantity $x$.
The two sensors output signals are estimates $\hat{x}_1$ and $\hat{x}_2$ of $x$.
A general sensor fusion architecture using complementary filters is shown in Fig.\nbsp{}ref:fig:sensor_fusion_overview where several sensors (here two) are measuring the same physical quantity $x$.
The two sensors output signals $\hat{x}_1$ and $\hat{x}_2$ are estimates of $x$.
Each of these estimates are then filtered out by complementary filters and combined to form a new estimate $\hat{x}$.
The resulting sensor, termed as "super sensor", can have larger bandwidth and better noise characteristics in comparison to the individual sensor.
The resulting sensor, termed as super sensor, can have larger bandwidth and better noise characteristics in comparison to the individual sensor.
This means that the super sensor provides an estimate $\hat{x}$ of $x$ which can be more accurate over a larger frequency band than the outputs of the individual sensors.
#+name: fig:sensor_fusion_overview
#+caption: Schematic of a sensor fusion architecture
#+caption: Schematic of a sensor fusion architecture using complementary filters
[[file:figs/sensor_fusion_overview.pdf]]
The complementary property of filters $H_1(s)$ and $H_2(s)$ implies that the summation of their transfer functions is equal to unity.
The complementary property of filters $H_1(s)$ and $H_2(s)$ implies that the sum of their transfer functions is equal to unity.
That is, unity magnitude and zero phase at all frequencies.
Therefore, a pair of strict complementary filter needs to satisfy the following condition:
Therefore, a pair of complementary filter needs to satisfy the following condition:
#+name: eq:comp_filter
\begin{equation}
H_1(s) + H_2(s) = 1
\end{equation}
It will soon become clear why the complementary property is important.
It will soon become clear why the complementary property is important for the sensor fusion architecture.
** Sensor Models and Sensor Normalization
<<sec:sensor_models>>
In order to study such sensor fusion architecture, a model of the sensors is required.
Such model is shown in Figure ref:fig:sensor_model and consists of a linear time invariant (LTI) system $G_i(s)$ representing the dynamics of the sensor and an additive noise input $n_i$ representing its noise.
In order to study such sensor fusion architecture, a model for the sensors is required.
Such model is shown in Fig.\nbsp{}ref:fig:sensor_model and consists of a linear time invariant (LTI) system $G_i(s)$ representing the sensor dynamics and an input $n_i$ representing the sensor noise.
The model input $x$ is the measured physical quantity and its output $\tilde{x}_i$ is the "raw" output of the sensor.
Before filtering the sensor outputs $\tilde{x}_i$ by the complementary filters, the sensors are usually normalized to simplify the fusion.
This normalization consists of first obtaining an estimate $\hat{G}_i(s)$ of the sensor dynamics $G_i(s)$.
It is supposed that the estimate of the sensor dynamics $\hat{G}_i(s)$ can be inverted and that its inverse $\hat{G}_i^{-1}(s)$ is proper and stable.
The raw output of the sensor $\tilde{x}_i$ is then passed through $\hat{G}_i^{-1}(s)$ as shown in Figure ref:fig:sensor_model_calibrated.
This normalization consists of using an estimate $\hat{G}_i(s)$ of the sensor dynamics $G_i(s)$, and filtering the sensor output by the inverse of this estimate $\hat{G}_i^{-1}(s)$ as shown in Fig.\nbsp{}ref:fig:sensor_model_calibrated.
It is here supposed that the sensor inverse $\hat{G}_i^{-1}(s)$ is proper and stable.
This way, the units of the estimates $\hat{x}_i$ are equal to the units of the physical quantity $x$.
The sensor dynamics estimate $\hat{G}_1(s)$ can be a simple gain or more complex transfer functions.
The sensor dynamics estimate $\hat{G}_1(s)$ can be a simple gain or a more complex transfer function.
#+begin_export latex
\begin{figure}[htbp]
\begin{subfigure}[b]{0.49\linewidth}
\centering
\includegraphics[scale=1]{figs/sensor_model.pdf}
\caption{\label{fig:sensor_model} Basic sensor model consisting of a noise input $n_i$ and a dynamics $G_i(s)$}
\caption{\label{fig:sensor_model} Basic sensor model consisting of a noise input $n_i$ and a linear time invariant transfer function $G_i(s)$}
\end{subfigure}
\hfill
\begin{subfigure}[b]{0.49\linewidth}
\centering
\includegraphics[scale=1]{figs/sensor_model_calibrated.pdf}
\caption{\label{fig:sensor_model_calibrated} Calibrated sensors using the inverse of an estimate $\hat{G}_1(s)$ of the sensor dynamics}
\caption{\label{fig:sensor_model_calibrated} Normalized sensors using the inverse of an estimate $\hat{G}_i(s)$ of the sensor dynamics}
\end{subfigure}
\caption{\label{fig:sensor_models}Sensor models with an without normalization}
\caption{\label{fig:sensor_models}Sensor models with and without normalization}
\centering
\end{figure}
#+end_export
Two calibrated sensors and then combined to form a super sensor as shown in Figure ref:fig:fusion_super_sensor.
Two normalized sensors are then combined to form a super sensor as shown in Fig.\nbsp{}ref:fig:fusion_super_sensor.
The two sensors are measuring the same physical quantity $x$ with dynamics $G_1(s)$ and $G_2(s)$, and with /uncorrelated/ noises $n_1$ and $n_2$.
The normalized signals from both calibrated sensors are fed into two complementary filters $H_1(s)$ and $H_2(s)$ and then combined to yield an estimate $\hat{x}$ of $x$ as shown in Fig. ref:fig:fusion_super_sensor.
The signals from both normalized sensors are fed into two complementary filters $H_1(s)$ and $H_2(s)$ and then combined to yield an estimate $\hat{x}$ of $x$.
The super sensor output is therefore equal to:
#+name: eq:comp_filter_estimate
@ -250,14 +222,13 @@ The super sensor output is therefore equal to:
\end{equation}
#+name: fig:fusion_super_sensor
#+caption: Sensor fusion architecture
#+attr_latex: :scale 1
#+caption: Sensor fusion architecture with two normalized sensors
[[file:figs/fusion_super_sensor.pdf]]
** Noise Sensor Filtering
<<sec:noise_filtering>>
In this section, it is supposed that all the sensors are perfectly calibrated, such that:
In this section, it is supposed that all the sensors are perfectly normalized, such that:
#+name: eq:perfect_dynamics
\begin{equation}
\frac{\hat{x}_i}{x} = \hat{G}_i(s) G_i(s) = 1
@ -265,7 +236,7 @@ In this section, it is supposed that all the sensors are perfectly calibrated, s
The effect of a non-perfect normalization will be discussed in the next section.
The super sensor output $\hat{x}$ is then:
Provided eqref:eq:perfect_dynamics is verified, the super sensor output $\hat{x}$ is then:
#+name: eq:estimate_perfect_dyn
\begin{equation}
\hat{x} = x + H_1(s) n_1 + H_2(s) n_2
@ -273,6 +244,7 @@ The super sensor output $\hat{x}$ is then:
From eqref:eq:estimate_perfect_dyn, the complementary filters $H_1(s)$ and $H_2(s)$ are shown to only operate on the sensor's noises.
Thus, this sensor fusion architecture permits to filter the noise of both sensors without introducing any distortion in the physical quantity to be measured.
This is why the two filters must be complementary.
The estimation error $\delta x$, defined as the difference between the sensor output $\hat{x}$ and the measured quantity $x$, is computed for the super sensor eqref:eq:estimate_error.
#+name: eq:estimate_error
@ -286,11 +258,11 @@ As shown in eqref:eq:noise_filtering_psd, the Power Spectral Density (PSD) of th
\Phi_{\delta x}(\omega) = \left|H_1(j\omega)\right|^2 \Phi_{n_1}(\omega) + \left|H_2(j\omega)\right|^2 \Phi_{n_2}(\omega)
\end{equation}
If the two sensors have identical noise characteristics ($\Phi_{n_1}(\omega) = \Phi_{n_2}(\omega)$), a simple averaging ($H_1(s) = H_2(s) = 0.5$) is what would minimize the super sensor noise.
If the two sensors have identical noise characteristics, $\Phi_{n_1}(\omega) = \Phi_{n_2}(\omega)$, a simple averaging ($H_1(s) = H_2(s) = 0.5$) is what would minimize the super sensor noise.
This is the simplest form of sensor fusion with complementary filters.
However, the two sensors have usually high noise levels over distinct frequency regions.
In such case, to lower the noise of the super sensor, the value of the norm $|H_1|$ has to be lowered when $\Phi_{n_1}$ is larger than $\Phi_{n_2}$ and that of $|H_2|$ lowered when $\Phi_{n_2}$ is larger than $\Phi_{n_1}$.
In such case, to lower the noise of the super sensor, the norm $|H_1(j\omega)|$ has to be small when $\Phi_{n_1}(\omega)$ is larger than $\Phi_{n_2}(\omega)$ and the norm $|H_2(j\omega)|$ has to be small when $\Phi_{n_2}(\omega)$ is larger than $\Phi_{n_1}(\omega)$.
Therefore, by properly shaping the norm of the complementary filters, it is possible to minimize the noise of the super sensor noise.
** Sensor Fusion Robustness
@ -298,11 +270,12 @@ Therefore, by properly shaping the norm of the complementary filters, it is poss
In practical systems the sensor normalization is not perfect and condition eqref:eq:perfect_dynamics is not verified.
In order to study such imperfection, a multiplicative input uncertainty is added to the sensor dynamics (Figure ref:fig:sensor_model_uncertainty), where the nominal model is taken as the estimated model for the normalization $\hat{G}_i(s)$, $\Delta_i$ is any stable transfer function satisfying $|\Delta_i(j\omega)| \le 1,\ \forall\omega$, and $w_i(s)$ is a weight representing the magnitude of the uncertainty.
In order to study such imperfection, a multiplicative input uncertainty is added to the sensor dynamics (Fig.\nbsp{}ref:fig:sensor_model_uncertainty).
The nominal model is taken as the estimated model for the normalization $\hat{G}_i(s)$, $\Delta_i$ is any stable transfer function satisfying $|\Delta_i(j\omega)| \le 1,\ \forall\omega$, and $w_i(s)$ is a weighting transfer function representing the magnitude of the uncertainty.
The weight $w_i(s)$ is chosen such that the real sensor dynamics is always contained in the uncertain region represented by a circle centered on $1$ and with a radius equal to $|w_i(j\omega)|$.
The weight $w_i(s)$ is chosen such that the real sensor dynamics $G(j\omega)$ is contained in the uncertain region represented by a circle in the complex plane, centered on $1$ and with a radius equal to $|w_i(j\omega)|$.
As the nominal sensor dynamics is taken as the normalized filter, the normalized sensor can be further simplified as shown in Figure ref:fig:sensor_model_uncertainty_simplified.
As the nominal sensor dynamics is taken as the normalized filter, the normalized sensor can be further simplified as shown in Fig.\nbsp{}ref:fig:sensor_model_uncertainty_simplified.
#+begin_export latex
\begin{figure}[htbp]
@ -323,30 +296,31 @@ As the nominal sensor dynamics is taken as the normalized filter, the normalized
\end{figure}
#+end_export
A sensor fusion architecture with two sensors with dynamical uncertainty is shown in Figure ref:fig:sensor_fusion_dynamic_uncertainty.
The sensor fusion architecture with two sensor models including dynamical uncertainty is shown in Fig.\nbsp{}ref:fig:sensor_fusion_dynamic_uncertainty.
#+name: fig:sensor_fusion_dynamic_uncertainty
#+caption: Sensor fusion architecture with sensor dynamics uncertainty
[[file:figs/sensor_fusion_dynamic_uncertainty.pdf]]
The super sensor dynamics eqref:eq:super_sensor_dyn_uncertainty is no longer equal to $1$ and now depends on the sensor dynamics uncertainty weights $w_i(s)$ as well as on the complementary filters $H_i(s)$.
The super sensor dynamics eqref:eq:super_sensor_dyn_uncertainty is no longer equal to $1$ and now depends on the sensor dynamical uncertainty weights $w_i(s)$ as well as on the complementary filters $H_i(s)$.
#+name: eq:super_sensor_dyn_uncertainty
\begin{equation}
\frac{\hat{x}}{x} = 1 + w_1(s) H_1(s) \Delta_1(s) + w_2(s) H_2(s) \Delta_2(s)
\end{equation}
The dynamical uncertainty of the super sensor can be graphically represented in the complex plane by a circle centered on $1$ with a radius equal to $|w_1(j\omega) H_1(j\omega)| + |w_2(j\omega) H_2(j\omega)|$ as shown in Figure ref:fig:uncertainty_set_super_sensor.
The dynamical uncertainty of the super sensor can be graphically represented in the complex plane by a circle centered on $1$ with a radius equal to $|w_1(j\omega) H_1(j\omega)| + |w_2(j\omega) H_2(j\omega)|$ as shown in Fig.\nbsp{}ref:fig:uncertainty_set_super_sensor.
#+name: fig:uncertainty_set_super_sensor
#+caption: Uncertainty region of the super sensor dynamics in the complex plane (solid circle). The contribution of both sensors 1 and 2 to the uncertainty are represented respectively by a blue circle and a red circle. The frequency dependency $\omega$ is here omitted.
#+caption: Uncertainty region of the super sensor dynamics in the complex plane (grey circle). The contribution of both sensors 1 and 2 to the total uncertainty are represented respectively by a blue circle and a red circle. The frequency dependency $\omega$ is here omitted.
[[file:figs/uncertainty_set_super_sensor.pdf]]
The super sensor dynamical uncertainty (i.e. the robustness of the fusion) clearly depends on the complementary filters norms.
For instance, the phase uncertainty $\Delta\phi(\omega)$ added by the super sensor dynamics at frequency $\omega$ can be found by drawing a tangent from the origin to the uncertainty circle of super sensor (Figure ref:fig:uncertainty_set_super_sensor) and is bounded by eqref:eq:max_phase_uncertainty.
For instance, the phase $\Delta\phi(\omega)$ added by the super sensor dynamics at frequency $\omega$ is bounded by $\Delta\phi_{\text{max}}(\omega)$ which can be found by drawing a tangent from the origin to the uncertainty circle of the super sensor (Fig.\nbsp{}ref:fig:uncertainty_set_super_sensor).
Therefore, the phase uncertainty of the super sensor dynamics depends on the Complementary filters norms eqref:eq:max_phase_uncertainty.
#+name: eq:max_phase_uncertainty
\begin{equation}
\Delta\phi(\omega) < \arcsin\big( |w_1(j\omega) H_1(j\omega)| + |w_2(j\omega) H_2(j\omega)| \big)
\Delta\phi_\text{max}(\omega) = \arcsin\big( |w_1(j\omega) H_1(j\omega)| + |w_2(j\omega) H_2(j\omega)| \big)
\end{equation}
As it is generally desired to limit the maximum phase added by the super sensor, $H_1(s)$ and $H_2(s)$ should be designed such that $\Delta \phi$ is bounded to acceptable values.
@ -355,16 +329,18 @@ Typically, the norm of the complementary filter $|H_i(j\omega)|$ should be made
* Complementary Filters Shaping
<<sec:hinf_method>>
** Introduction :ignore:
As shown in Section ref:sec:requirements, the noise and robustness of the "super sensor" are determined by the complementary filters norms.
Therefore, a complementary filters synthesis method that allows to shape their norms would be of great use.
In this section, such synthesis is proposed by expressing this problem as a $\mathcal{H}_\infty$ norm optimization.
As shown in Section\nbsp{}ref:sec:requirements, the noise and robustness of the super sensor are a function of the complementary filters norms.
Therefore, a complementary filters synthesis method that allows to shape their norms would be of great use.
In this section, such synthesis is proposed by writing the synthesis objective as a standard $\mathcal{H}_\infty$ optimization problem.
As weighting functions are used to represent the wanted complementary filters shapes during the synthesis, the proper design of weighting functions is discussed.
Finally, the synthesis method is validated on an simple example.
** Synthesis Objective
<<sec:synthesis_objective>>
The synthesis objective is to shape the norm of two filters $H_1(s)$ and $H_2(s)$ while ensuring their complementary property eqref:eq:comp_filter.
This is equivalent as to finding proper and stable transfer functions $H_1(s)$ and $H_2(s)$ such that conditions eqref:eq:comp_filter_problem_form are satisfied.
This is equivalent as to finding proper and stable transfer functions $H_1(s)$ and $H_2(s)$ such that conditions eqref:eq:hinf_cond_complementarity, eqref:eq:hinf_cond_h1 and eqref:eq:hinf_cond_h2 are satisfied.
#+name: eq:comp_filter_problem_form
\begin{subequations}
@ -374,26 +350,41 @@ This is equivalent as to finding proper and stable transfer functions $H_1(s)$ a
& |H_2(j\omega)| \le \frac{1}{|W_2(j\omega)|} \quad \forall\omega \label{eq:hinf_cond_h2}
\end{align}
\end{subequations}
where $W_1(s)$ and $W_2(s)$ are two weighting transfer functions that are chosen to specify the maximum wanted norms of the complementary filters during the synthesis.
$W_1(s)$ and $W_2(s)$ are two weighting transfer functions that are carefully chosen to specify the maximum wanted norms of the complementary filters during the synthesis.
** Shaping of Complementary Filters using $\mathcal{H}_\infty$ synthesis
<<sec:hinf_synthesis>>
In this section, it is shown that the synthesis objective can be easily expressed as a standard $\mathcal{H}_\infty$ optimal control problem and therefore solved using convenient tools readily available.
In this section, it is shown that the synthesis objective can be easily expressed as a standard $\mathcal{H}_\infty$ optimization problem and therefore solved using convenient tools readily available.
Consider the generalized plant $P(s)$ shown in Figure ref:fig:h_infinity_robust_fusion and mathematically described by eqref:eq:generalized_plant.
Consider the generalized plant $P(s)$ shown in Fig.\nbsp{}ref:fig:h_infinity_robust_fusion_plant and mathematically described by eqref:eq:generalized_plant.
#+name: eq:generalized_plant
\begin{equation}
\begin{bmatrix} z_1 \\ z_2 \\ v \end{bmatrix} = P(s) \begin{bmatrix} w\\u \end{bmatrix}; \quad P(s) = \begin{bmatrix}W_1(s) & -W_1(s) \\ 0 & \phantom{+}W_2(s) \\ 1 & 0 \end{bmatrix}
\end{equation}
#+name: fig:h_infinity_robust_fusion
#+caption: Generalized plant used for $\mathcal{H}_\infty$ synthesis of complementary filters
#+attr_latex: :scale 1
[[file:figs/h_infinity_robust_fusion.pdf]]
#+begin_export latex
\begin{figure}[htbp]
\begin{subfigure}[t]{0.5\linewidth}
\centering
\includegraphics[scale=1]{figs/h_infinity_robust_fusion_plant.pdf}
\caption{\label{fig:h_infinity_robust_fusion_plant} Generalized plant}
\vfill
\end{subfigure}
\hfill
\begin{subfigure}[t]{0.5\linewidth}
\centering
\includegraphics[scale=1]{figs/h_infinity_robust_fusion_fb.pdf}
\caption{\label{fig:h_infinity_robust_fusion_fb} Generalized plant with the synthesized filter}
\end{subfigure}
\hfill
\caption{\label{fig:h_infinity_robust_fusion} Architecture for the $\mathcal{H}_\infty$ synthesis of complementary filters}
\centering
\end{figure}
#+end_export
Applying the standard $\mathcal{H}_\infty$ synthesis on the generalized plant $P(s)$ is then equivalent as finding a stable filter $H_2(s)$ which based on $v$, generates a signal $u$ such that the $\mathcal{H}_\infty$ norm from $w$ to $[z_1, \ z_2]$ is less than one eqref:eq:hinf_syn_obj.
Applying the standard $\mathcal{H}_\infty$ synthesis on the generalized plant $P(s)$ is then equivalent as finding a stable filter $H_2(s)$ which based on $v$, generates a signal $u$ such that the $\mathcal{H}_\infty$ norm of the system in Fig.\nbsp{}ref:fig:h_infinity_robust_fusion_fb from $w$ to $[z_1, \ z_2]$ is less than one eqref:eq:hinf_syn_obj.
#+name: eq:hinf_syn_obj
\begin{equation}
\left\|\begin{matrix} \left(1 - H_2(s)\right) W_1(s) \\ H_2(s) W_2(s) \end{matrix}\right\|_\infty \le 1
@ -413,21 +404,23 @@ By then defining $H_1(s)$ to be the complementary of $H_2(s)$ eqref:eq:definitio
Therefore, applying the $\mathcal{H}_\infty$ synthesis on the standard plant $P(s)$ eqref:eq:generalized_plant will generate two filters $H_2(s)$ and $H_1(s) \triangleq 1 - H_2(s)$ that are complementary eqref:eq:comp_filter_problem_form and such that there norms are bellow specified bounds [[eqref:eq:hinf_cond_h1]],eqref:eq:hinf_cond_h2.
The above optimization problem can be efficiently solved in Matlab cite:matlab20 using the Robust Control Toolbox.
Note that there is not an equivalence between the $\mathcal{H}_\infty$ norm condition\nbsp{}eqref:eq:hinf_problem and the initial synthesis objectives\nbsp{}eqref:eq:hinf_cond_h1 and\nbsp{}eqref:eq:hinf_cond_h2, but only an implication.
Hence, the optimization may be a little bit conservative with respect to the "set" of filters on which it is performed.
There might be solutions were the objectives\nbsp{}eqref:eq:hinf_cond_h1 and\nbsp{}eqref:eq:hinf_cond_h2 are valid but where the $\mathcal{H}_\infty$ norm\nbsp{}eqref:eq:hinf_problem is larger than one.
In practice, this is however not an found to be an issue.
** Weighting Functions Design
<<sec:hinf_weighting_func>>
Weighting functions are used during the synthesis to specify what is the maximum allowed norms of the complementary filters.
The proper design of these weighting functions is of primary importance for the success of the presented complementary filters $\mathcal{H}_\infty$ synthesis.
Weighting functions are used during the synthesis to specify the maximum allowed norms of the complementary filters.
The proper design of these weighting functions is of primary importance for the success of the presented $\mathcal{H}_\infty$ synthesis of complementary filters.
First, only proper and stable transfer functions should be used.
Second, the order of the weighting functions should stay reasonably small in order to reduce the computational costs associated with the solving of the optimization problem and for the physical implementation of the filters (the order of the synthesized filters being equal to the sum of the weighting functions order).
Second, the order of the weighting functions should stay reasonably small in order to reduce the computational costs associated with the solving of the optimization problem and for the physical implementation of the filters (the order of the synthesized filters being equal to the sum of the weighting functions orders).
Third, one should not forget the fundamental limitations imposed by the complementary property eqref:eq:comp_filter.
This implies for instance that $|H_1(j\omega)|$ and $|H_2(j\omega)|$ cannot be made small at the same frequency.
When designing complementary filters, it is usually desired to specify its slope, its crossover frequency and its maximum gain at low and high frequency.
When designing complementary filters, it is usually desired to specify their slopes, their "blending" frequency and their maximum gains at low and high frequency.
To help with the design of the weighting functions such that the above specification can be easily expressed, the formula eqref:eq:weight_formula is proposed.
#+name: eq:weight_formula
@ -442,8 +435,8 @@ To help with the design of the weighting functions such that the above specifica
The parameters in formula eqref:eq:weight_formula are:
- $G_0 = lim_{\omega \to 0} |W(j\omega)|$: the low frequency gain
- $G_\infty = lim_{\omega \to \infty} |W(j\omega)|$: the high frequency gain
- $G_c = |W(j\omega_c)|$: the gain at $\omega_c$
- $n$: the slope between high and low frequency. It is also the order of the weighting function.
- $G_c = |W(j\omega_c)|$: the gain at a specific frequency $\omega_c$ in $\si{rad/s}$.
- $n$: the slope between high and low frequency. It also corresponds to the order of the weighting function.
The parameters $G_0$, $G_c$ and $G_\infty$ should either satisfy condition eqref:eq:cond_formula_1 or eqref:eq:cond_formula_2.
#+name: eq:condition_params_formula
@ -454,47 +447,63 @@ The parameters $G_0$, $G_c$ and $G_\infty$ should either satisfy condition eqref
\end{align}
\end{subequations}
The typical shape of a weighting function generated using eqref:eq:weight_formula is shown in Figure ref:fig:weight_formula.
An example of the obtained magnitude of a weighting function generated using eqref:eq:weight_formula is shown in Fig.\nbsp{}ref:fig:weight_formula.
#+name: fig:weight_formula
#+caption: Magnitude of a weighting function generated using the proposed formula eqref:eq:weight_formula, $G_0 = 1e^{-3}$, $G_\infty = 10$, $\omega_c = \SI{10}{Hz}$, $G_c = 2$, $n = 3$
#+attr_latex: :scale 1
[[file:figs/weight_formula.pdf]]
** Validation of the proposed synthesis method
<<sec:hinf_example>>
The proposed methodology for the design of complementary filters is now applied on a simple example where two complementary filters $H_1(s)$ and $H_2(s)$ have to be designed such that:
- the merging frequency is around $\SI{10}{Hz}$
- the slope of $|H_1(j\omega)|$ is $-2$ above $\SI{10}{Hz}$
- the slope of $|H_2(j\omega)|$ is $+3$ below $\SI{10}{Hz}$
- the maximum gain of both filters is $10^{-3}$ away from the merging frequency
- the blending frequency is around $\SI{10}{Hz}$
- the slope of $|H_1(j\omega)|$ is $+2$ below $\SI{10}{Hz}$, its low frequency gain is $10^{-3}$
- the slope of $|H_2(j\omega)|$ is $-3$ above $\SI{10}{Hz}$, its high frequency gain is $10^{-3}$
The first step is to design weighting functions that translate the above requirements.
They are here designed using eqref:eq:weight_formula with parameters summarized in table ref:tab:weights_params.
The magnitudes of the weighting functions are shown by dashed lines in Figure ref:fig:hinf_filters_results.
The first step is to translate the above requirements into the design of the weighting functions.
The proposed formula eqref:eq:weight_formula is here used for such purpose.
Parameters used are summarized in Table\nbsp{}ref:tab:weights_params.
The inverse magnitudes of the designed weighting functions, which are representing the maximum allowed norms of the complementary filters, are shown by the dashed lines in Fig.\nbsp{}ref:fig:weights_W1_W2.
#+name: tab:weights_params
#+caption: Parameters used for weighting functions $W_1(s)$ and $W_2(s)$ using eqref:eq:weight_formula
#+ATTR_LATEX: :environment tabularx :width 0.29\linewidth :align ccc
#+ATTR_LATEX: :center t :booktabs t
| Parameters | $W_1(s)$ | $W_2(s)$ |
|------------+---------------+---------------|
| $G_0$ | $0.1$ | $1000$ |
| $G_\infty$ | $1000$ | $0.1$ |
| $\omega_c$ | $2\pi\cdot10$ | $2\pi\cdot10$ |
| $G_c$ | $0.45$ | $0.45$ |
| $n$ | $2$ | $3$ |
#+begin_export latex
\begin{figure}
\begin{minipage}[b]{0.49\linewidth}
\centering
\begin{tabularx}{0.65\linewidth}{ccc}
\toprule
Parameters & \(W_1(s)\) & \(W_2(s)\)\\
\midrule
\(G_0\) & \(0.1\) & \(1000\)\\
\(G_\infty\) & \(1000\) & \(0.1\)\\
\(\omega_c\) & \(2\pi\cdot10\) & \(2\pi\cdot10\)\\
\(G_c\) & \(0.45\) & \(0.45\)\\
\(n\) & \(2\) & \(3\)\\
\bottomrule
\end{tabularx}
\captionof{table}{\label{tab:weights_params}Parameters used for weighting functions \(W_1(s)\) and \(W_2(s)\) using \eqref{eq:weight_formula}}
\end{minipage}
\hfill
\begin{minipage}[b]{0.49\linewidth}
\centering
\includegraphics[scale=1]{figs/weights_W1_W2.pdf}
\caption{\label{fig:weights_W1_W2}Inverse magnitude of the weighting functions}
\end{minipage}
\end{figure}
#+end_export
The $\mathcal{H}_\infty$ synthesis is applied on the generalized plant of Figure ref:fig:h_infinity_robust_fusion using the Matlab =hinfsyn= command.
The synthesized filter $H_2(s)$ is such that $\mathcal{H}_\infty$ norm between $w$ and $[z_1,\ z_2]^T$ is minimized and here found close to one eqref:eq:hinf_synthesis_result.
The $\mathcal{H}_\infty$ synthesis is then applied to the generalized plant of Fig.\nbsp{}ref:fig:h_infinity_robust_fusion_plant on efficiently solved in Matlab\nbsp{}cite:matlab20 using the Robust Control Toolbox.
The filter $H_2(s)$ that minimizes the $\mathcal{H}_\infty$ norm between $w$ and $[z_1,\ z_2]^T$ is obtained.
The $\mathcal{H}_\infty$ norm is here found to be close to one eqref:eq:hinf_synthesis_result which indicates that the synthesis is successful: the complementary filters norms are below the maximum specified upper bounds.
This is confirmed by the bode plots of the obtained complementary filters in Fig.\nbsp{}ref:fig:hinf_filters_results.
#+name: eq:hinf_synthesis_result
\begin{equation}
\left\|\begin{matrix} \left(1 - H_2(s)\right) W_1(s) \\ H_2(s) W_2(s) \end{matrix}\right\|_\infty \approx 1
\end{equation}
The bode plots of the obtained complementary filters are shown by solid lines in Figure ref:fig:hinf_filters_results and their transfer functions in the Laplace domain are given in eqref:eq:hinf_synthesis_result_tf.
Their transfer functions in the Laplace domain are given in eqref:eq:hinf_synthesis_result_tf.
As expected, the obtained filters are of order $5$, that is the sum of the weighting functions orders.
#+name: eq:hinf_synthesis_result_tf
\begin{subequations}
@ -504,83 +513,78 @@ The bode plots of the obtained complementary filters are shown by solid lines in
\end{align}
\end{subequations}
The obtained transfer functions are of order $5$ as expected (sum of the weighting functions orders), and their magnitudes are bellow the maximum specified ones as ensured by eqref:eq:hinf_synthesis_result.
#+name: fig:hinf_filters_results
#+caption: Frequency response of the weighting functions and complementary filters obtained using $\mathcal{H}_\infty$ synthesis
#+caption: Bode plot of the obtained complementary filters
[[file:figs/hinf_filters_results.pdf]]
This simple example illustrates the fact that the proposed methodology for complementary filters shaping is quite easy to use and effective.
This simple example illustrates the fact that the proposed methodology for complementary filters shaping is easy to use and effective.
A more complex real life example is taken up in the next section.
* Application: Design of Complementary Filters used in the Active Vibration Isolation System at the LIGO
<<sec:application_ligo>>
** Introduction :ignore:
Sensor fusion using complementary filters are widely used in active vibration isolation systems in gravitational wave detectors such at the LIGO cite:matichard15_seism_isolat_advan_ligo,hua05_low_ligo, the VIRGO cite:lucia18_low_frequen_optim_perfor_advan,heijningen18_low and the KAGRA [[cite:sekiguchi16_study_low_frequen_vibrat_isolat_system][Chap. 5]].
Sensor fusion using complementary filters are widely used in active vibration isolation systems in gravitational wave detectors such at the LIGO\nbsp{}cite:matichard15_seism_isolat_advan_ligo,hua05_low_ligo, the VIRGO\nbsp{}cite:lucia18_low_frequen_optim_perfor_advan,heijningen18_low and the KAGRA [[cite:sekiguchi16_study_low_frequen_vibrat_isolat_system][Chap. 5]].
In the first isolation stage at the LIGO, two sets of complementary filters are used and included in a feedback loop cite:hua04_low_ligo.
In the first isolation stage at the LIGO, two sets of complementary filters are used and included in a feedback loop\nbsp{}cite:hua04_low_ligo.
A set of complementary filters ($L_2,H_2$) is first used to fuse a seismometer and a geophone.
Then, another set of complementary filters ($L_1,H_1$) is used to merge the output of the first "inertial super sensor" with a position sensor.
A simplified block diagram of the sensor fusion architecture is shown in Figure ref:fig:ligo_super_sensor_architecture.
A simplified block diagram of the sensor fusion architecture is shown in Fig.\nbsp{}ref:fig:ligo_super_sensor_architecture.
#+name: fig:ligo_super_sensor_architecture
#+caption: Simplified block diagram of the sensor blending strategy for the first stage at the LIGO cite:hua04_low_ligo
#+attr_latex: :scale 1
#+caption: Simplified block diagram of the sensor blending strategy for the first stage at the LIGO\nbsp{}cite:hua04_low_ligo
[[file:figs/ligo_super_sensor_architecture.pdf]]
The fusion of the position sensor at low frequency with the "inertial super sensor" at high frequency using the complementary filters ($L_1,H_1$) is done for several reasons, first of which is to give the super sensor a DC sensibility and therefore allow the feedback loop to have authority at zero frequency.
The requirements on those filters are very tight and thus their design is complex and should be expressed as an optimization problem.
The approach used in cite:hua05_low_ligo is to use FIR complementary filters and to write the synthesis as a convex optimization problem.
The approach used in\nbsp{}cite:hua05_low_ligo is to use FIR complementary filters and to write the synthesis as a convex optimization problem.
After synthesis, the obtained FIR filters were found to be compliant with the requirements.
However they are of very high order so their implementation is quite complex.
In this section, the effectiveness of the proposed complementary filter synthesis strategy is demonstrated on the same set of requirements.
# Example where clearly manual tuning of the complementary filters is not an option
** Complementary Filters Specifications
<<sec:ligo_specifications>>
The specifications for the set of complementary filters ($L_1,H_1$) used at the LIGO are summarized below (for further details, refer to cite:hua04_polyp_fir_compl_filter_contr_system):
The specifications for the set of complementary filters ($L_1,H_1$) used at the LIGO are summarized below (for further details, refer to\nbsp{}cite:hua04_polyp_fir_compl_filter_contr_system):
- From $0$ to $\SI{0.008}{Hz}$, the magnitude $|L_1(j\omega)|$ should be less or equal to $8 \times 10^{-4}$
- Between $\SI{0.008}{Hz}$ to $\SI{0.04}{Hz}$, the filter $L_1(s)$ should attenuate the input signal proportional to frequency cubed
- Between $\SI{0.04}{Hz}$ to $\SI{0.1}{Hz}$, the magnitude $|L_1(j\omega)|$ should be less than $3$
- Above $\SI{0.1}{Hz}$, the magnitude $|H_1(j\omega)|$ should be less than $0.045$
These specifications are therefore upper bounds on the complementary filters' magnitudes.
They are physically represented in Figure ref:fig:fir_filter_ligo as well as the obtained magnitude of the FIR filters in cite:hua05_low_ligo.
# Replicated using SeDuMi matlab toolbox cite:sturm99_using_sedum
They are physically represented in Fig.\nbsp{}ref:fig:fir_filter_ligo as well as the obtained magnitude of the FIR filters in\nbsp{}cite:hua05_low_ligo.
#+name: fig:fir_filter_ligo
#+caption: Specifications and Bode plot of the obtained FIR filters in cite:hua05_low_ligo
#+attr_latex: :scale 1
#+caption: Specifications and Bode plot of the obtained FIR complementary filters in\nbsp{}cite:hua05_low_ligo. The filters are here obtained using the SeDuMi Matlab toolbox\nbsp{}cite:sturm99_using_sedum
[[file:figs/fir_filter_ligo.pdf]]
** Weighting Functions Design
<<sec:ligo_weights>>
The weighting functions should be designed such that their inverse magnitude is as close as possible to the specifications in order to not over-constrain the synthesis problem.
However, the order of each weight should stay reasonably small in order to reduce the computational costs of the optimization problem as well as for the physical implementation of the filters.
A Type I Chebyshev filter of order $20$ is used as the weighting transfer function $w_L(s)$ corresponding to the low pass filter.
For the one corresponding to the high pass filter $w_H(s)$, a $7^{\text{th}}$ order transfer function is designed.
The magnitudes of the weighting functions are shown in Fig. ref:fig:ligo_weights.
A Type I Chebyshev filter of order $20$ is used for the weighting transfer function $W_L(s)$ corresponding to the low pass filter.
For the one corresponding to the high pass filter $W_H(s)$, a $7^{\text{th}}$ order transfer function is designed.
The magnitudes of the weighting functions are shown in Fig.\nbsp{}ref:fig:ligo_weights.
#+name: fig:ligo_weights
#+caption: Specifications and weighting functions magnitudes
#+attr_latex: :scale 1
#+caption: Specifications and weighting functions inverse magnitudes
[[file:figs/ligo_weights.pdf]]
** $\mathcal{H}_\infty$ Synthesis
** $\mathcal{H}_\infty$ Synthesis of the complementary filters
<<sec:ligo_results>>
$\mathcal{H}_\infty$ synthesis is performed using the architecture shown in Fig. ref:eq:generalized_plant.
The complementary filters obtained are of order $27$.
In Fig. ref:fig:comp_fir_ligo_hinf, their bode plot is compared with the FIR filters of order 512 obtained in cite:hua05_low_ligo.
They are found to be very close to each other and this shows the effectiveness of the proposed synthesis method.
The proposed $\mathcal{H}_\infty$ synthesis is performed on the generalized plant shown in Fig.\nbsp{}ref:fig:h_infinity_robust_fusion_plant.
After optimization, the $\mathcal{H}_\infty$ norm from $w$ to $[z_1,\ z_2]^T$ is found close to one indication successful synthesis.
In Fig.\nbsp{}ref:fig:comp_fir_ligo_hinf, the bode plot of the obtained complementary filters are compared with the FIR filters of order 512 obtained in\nbsp{}cite:hua05_low_ligo.
Even though the complementary filters using the $\mathcal{H}_\infty$ synthesis are of much lower order (order 27), they are found to be very close to the FIR filters.
This confirms the effectiveness of the proposed synthesis method even when the complementary filters are subject to complex requirements.
#+name: fig:comp_fir_ligo_hinf
#+caption: Comparison of the FIR filters (solid) designed in cite:hua05_low_ligo with the filters obtained with $\mathcal{H}_\infty$ synthesis (dashed)
#+attr_latex: :scale 1
#+caption: Comparison of the FIR filters (dashed) designed in\nbsp{}cite:hua05_low_ligo with the filters obtained with $\mathcal{H}_\infty$ synthesis (solid)
[[file:figs/comp_fir_ligo_hinf.pdf]]
* Discussion
@ -589,49 +593,41 @@ They are found to be very close to each other and this shows the effectiveness o
** "Closed-Loop" complementary filters
<<sec:closed_loop_complementary_filters>>
*** Introduction to using feedback architecture for CF :ignore:
It is possible to use the fundamental properties of a feedback architecture to generate complementary filters.
It has been proposed by:
- cite:plummer06_optim_compl_filter_their_applic_motion_measur use H-Infinity to optimize complementary filters (flatten the super sensor noise spectral density)
- cite:jensen13_basic_uas design of complementary filters with classical control theory, PID
- Maybe also cite cite:mahony05_compl_filter_desig_special_orthog
Consider the feedback architecture of Figure ref:fig:feedback_sensor_fusion, with two inputs $\hat{x}_1$ and $\hat{x}_2$, and one output $\hat{x}$.
An alternative way to implement complementary filters is by using a fundamental property of the classical feedback architecture shown in Fig.\nbsp{}ref:fig:feedback_sensor_fusion.
This is for instance presented in cite:mahony05_compl_filter_desig_special_orthog,plummer06_optim_compl_filter_their_applic_motion_measur,jensen13_basic_uas.
#+name: fig:feedback_sensor_fusion
#+caption: "Closed-Loop" complementary filters
#+attr_latex: :scale 1
[[file:figs/feedback_sensor_fusion.pdf]]
The output $\hat{x}$ is described by eqref:eq:closed_loop_complementary_filters.
Consider the feedback architecture of Fig.\nbsp{}ref:fig:feedback_sensor_fusion, with two inputs $\hat{x}_1$ and $\hat{x}_2$, and one output $\hat{x}$.
The output $\hat{x}$ is linked to the inputs by eqref:eq:closed_loop_complementary_filters.
#+name: eq:closed_loop_complementary_filters
\begin{equation}
\hat{x} = \underbrace{\frac{1}{1 + L(s)}}_{S(s)} \hat{x}_1 + \underbrace{\frac{L(s)}{1 + L(s)}}_{T(s)} \hat{x}_2
\end{equation}
with the famous relationship
As for any classical feedback architecture, we have that the sum of the sensitivity and complementary sensitivity transfer function is equal to one\nbsp{}eqref:eq:sensitivity_sum.
#+name: eq:sensitivity_sum
\begin{equation}
T(s) + S(s) = 1
\end{equation}
Provided that the closed-loop system is stable, this indeed forms two complementary filters.
*** Sensor Fusion with "closed-loop" complementary filters :ignore:
Therefore, two filters can be merged as shown in Figure ref:fig:feedback_sensor_fusion_arch.
Therefore, provided that the closed-loop system is stable, the closed-loop system in Fig.\nbsp{}ref:fig:feedback_sensor_fusion is corresponding to two complementary filters.
Two sensors can then be merged as shown in Fig.\nbsp{}ref:fig:feedback_sensor_fusion_arch.
#+name: fig:feedback_sensor_fusion_arch
#+caption: Classical feedback architecture for sensor fusion
#+attr_latex: :scale 1
#+caption: Classical feedback architecture used for sensor fusion
[[file:figs/feedback_sensor_fusion_arch.pdf]]
One of the main advantage of this configuration is that standard tools of the linear control theory can be applied.
One of the main advantage of implementing and designing complementary filters using the feedback architecture of Fig.\nbsp{}ref:fig:feedback_sensor_fusion is that all the tools of the linear control theory can be applied for the design of the filters.
If one want to shape both $\frac{\hat{x}}{\hat{x}_1}(s) = S(s)$ and $\frac{\hat{x}}{\hat{x}_2}(s) = T(s)$, the $\mathcal{H}_\infty$ mixed-sensitivity synthesis can be easily applied.
*** Mixed Sensitivity Synthesis :ignore:
If one want to shape both the transfer functions $\frac{\hat{x}}{\hat{x}_1}(s) = S(s)$ and $\frac{\hat{x}}{\hat{x}_2}(s) = T(s)$, this corresponds to the $\mathcal{H}_\infty$ mixed-sensitivity synthesis.
The $\mathcal{H}_\infty$ mixed-sensitivity synthesis can be perform by applying the $\mathcal{H}_\infty$ synthesis to the generalized plant $P_L(s)$ shown in Figure ref:fig:feedback_synthesis_architecture_generalized_plant and described by eqref:eq:generalized_plant_mixed_sensitivity where $W_1(s)$ and $W_2(s)$ are weighting functions used to respectively shape $S(s)$ and $T(s)$.
To do so, weighting functions $W_1(s)$ and $W_2(s)$ are added to respectively shape $S(s)$ and $T(s)$ (Fig.\nbsp{}ref:fig:feedback_synthesis_architecture).
Then the system is re-organized to form the generalized plant $P_L(s)$ shown in Fig.\nbsp{}ref:fig:feedback_synthesis_architecture_generalized_plant.
The $\mathcal{H}_\infty$ mixed-sensitivity synthesis can finally be performed by applying the $\mathcal{H}_\infty$ synthesis to the generalized plant $P_L(s)$ which is described by eqref:eq:generalized_plant_mixed_sensitivity.
#+name: eq:generalized_plant_mixed_sensitivity
\begin{equation}
@ -641,66 +637,64 @@ The $\mathcal{H}_\infty$ mixed-sensitivity synthesis can be perform by applying
\end{bmatrix}
\end{equation}
This is equivalent as to find a filter $L(s)$ such that eqref:eq:comp_filters_feedback_obj is verified.
The output of the synthesis is a filter $L(s)$ such that the "closed-loop" $\mathcal{H}_\infty$ norm from $[w_1,\ w_2]$ to $z$ is less than one eqref:eq:comp_filters_feedback_obj.
#+name: eq:comp_filters_feedback_obj
\begin{equation}
\left\|\begin{matrix} \frac{1}{1 + L(s)} W_1(s) \\ \frac{L(s)}{1 + L(s)} W_2(s) \end{matrix}\right\|_\infty \le 1
\left\| \begin{matrix} \frac{z}{w_1} \\ \frac{z}{w_2} \end{matrix} \right\|_\infty = \left\| \begin{matrix} \frac{1}{1 + L(s)} W_1(s) \\ \frac{L(s)}{1 + L(s)} W_2(s) \end{matrix} \right\|_\infty \le 1
\end{equation}
The sensor fusion can be implemented as shown in Figure ref:fig:feedback_sensor_fusion_arch using the feedback architecture or more classically as shown in Figure ref:fig:sensor_fusion_overview using eqref:eq:comp_filters_feedback.
If the synthesis is successful, two complementary filters are obtained with their magnitudes bounded by the inverse magnitudes of the weighting functions.
The sensor fusion can then be implemented as shown in Fig.\nbsp{}ref:fig:feedback_sensor_fusion_arch using the feedback architecture or more classically as shown in Fig.\nbsp{}ref:fig:sensor_fusion_overview by defining the two complementary filters as in\nbsp{}eqref:eq:comp_filters_feedback.
#+name: eq:comp_filters_feedback
\begin{equation}
H_1(s) = \frac{1}{1 + L(s)}; \quad H_2(s) = \frac{L(s)}{1 + L(s)}
\end{equation}
The two being equivalent considering only the inputs/outputs relationships.
The two architecture are equivalent regarding their inputs/outputs relationships.
#+name: fig:feedback_synthesis_architecture_generalized_plant
#+caption: Generalized plant for the $\mathcal{H}_\infty$ mixed-sensitivity synthesis
#+attr_latex: :scale 1
[[file:figs/feedback_synthesis_architecture_generalized_plant.pdf]]
#+begin_export latex
\begin{figure}[htbp]
\begin{subfigure}[t]{0.6\linewidth}
\centering
\includegraphics[scale=1]{figs/feedback_synthesis_architecture.pdf}
\caption{\label{fig:feedback_synthesis_architecture} Feedback architecture with included weights}
\end{subfigure}
\hfill
\begin{subfigure}[t]{0.4\linewidth}
\centering
\includegraphics[scale=1]{figs/feedback_synthesis_architecture_generalized_plant.pdf}
\caption{\label{fig:feedback_synthesis_architecture_generalized_plant} Generalized plant}
\end{subfigure}
\hfill
\caption{\label{fig:h_inf_mixed_sensitivity_synthesis} $\mathcal{H}_\infty$ mixed-sensitivity synthesis}
\centering
\end{figure}
#+end_export
*** Example and equivalence with our synthesis method :ignore:
As an example, two "closed-loop" complementary filters are designed using the $\mathcal{H}_\infty$ mixed-sensitivity synthesis.
The weighting functions are designed using formula\nbsp{}eqref:eq:weight_formula with parameters shown in Table\nbsp{}ref:tab:weights_params.
After synthesis, a filter $L(s)$ is obtained, its magnitude is shown in Fig.\nbsp{}ref:fig:hinf_filters_results_mixed_sensitivity by the dashed line.
The "closed-loop" complementary filters are compared with the inverse magnitude of the weighting functions in Fig.\nbsp{}ref:fig:hinf_filters_results_mixed_sensitivity confirming that the synthesis is successful.
The obtained "closed-loop" complementary filters are indeed equal to the ones obtained in Section\nbsp{}ref:sec:hinf_example.
Example: same weights as in ref:tab:weights_params.
#+name: fig:hinf_filters_results_mixed_sensitivity
#+caption: Bode plot of the obtained complementary filters after $\mathcal{H}_\infty$ mixed-sensitivity synthesis
[[file:figs/hinf_filters_results_mixed_sensitivity.pdf]]
Therefore, complementary filter design is very similar to mixed-sensitivity synthesis.
They are actually equivalent by taking
\begin{equation}
L = H_H^{-1} - 1
\end{equation}
(provided $H_H$ is invertible, therefore bi-proper)
** Imposing zero at origin / roll-off
<<sec:add_features_in_filters>>
3 methods:
Link to literature about doing that with mixed sensitivity
** Synthesis of Three Complementary Filters
** Synthesis of more than two Complementary Filters
<<sec:hinf_three_comp_filters>>
*** Why it is used sometimes :ignore:
Some applications may require to merge more than two sensors.
For instance at the LIGO, three sensors (an LVDT, a seismometer and a geophone) are merged to form a super sensor (Figure ref:fig:ligo_super_sensor_architecture). \par
Some applications may require to merge more than two sensors\nbsp{}cite:stoten01_fusion_kinet_data_using_compos_filter,becker15_compl_filter_desig_three_frequen_bands.
For instance at the LIGO\nbsp{}cite:matichard15_seism_isolat_advan_ligo, three sensors (an LVDT, a seismometer and a geophone) are merged to form a super sensor (Fig.\nbsp{}ref:fig:ligo_super_sensor_architecture).\par
- [ ] cite:becker15_compl_filter_desig_three_frequen_bands
*** Sequential vs Parallel :ignore:
When merging $n>2$ sensors using complementary filters, two architectures can be used as shown in Figure ref:fig:sensor_fusion_three.
The fusion can either be done in a "sequential" way where $n-1$ sets of two complementary filters are used (Figure ref:fig:sensor_fusion_three_sequential), or in a "parallel" way where one set of $n$ complementary filters is used (Figure ref:fig:sensor_fusion_three_parallel).
When merging $n>2$ sensors using complementary filters, two architectures can be used as shown in Fig.\nbsp{}ref:fig:sensor_fusion_three.
The fusion can either be done in a "sequential" way where $n-1$ sets of two complementary filters are used (Fig.\nbsp{}ref:fig:sensor_fusion_three_sequential), or in a "parallel" way where one set of $n$ complementary filters is used (Fig.\nbsp{}ref:fig:sensor_fusion_three_parallel).
In the first case, typical sensor fusion synthesis techniques can be used.
However, when a parallel architecture is used, a new synthesis method for a set of more than two complementary filters is required.
Such synthesis method is presented in this section. \par
*************** TODO Say possible advantages of parallel architecture
*************** END
However, when a parallel architecture is used, a new synthesis method for a set of more than two complementary filters is required as only simple analytical formulas have been proposed in the literature\nbsp{cite:stoten01_fusion_kinet_data_using_compos_filter,becker15_compl_filter_desig_three_frequen_bands}.
A generalization of the proposed synthesis method of complementary filters is presented in this section. \par
#+begin_export latex
\begin{figure}[htbp]
@ -715,13 +709,12 @@ Such synthesis method is presented in this section. \par
\includegraphics[scale=1]{figs/sensor_fusion_three_parallel.pdf}
\caption{\label{fig:sensor_fusion_three_parallel}Parallel fusion}
\end{subfigure}
\caption{\label{fig:sensor_fusion_three}Sensor fusion architecture with more than two sensors}
\caption{\label{fig:sensor_fusion_three}Possible sensor fusion architecture when more than two sensors are to be merged}
\centering
\end{figure}
#+end_export
*** Mathematical Problem :ignore:
The synthesis objective is to compute a set of $n$ stable transfer functions $[H_1(s),\ H_2(s),\ \dots,\ H_n(s)]$ such that eqref:eq:hinf_problem_gen is satisfied.
The synthesis objective is to compute a set of $n$ stable transfer functions $[H_1(s),\ H_2(s),\ \dots,\ H_n(s)]$ such that conditions\nbsp{}eqref:eq:hinf_cond_compl_gen and\nbsp{}eqref:eq:hinf_cond_perf_gen are satisfied.
#+name: eq:hinf_problem_gen
\begin{subequations}
\begin{align}
@ -731,22 +724,37 @@ The synthesis objective is to compute a set of $n$ stable transfer functions $[H
\end{subequations}
where $[W_1(s),\ W_2(s),\ \dots,\ W_n(s)]$ are weighting transfer functions that are chosen to specify the maximum wanted norms of the complementary filters during the synthesis.
Such synthesis objective is very close to the one described in Section ref:sec:synthesis_objective, and indeed the proposed synthesis architecture is also very similar. \par
Such synthesis objective is very close to the one described in Section\nbsp{}ref:sec:synthesis_objective, and indeed the proposed synthesis method is a generalization of the one presented in Section\nbsp{}ref:sec:hinf_synthesis. \par
*** H-Infinity Architecture :ignore:
Consider the generalized plant $P_3(s)$ shown in Figure ref:fig:comp_filter_three_hinf which is also described by eqref:eq:generalized_plant_three_filters.
Before presenting the generalized synthesis method, the case with three sensors is presented.
Consider the generalized plant $P_3(s)$ shown in Fig.\nbsp{}ref:fig:comp_filter_three_hinf_gen_plant which is also described by\nbsp{}eqref:eq:generalized_plant_three_filters.
#+name: eq:generalized_plant_three_filters
\begin{equation}
\begin{bmatrix} z_1 \\ z_2 \\ z_3 \\ v \end{bmatrix} = P_3(s) \begin{bmatrix} w \\ u_1 \\ u_2 \end{bmatrix}; \quad P_3(s) = \begin{bmatrix}W_1(s) & -W_1(s) & -W_1(s) \\ 0 & \phantom{+}W_2(s) & 0 \\ 0 & 0 & \phantom{+}W_3(s) \\ 1 & 0 & 0 \end{bmatrix}
\end{equation}
#+name: fig:comp_filter_three_hinf
#+caption: Architecture for $\mathcal{H}_\infty$ synthesis of three complementary filters
#+attr_latex: :scale 1
[[file:figs/comp_filter_three_hinf.pdf]]
#+begin_export latex
\begin{figure}[htbp]
\begin{subfigure}[t]{0.5\linewidth}
\centering
\includegraphics[scale=1]{figs/comp_filter_three_hinf_gen_plant.pdf}
\caption{\label{fig:comp_filter_three_hinf_gen_plant} Generalized plant}
\vfill
\end{subfigure}
\hfill
\begin{subfigure}[t]{0.5\linewidth}
\centering
\includegraphics[scale=1]{figs/comp_filter_three_hinf_fb.pdf}
\caption{\label{fig:comp_filter_three_hinf_fb} Generalized plant with the synthesized filter}
\end{subfigure}
\hfill
\caption{\label{fig:comp_filter_three_hinf} Architecture for the $\mathcal{H}_\infty$ synthesis of three complementary filters}
\centering
\end{figure}
#+end_export
Applying the $\mathcal{H}_\infty$ synthesis on the generalized plant $P_3(s)$ is equivalent as to find two stable filters $[H_2(s),\ H_3(s)]$ (shown in Figure ref:fig:comp_filter_three_hinf) such that the $\mathcal{H}_\infty$ norm of the transfer function from $w$ to $[z_1,\ z_2, \ z_3]$ is less than one eqref:eq:hinf_syn_obj_three.
Applying the $\mathcal{H}_\infty$ synthesis on the generalized plant $P_3(s)$ is equivalent as to find two stable filters $[H_2(s),\ H_3(s)]$ (shown in Fig.\nbsp{}ref:fig:comp_filter_three_hinf_fb) such that the $\mathcal{H}_\infty$ norm of the transfer function from $w$ to $[z_1,\ z_2, \ z_3]$ is less than one\nbsp{}eqref:eq:hinf_syn_obj_three.
#+name: eq:hinf_syn_obj_three
\begin{equation}
@ -755,18 +763,17 @@ Applying the $\mathcal{H}_\infty$ synthesis on the generalized plant $P_3(s)$ is
By defining $H_1(s) \triangleq 1 - H_2(s) - H_3(s)$, the proposed $\mathcal{H}_\infty$ synthesis solves the design problem eqref:eq:hinf_problem_gen with $n=3$. \par
*** Example of generated complementary filters :ignore:
An example is given to validate the method where three sensors are used in different frequency bands (up to $\SI{1}{Hz}$, from $1$ to $\SI{10}{Hz}$ and above $\SI{10}{Hz}$ respectively).
Three weighting functions are designed using eqref:eq:weight_formula and shown by dashed curves in Fig. ref:fig:three_complementary_filters_results.
The bode plots of the obtained complementary filters are shown in Fig. ref:fig:three_complementary_filters_results. \par
An example is given to validate the method where three sensors are used in different frequency bands.
For instance a displacement sensor from DC up to $\SI{1}{Hz}$, a geophone from $1$ to $\SI{10}{Hz}$ and an accelerometer above $\SI{10}{Hz}$.
Three weighting functions are designed using formula\nbsp{}eqref:eq:weight_formula and their inverse magnitudes are shown in Fig.\nbsp{}ref:fig:three_complementary_filters_results (dashed curves).
The $\mathcal{H}_\infty$ synthesis is performed on the generalized plant $P_3(s)$ and the bode plot of the obtained complementary filters are shown in Fig.\nbsp{}ref:fig:three_complementary_filters_results (solid lines). \par
#+name: fig:three_complementary_filters_results
#+caption: Frequency response of the weighting functions and three complementary filters obtained using $\mathcal{H}_\infty$ synthesis
#+attr_latex: :scale 1
#+caption: Bode plot of the inverse weighting functions and of the three complementary filters obtained using the $\mathcal{H}_\infty$ synthesis
[[file:figs/three_complementary_filters_results.pdf]]
*** Generalization :ignore:
Such synthesis method can be generalized to a set of $n$ complementary filters, even though there might not be any practical application for $n>3$.
Even though there might not be any practical application for a set of more than 3 complementary filters, it can still be designed using the same procedure.
A set of $n$ complementary filters can be shaped using the generalized plant $P_n(s)$ described by\nbsp{}eqref:eq:generalized_plant_n_filters.
#+name: eq:generalized_plant_n_filters
\begin{equation}
@ -783,16 +790,32 @@ Such synthesis method can be generalized to a set of $n$ complementary filters,
* Conclusion
<<sec:conclusion>>
This paper has shown how complementary filters can be used to combine multiple sensors in order to obtain a super sensor.
Typical specification on the super sensor noise and on the robustness of the sensor fusion has been shown to be linked to the norm of the complementary filters.
Therefore, a synthesis method that permits the shaping of the complementary filters norms has been proposed and has been successfully applied for the design of complex filters.
Future work will aim at further developing this synthesis method for the robust and optimal synthesis of complementary filters used in sensor fusion.
Sensors measuring a physical quantities are always subject to limitations both in terms of bandwidth or accuracy.
Complementary filters can be used to fuse multiple sensors with different characteristics in order to combine their benefits and yield a better estimate of the measured physical quantity.
The sensor fusion robustness and the obtained super sensor noise can be linked to the magnitude of the complementary filters.
In this paper, a synthesis method that enables the shaping of the complementary filters norms has been proposed.
Several example were used to emphasize the simplicity and the effectiveness of the proposed method.
Links with "closed-loop" complementary filters where highlighted, and the proposed method was generalized for the design of a set of more than two complementary filters.
Future work will aim at developing a complementary filter synthesis method that minimizes the super sensor noise while ensuring the robustness of the fusion.
* Acknowledgment
:PROPERTIES:
:UNNUMBERED: t
:END:
This research benefited from a FRIA grant from the French Community of Belgium.
This paper is based on a paper previously presented at the ICCMA conference\nbsp{}cite:dehaeze19_compl_filter_shapin_using_synth.
* Data Availability
:PROPERTIES:
:UNNUMBERED: t
:END:
Matlab cite:matlab20 was used for this study.
The source code is available under a MIT License and archived in Zenodo\nbsp{}cite:dehaeze21_new_method_desig_compl_filter_code.
* Bibliography :ignore:
\bibliographystyle{elsarticle-num}

BIN
journal/dehaeze21_desig_compl_filte.pdf
View File

Binary file not shown.

580
journal/dehaeze21_desig_compl_filte.tex
View File

@ -1,4 +1,4 @@
% Created 2021-08-27 ven. 11:24
% Created 2021-08-31 mar. 14:16
% Intended LaTeX compiler: pdflatex
\documentclass[preprint, sort&compress]{elsarticle}
\usepackage[utf8]{inputenc}
@ -14,14 +14,14 @@
\usepackage{amssymb}
\usepackage{capt-of}
\usepackage{hyperref}
\usepackage{subcaption}
\usepackage{caption,subcaption}
\usepackage{amsfonts}
\usepackage{siunitx}
\journal{Mechanical Systems and Signal Processing}
\author[a1,a2]{Thomas Dehaeze\corref{cor1}}
\author[a3,a4]{Mohit Verma}
\author[a2,a4]{Christophe Collette}
\cortext[cor1]{Corresponding author. Email Address: dehaeze.thomas@gmail.com}
\cortext[cor1]{Corresponding author. Email Address: thomas.dehaeze@esrf.fr}
\address[a1]{European Synchrotron Radiation Facility, Grenoble, France}
\address[a2]{University of Li\`{e}ge, Department of Aerospace and Mechanical Engineering, 4000 Li\`{e}ge, Belgium.}
\address[a3]{CSIR --- Structural Engineering Research Centre, Taramani, Chennai --- 600113, India.}
@ -32,8 +32,9 @@
\usepackage[hyperref]{xcolor}
\usepackage[top=2cm, bottom=2cm, left=2cm, right=2cm]{geometry}
\hypersetup{colorlinks=true}
\setlength{\parskip}{1em}
\date{}
\title{A new method of designing complementary filters for sensor fusion using \(\mathcal{H}_\infty\) synthesis}
\title{A new method of designing complementary filters for sensor fusion using the \(\mathcal{H}_\infty\) synthesis}
\begin{document}
@ -41,170 +42,134 @@
\begin{frontmatter}
\begin{abstract}
Sensor have limited bandwidth and are accurate only in a certain frequency band.
In many applications, the signals of different sensor are fused together in order to either enhance the stability or improve the operational bandwidth of the system.
The sensor signals can be fused using complementary filters.
The tuning of complementary filters is a complex task and is the subject of this paper.
The filters needs to meet design specifications while satisfying the complementary property.
This paper presents a framework to shape the norm of complementary filters using the \(\mathcal{H}_\infty\) norm minimization.
The design specifications are imposed as constraints in the optimization problem by appropriate selection of weighting functions.
The proposed method is quite general and easily extendable to cases where more than two sensors are fused.
Finally, the proposed method is applied to the design of complementary filter design for active vibration isolation of the Laser Interferometer Gravitation-wave Observatory (LIGO).
In order to obtain a better estimate of a quantity being measured, several sensors having different characteristics can be merged with a technique called ``sensor fusion''.
The obtained ``super sensor'' can combine the benefits of the individual sensors provided that the complementary filters used in the fusion are well designed.
Indeed, properties of the super sensor are linked to the magnitude of the complementary filters.
Properly shaping the magnitude of complementary filters is a difficult and time-consuming task.
In this study, we address this issue and propose a new method for designing complementary filters.
This method uses weighting functions to specify the wanted shape of the complementary filter that are then easily obtained using the standard \(\mathcal{H}_\infty\) synthesis.
The proper choice of the weighting functions is discussed, and the effectiveness and simplicity of the design method is highlighted using several examples.
Such synthesis method is further extended for the shaping of more than two complementary filters.
\end{abstract}
\begin{keyword}
Sensor fusion \sep{} Optimal filters \sep{} \(\mathcal{H}_\infty\) synthesis \sep{} Vibration isolation \sep{} Precision
Sensor fusion \sep{} Complementary filters \sep{} \(\mathcal{H}_\infty\) synthesis \sep{} Vibration isolation \sep{} Motion control
\end{keyword}
\end{frontmatter}
\section{Introduction}
\label{sec:org5737795}
\label{sec:orgb3159e9}
\label{sec:introduction}
\begin{itemize}
\item \cite{anderson53_instr_approac_system_steer_comput} earliest application of complementary filters (A simple RC circuit was used to physically realize the complementary filters)
\item \cite{bendat57_optim_filter_indep_measur_two} roots of sensor fusion
\end{itemize}
Sensor fusion can have many advantages.
Measuring a physical quantity using sensors is always subject to several limitations.
First, the accuracy of the measurement will be affected by several noise sources, such as the electrical noise of the conditioning electronics being used.
Second, the frequency range in which the measurement is relevant is bounded by the bandwidth of the sensor.
One way to overcome these limitations is to combine several sensors using a technique called ``sensor fusion''~\cite{bendat57_optim_filter_indep_measur_two}.
Fortunately, a wide variety of sensors exist, each with different characteristics.
By carefully choosing the fused sensors, a so called ``super sensor'' is obtain that combines benefits of individual sensors and yields a better estimate of the measured physical quantity. \par
In some situations, sensor fusion is used to increase the bandwidth of the measurement~\cite{shaw90_bandw_enhan_posit_measur_using_measur_accel,zimmermann92_high_bandw_orien_measur_contr,min15_compl_filter_desig_angle_estim}.
For instance, in~\cite{shaw90_bandw_enhan_posit_measur_using_measur_accel}, the bandwidth of a position sensor is increased by fusing it with an accelerometer providing the high frequency motion information.
For other applications, sensor fusion is used to obtain a estimate of the measured quantity with lower noise~\cite{hua05_low_ligo,hua04_polyp_fir_compl_filter_contr_system,plummer06_optim_compl_filter_their_applic_motion_measur,robert12_introd_random_signal_applied_kalman}.
More recently, the fusion of sensors measuring different physical quantities has been proposed to obtained interesting properties for control~\cite{collette15_sensor_fusion_method_high_perfor,yong16_high_speed_vertic_posit_stage}.
In~\cite{collette15_sensor_fusion_method_high_perfor}, an inertial sensor used for active vibration isolation is fused with a sensor collocated with the actuator to improve the stability margins of the feedback controller. \par
Practical applications of sensor fusion are numerous.
It is widely used for the attitude estimation of several autonomous vehicles such as unmanned aerial vehicle~\cite{baerveldt97_low_cost_low_weigh_attit,corke04_inert_visual_sensin_system_small_auton_helic,jensen13_basic_uas} and underwater vehicles~\cite{pascoal99_navig_system_desig_using_time,batista10_optim_posit_veloc_navig_filter_auton_vehic}.
Naturally, it is of great benefits for high performance positioning control as shown in~\cite{shaw90_bandw_enhan_posit_measur_using_measur_accel,zimmermann92_high_bandw_orien_measur_contr,min15_compl_filter_desig_angle_estim,yong16_high_speed_vertic_posit_stage}.
Sensor fusion was also shown to be a key technology to improve the performances of active vibration isolation systems~\cite{tjepkema12_sensor_fusion_activ_vibrat_isolat_precis_equip}.
This is particularly apparent for the isolation stages of gravitational wave observer~\cite{collette15_sensor_fusion_method_high_perfor,heijningen18_low} such as the ones used at the LIGO~\cite{hua05_low_ligo,hua04_polyp_fir_compl_filter_contr_system} and at the VIRGO~\cite{lucia18_low_frequen_optim_perfor_advan}. \par
There are mainly two ways to perform sensor fusion: either using a set of complementary filters~\cite{anderson53_instr_approac_system_steer_comput} or using Kalman filtering~\cite{brown72_integ_navig_system_kalman_filter,odry18_kalman_filter_mobil_robot_attit_estim}.
For sensor fusion applications, both methods are sharing many relationships~\cite{brown72_integ_navig_system_kalman_filter,higgins75_compar_compl_kalman_filter,robert12_introd_random_signal_applied_kalman,becker15_compl_filter_desig_three_frequen_bands}.
However, for Kalman filtering, we are forced to make assumption about the probabilistic character of the sensor noises~\cite{robert12_introd_random_signal_applied_kalman} whereas it is not the case for complementary filters.
Furthermore, the main advantages of complementary filters over Kalman filtering for sensor fusion are their very general applicability, their low computational cost~\cite{higgins75_compar_compl_kalman_filter}, and the fact that they are very intuitive as their effects can be easily interpreted in the frequency domain. \par
A set of filters is said to be complementary if the sum of their transfer functions is equal to one at all frequencies.
For the earliest used of complementary filtering for sensor fusion, analog circuits were used to physically realize the filters~\cite{anderson53_instr_approac_system_steer_comput}.
Analog complementary filters are still used today~\cite{yong16_high_speed_vertic_posit_stage,moore19_capac_instr_sensor_fusion_high_bandw_nanop}, but most of the time they are now implemented numerically as it allows for much more flexibility. \par
In some situations, it is used to increase the bandwidth of the sensor \cite{shaw90_bandw_enhan_posit_measur_using_measur_accel,zimmermann92_high_bandw_orien_measur_contr,min15_compl_filter_desig_angle_estim}.
For instance by increasing the high frequency bandwidth of a position sensor using an accelerometer.
Several design methods have been developed over the years to optimize complementary filters.
The easiest way to design complementary filters is to use analytical formulas.
Depending on the application, the formulas used are of first order~\cite{corke04_inert_visual_sensin_system_small_auton_helic,yeh05_model_contr_hydraul_actuat_two,yong16_high_speed_vertic_posit_stage}, second order~\cite{baerveldt97_low_cost_low_weigh_attit,stoten01_fusion_kinet_data_using_compos_filter,jensen13_basic_uas} or even higher orders~\cite{shaw90_bandw_enhan_posit_measur_using_measur_accel,zimmermann92_high_bandw_orien_measur_contr,stoten01_fusion_kinet_data_using_compos_filter,collette15_sensor_fusion_method_high_perfor,matichard15_seism_isolat_advan_ligo}. \par
Decrease the noise: \cite{hua05_low_ligo,hua04_polyp_fir_compl_filter_contr_system,plummer06_optim_compl_filter_their_applic_motion_measur}
\cite[chapter 8]{robert12_introd_random_signal_applied_kalman}
Increased robustness (sensor measuring different quantities): \cite{collette15_sensor_fusion_method_high_perfor,yong16_high_speed_vertic_posit_stage}
\par
The applications of sensor fusion are numerous.
It is widely used for attitude estimation of unmanned aerial vehicle
\cite{baerveldt97_low_cost_low_weigh_attit,pascoal99_navig_system_desig_using_time,corke04_inert_visual_sensin_system_small_auton_helic,batista10_optim_posit_veloc_navig_filter_auton_vehic,jensen13_basic_uas,min15_compl_filter_desig_angle_estim}
Motion control
\cite{shaw90_bandw_enhan_posit_measur_using_measur_accel,zimmermann92_high_bandw_orien_measur_contr}
Tjepkema et al. \cite{tjepkema12_sensor_fusion_activ_vibrat_isolat_precis_equip} used sensor fusion to isolate precision equipment from the ground motion.
Gravitational wave observer \cite{heijningen18_low}:
LIGO \cite{hua05_low_ligo,hua04_polyp_fir_compl_filter_contr_system}
VIRGO \cite{lucia18_low_frequen_optim_perfor_advan}
There are mainly two ways to perform sensor fusion: using complementary filters or using Kalman filtering \cite{brown72_integ_navig_system_kalman_filter}.
Kalman filtering \cite{odry18_kalman_filter_mobil_robot_attit_estim}
Relations between CF and Kalman: \cite{becker15_compl_filter_desig_three_frequen_bands}
Advantages of complementary filtering over Kalman filtering for sensor fusion:
\begin{itemize}
\item Less computation \cite{higgins75_compar_compl_kalman_filter}
\item For Kalman filtering, we are forced to make assumption about the probabilistic character of the sensor noises \cite{robert12_introd_random_signal_applied_kalman}
\item More intuitive frequency domain technique
\end{itemize}
In some cases, complementary filters are implemented in an analog way such as in \cite{yong16_high_speed_vertic_posit_stage,moore19_capac_instr_sensor_fusion_high_bandw_nanop}, but most of the time it is implemented numerically which allows much more complex
Multiple design methods have been used for complementary filters
\begin{itemize}
\item Analytical methods:
\begin{itemize}
\item first order: \cite{corke04_inert_visual_sensin_system_small_auton_helic,yong16_high_speed_vertic_posit_stage}
\item second order: \cite{baerveldt97_low_cost_low_weigh_attit}, \cite{stoten01_fusion_kinet_data_using_compos_filter}, \cite{jensen13_basic_uas}
\item higher order: \cite{shaw90_bandw_enhan_posit_measur_using_measur_accel}, \cite{zimmermann92_high_bandw_orien_measur_contr}, \cite{collette15_sensor_fusion_method_high_perfor}, \cite{matichard15_seism_isolat_advan_ligo}
\end{itemize}
\item \cite{pascoal99_navig_system_desig_using_time} use LMI to generate complementary filters (convex optimization techniques), specific for navigation systems
\item \cite{hua05_low_ligo,hua04_polyp_fir_compl_filter_contr_system}: FIR + convex optimization
\item Similar to feedback system:
\begin{itemize}
\item \cite{plummer06_optim_compl_filter_their_applic_motion_measur} use H-Infinity to optimize complementary filters (flatten the super sensor noise spectral density)
\item \cite{jensen13_basic_uas} design of complementary filters with classical control theory, PID
\end{itemize}
\item 3 complementary filters: \cite{becker15_compl_filter_desig_three_frequen_bands}
\end{itemize}
\begin{itemize}
\item Robustness problems: \cite{zimmermann92_high_bandw_orien_measur_contr,plummer06_optim_compl_filter_their_applic_motion_measur} change of phase near the merging frequency
\item Trial and error
\item Although many design methods of complementary filters have been proposed in the literature, no simple method that allows to shape the norm of the complementary filters is available.
\end{itemize}
Most of the requirements => shape of the complementary filters
=> propose a way to shape complementary filters.
Section \ref{sec:requirements}
Section \ref{sec:hinf_method}
Section \ref{sec:application_ligo}
Section \ref{sec:discussion}
As the characteristics of the ``super sensor'' depends on the design of the complementary filters~\cite{dehaeze19_compl_filter_shapin_using_synth}, several optimization techniques were developed over the years.
Some are based on the finding the optimal parameters of analytical formulas~\cite{jensen13_basic_uas,min15_compl_filter_desig_angle_estim,becker15_compl_filter_desig_three_frequen_bands}, while other are using convex optimization tools~\cite{hua04_polyp_fir_compl_filter_contr_system,hua05_low_ligo} such as linear matrix inequalities~\cite{pascoal99_navig_system_desig_using_time}.
As shown in~\cite{plummer06_optim_compl_filter_their_applic_motion_measur}, the design of complementary filters can also be linked to the standard mixed-sensitivity control problem.
Therefore, all the powerful tools developed for the classical control theory can also be used for the design of complementary filters.
For instance, in~\cite{jensen13_basic_uas} the two gains of a Proportional Integral (PI) controller are optimized to minimize the noise of the super sensor. \par
The common objective to all these complementary filters design methods is to obtain a super sensor that has desired characteristics, usually in terms of noise and dynamics.
Moreover, as reported in~\cite{zimmermann92_high_bandw_orien_measur_contr,plummer06_optim_compl_filter_their_applic_motion_measur}, phase shifts and magnitude bumps of the ``super sensors'' dynamics can be observed if either the complementary filters are poorly designed or if the sensors are not well calibrated.
Therefore, the robustness of the fusion is also of concerned when designing the complementary filters.
Although many design methods of complementary filters have been proposed in the literature, no simple method that allows to specify the wanted super sensor characteristic while ensuring good fusion robustness has been proposed. \par
Fortunately, both the robustness of the fusion and the super sensor characteristics can be linked to the magnitude of the complementary filters~\cite{dehaeze19_compl_filter_shapin_using_synth}.
Based on that, this paper introduces a new way to design complementary filters using the \(\mathcal{H}_\infty\) synthesis which allows to shape the magnitude of the complementary filters in an easy and intuitive way. \par
Section~\ref{sec:requirements} introduces the sensor fusion architecture and demonstrates how typical requirements can be linked to the complementary filters magnitudes.
In section~\ref{sec:hinf_method}, the shaping of complementary filters is written as an \(\mathcal{H}_\infty\) optimization problem using weighting functions, and the simplicity of the proposed method is illustrated with an example.
The synthesis method is further validated in Section~\ref{sec:application_ligo} by designing complex complementary filters.
Section~\ref{sec:discussion} compares the proposed synthesis method with the classical mixed-sensitivity synthesis, and extends it to the shaping of more than two complementary filters.
\section{Sensor Fusion and Complementary Filters Requirements}
\label{sec:orgbd86d49}
\label{sec:orge37bf43}
\label{sec:requirements}
Complementary filters provides a framework for fusing signals from different sensors.
As the effectiveness of the fusion depends on the proper design of the complementary filters, they are expected to fulfill certain requirements.
These requirements are discussed in this section.
\subsection{Sensor Fusion Architecture}
\label{sec:org56b9e47}
\label{sec:orgd0086e9}
\label{sec:sensor_fusion}
A general sensor fusion architecture using complementary filters is shown in Figure \ref{fig:sensor_fusion_overview} where several sensors (here two) are measuring the same physical quantity \(x\).
The two sensors output signals are estimates \(\hat{x}_1\) and \(\hat{x}_2\) of \(x\).
A general sensor fusion architecture using complementary filters is shown in Fig.~\ref{fig:sensor_fusion_overview} where several sensors (here two) are measuring the same physical quantity \(x\).
The two sensors output signals \(\hat{x}_1\) and \(\hat{x}_2\) are estimates of \(x\).
Each of these estimates are then filtered out by complementary filters and combined to form a new estimate \(\hat{x}\).
The resulting sensor, termed as ``super sensor'', can have larger bandwidth and better noise characteristics in comparison to the individual sensor.
The resulting sensor, termed as super sensor, can have larger bandwidth and better noise characteristics in comparison to the individual sensor.
This means that the super sensor provides an estimate \(\hat{x}\) of \(x\) which can be more accurate over a larger frequency band than the outputs of the individual sensors.
\begin{figure}[htbp]
\centering
\includegraphics[scale=1]{figs/sensor_fusion_overview.pdf}
\caption{\label{fig:sensor_fusion_overview}Schematic of a sensor fusion architecture}
\caption{\label{fig:sensor_fusion_overview}Schematic of a sensor fusion architecture using complementary filters}
\end{figure}
The complementary property of filters \(H_1(s)\) and \(H_2(s)\) implies that the summation of their transfer functions is equal to unity.
The complementary property of filters \(H_1(s)\) and \(H_2(s)\) implies that the sum of their transfer functions is equal to unity.
That is, unity magnitude and zero phase at all frequencies.
Therefore, a pair of strict complementary filter needs to satisfy the following condition:
Therefore, a pair of complementary filter needs to satisfy the following condition:
\begin{equation}
\label{eq:comp_filter}
H_1(s) + H_2(s) = 1
\end{equation}
It will soon become clear why the complementary property is important.
It will soon become clear why the complementary property is important for the sensor fusion architecture.
\subsection{Sensor Models and Sensor Normalization}
\label{sec:org684f136}
\label{sec:orgf4072c3}
\label{sec:sensor_models}
In order to study such sensor fusion architecture, a model of the sensors is required.
Such model is shown in Figure \ref{fig:sensor_model} and consists of a linear time invariant (LTI) system \(G_i(s)\) representing the dynamics of the sensor and an additive noise input \(n_i\) representing its noise.
In order to study such sensor fusion architecture, a model for the sensors is required.
Such model is shown in Fig.~\ref{fig:sensor_model} and consists of a linear time invariant (LTI) system \(G_i(s)\) representing the sensor dynamics and an input \(n_i\) representing the sensor noise.
The model input \(x\) is the measured physical quantity and its output \(\tilde{x}_i\) is the ``raw'' output of the sensor.
Before filtering the sensor outputs \(\tilde{x}_i\) by the complementary filters, the sensors are usually normalized to simplify the fusion.
This normalization consists of first obtaining an estimate \(\hat{G}_i(s)\) of the sensor dynamics \(G_i(s)\).
It is supposed that the estimate of the sensor dynamics \(\hat{G}_i(s)\) can be inverted and that its inverse \(\hat{G}_i^{-1}(s)\) is proper and stable.
The raw output of the sensor \(\tilde{x}_i\) is then passed through \(\hat{G}_i^{-1}(s)\) as shown in Figure \ref{fig:sensor_model_calibrated}.
This normalization consists of using an estimate \(\hat{G}_i(s)\) of the sensor dynamics \(G_i(s)\), and filtering the sensor output by the inverse of this estimate \(\hat{G}_i^{-1}(s)\) as shown in Fig.~\ref{fig:sensor_model_calibrated}.
It is here supposed that the sensor inverse \(\hat{G}_i^{-1}(s)\) is proper and stable.
This way, the units of the estimates \(\hat{x}_i\) are equal to the units of the physical quantity \(x\).
The sensor dynamics estimate \(\hat{G}_1(s)\) can be a simple gain or more complex transfer functions.
The sensor dynamics estimate \(\hat{G}_1(s)\) can be a simple gain or a more complex transfer function.
\begin{figure}[htbp]
\begin{subfigure}[b]{0.49\linewidth}
\centering
\includegraphics[scale=1]{figs/sensor_model.pdf}
\caption{\label{fig:sensor_model} Basic sensor model consisting of a noise input $n_i$ and a dynamics $G_i(s)$}
\caption{\label{fig:sensor_model} Basic sensor model consisting of a noise input $n_i$ and a linear time invariant transfer function $G_i(s)$}
\end{subfigure}
\hfill
\begin{subfigure}[b]{0.49\linewidth}
\centering
\includegraphics[scale=1]{figs/sensor_model_calibrated.pdf}
\caption{\label{fig:sensor_model_calibrated} Calibrated sensors using the inverse of an estimate $\hat{G}_1(s)$ of the sensor dynamics}
\caption{\label{fig:sensor_model_calibrated} Normalized sensors using the inverse of an estimate $\hat{G}_i(s)$ of the sensor dynamics}
\end{subfigure}
\caption{\label{fig:sensor_models}Sensor models with an without normalization}
\caption{\label{fig:sensor_models}Sensor models with and without normalization}
\centering
\end{figure}
Two calibrated sensors and then combined to form a super sensor as shown in Figure \ref{fig:fusion_super_sensor}.
Two normalized sensors are then combined to form a super sensor as shown in Fig.~\ref{fig:fusion_super_sensor}.
The two sensors are measuring the same physical quantity \(x\) with dynamics \(G_1(s)\) and \(G_2(s)\), and with \emph{uncorrelated} noises \(n_1\) and \(n_2\).
The normalized signals from both calibrated sensors are fed into two complementary filters \(H_1(s)\) and \(H_2(s)\) and then combined to yield an estimate \(\hat{x}\) of \(x\) as shown in Fig. \ref{fig:fusion_super_sensor}.
The signals from both normalized sensors are fed into two complementary filters \(H_1(s)\) and \(H_2(s)\) and then combined to yield an estimate \(\hat{x}\) of \(x\).
The super sensor output is therefore equal to:
\begin{equation}
@ -214,15 +179,15 @@ The super sensor output is therefore equal to:
\begin{figure}[htbp]
\centering
\includegraphics[scale=1,scale=1]{figs/fusion_super_sensor.pdf}
\caption{\label{fig:fusion_super_sensor}Sensor fusion architecture}
\includegraphics[scale=1]{figs/fusion_super_sensor.pdf}
\caption{\label{fig:fusion_super_sensor}Sensor fusion architecture with two normalized sensors}
\end{figure}
\subsection{Noise Sensor Filtering}
\label{sec:org99631b9}
\label{sec:org72d0e25}
\label{sec:noise_filtering}
In this section, it is supposed that all the sensors are perfectly calibrated, such that:
In this section, it is supposed that all the sensors are perfectly normalized, such that:
\begin{equation}
\label{eq:perfect_dynamics}
\frac{\hat{x}_i}{x} = \hat{G}_i(s) G_i(s) = 1
@ -230,7 +195,7 @@ In this section, it is supposed that all the sensors are perfectly calibrated, s
The effect of a non-perfect normalization will be discussed in the next section.
The super sensor output \(\hat{x}\) is then:
Provided \eqref{eq:perfect_dynamics} is verified, the super sensor output \(\hat{x}\) is then:
\begin{equation}
\label{eq:estimate_perfect_dyn}
\hat{x} = x + H_1(s) n_1 + H_2(s) n_2
@ -238,6 +203,7 @@ The super sensor output \(\hat{x}\) is then:
From \eqref{eq:estimate_perfect_dyn}, the complementary filters \(H_1(s)\) and \(H_2(s)\) are shown to only operate on the sensor's noises.
Thus, this sensor fusion architecture permits to filter the noise of both sensors without introducing any distortion in the physical quantity to be measured.
This is why the two filters must be complementary.
The estimation error \(\delta x\), defined as the difference between the sensor output \(\hat{x}\) and the measured quantity \(x\), is computed for the super sensor \eqref{eq:estimate_error}.
\begin{equation}
@ -251,24 +217,25 @@ As shown in \eqref{eq:noise_filtering_psd}, the Power Spectral Density (PSD) of
\Phi_{\delta x}(\omega) = \left|H_1(j\omega)\right|^2 \Phi_{n_1}(\omega) + \left|H_2(j\omega)\right|^2 \Phi_{n_2}(\omega)
\end{equation}
If the two sensors have identical noise characteristics (\(\Phi_{n_1}(\omega) = \Phi_{n_2}(\omega)\)), a simple averaging (\(H_1(s) = H_2(s) = 0.5\)) is what would minimize the super sensor noise.
If the two sensors have identical noise characteristics, \(\Phi_{n_1}(\omega) = \Phi_{n_2}(\omega)\), a simple averaging (\(H_1(s) = H_2(s) = 0.5\)) is what would minimize the super sensor noise.
This is the simplest form of sensor fusion with complementary filters.
However, the two sensors have usually high noise levels over distinct frequency regions.
In such case, to lower the noise of the super sensor, the value of the norm \(|H_1|\) has to be lowered when \(\Phi_{n_1}\) is larger than \(\Phi_{n_2}\) and that of \(|H_2|\) lowered when \(\Phi_{n_2}\) is larger than \(\Phi_{n_1}\).
In such case, to lower the noise of the super sensor, the norm \(|H_1(j\omega)|\) has to be small when \(\Phi_{n_1}(\omega)\) is larger than \(\Phi_{n_2}(\omega)\) and the norm \(|H_2(j\omega)|\) has to be small when \(\Phi_{n_2}(\omega)\) is larger than \(\Phi_{n_1}(\omega)\).
Therefore, by properly shaping the norm of the complementary filters, it is possible to minimize the noise of the super sensor noise.
\subsection{Sensor Fusion Robustness}
\label{sec:org3f9e403}
\label{sec:org6b3c1f3}
\label{sec:fusion_robustness}
In practical systems the sensor normalization is not perfect and condition \eqref{eq:perfect_dynamics} is not verified.
In order to study such imperfection, a multiplicative input uncertainty is added to the sensor dynamics (Figure \ref{fig:sensor_model_uncertainty}), where the nominal model is taken as the estimated model for the normalization \(\hat{G}_i(s)\), \(\Delta_i\) is any stable transfer function satisfying \(|\Delta_i(j\omega)| \le 1,\ \forall\omega\), and \(w_i(s)\) is a weight representing the magnitude of the uncertainty.
In order to study such imperfection, a multiplicative input uncertainty is added to the sensor dynamics (Fig.~\ref{fig:sensor_model_uncertainty}).
The nominal model is taken as the estimated model for the normalization \(\hat{G}_i(s)\), \(\Delta_i\) is any stable transfer function satisfying \(|\Delta_i(j\omega)| \le 1,\ \forall\omega\), and \(w_i(s)\) is a weighting transfer function representing the magnitude of the uncertainty.
The weight \(w_i(s)\) is chosen such that the real sensor dynamics is always contained in the uncertain region represented by a circle centered on \(1\) and with a radius equal to \(|w_i(j\omega)|\).
The weight \(w_i(s)\) is chosen such that the real sensor dynamics \(G(j\omega)\) is contained in the uncertain region represented by a circle in the complex plane, centered on \(1\) and with a radius equal to \(|w_i(j\omega)|\).
As the nominal sensor dynamics is taken as the normalized filter, the normalized sensor can be further simplified as shown in Figure \ref{fig:sensor_model_uncertainty_simplified}.
As the nominal sensor dynamics is taken as the normalized filter, the normalized sensor can be further simplified as shown in Fig.~\ref{fig:sensor_model_uncertainty_simplified}.
\begin{figure}[htbp]
\begin{subfigure}[b]{0.59\linewidth}
@ -287,7 +254,7 @@ As the nominal sensor dynamics is taken as the normalized filter, the normalized
\centering
\end{figure}
A sensor fusion architecture with two sensors with dynamical uncertainty is shown in Figure \ref{fig:sensor_fusion_dynamic_uncertainty}.
The sensor fusion architecture with two sensor models including dynamical uncertainty is shown in Fig.~\ref{fig:sensor_fusion_dynamic_uncertainty}.
\begin{figure}[htbp]
\centering
@ -295,44 +262,46 @@ A sensor fusion architecture with two sensors with dynamical uncertainty is show
\caption{\label{fig:sensor_fusion_dynamic_uncertainty}Sensor fusion architecture with sensor dynamics uncertainty}
\end{figure}
The super sensor dynamics \eqref{eq:super_sensor_dyn_uncertainty} is no longer equal to \(1\) and now depends on the sensor dynamics uncertainty weights \(w_i(s)\) as well as on the complementary filters \(H_i(s)\).
The super sensor dynamics \eqref{eq:super_sensor_dyn_uncertainty} is no longer equal to \(1\) and now depends on the sensor dynamical uncertainty weights \(w_i(s)\) as well as on the complementary filters \(H_i(s)\).
\begin{equation}
\label{eq:super_sensor_dyn_uncertainty}
\frac{\hat{x}}{x} = 1 + w_1(s) H_1(s) \Delta_1(s) + w_2(s) H_2(s) \Delta_2(s)
\end{equation}
The dynamical uncertainty of the super sensor can be graphically represented in the complex plane by a circle centered on \(1\) with a radius equal to \(|w_1(j\omega) H_1(j\omega)| + |w_2(j\omega) H_2(j\omega)|\) as shown in Figure \ref{fig:uncertainty_set_super_sensor}.
The dynamical uncertainty of the super sensor can be graphically represented in the complex plane by a circle centered on \(1\) with a radius equal to \(|w_1(j\omega) H_1(j\omega)| + |w_2(j\omega) H_2(j\omega)|\) as shown in Fig.~\ref{fig:uncertainty_set_super_sensor}.
\begin{figure}[htbp]
\centering
\includegraphics[scale=1]{figs/uncertainty_set_super_sensor.pdf}
\caption{\label{fig:uncertainty_set_super_sensor}Uncertainty region of the super sensor dynamics in the complex plane (solid circle). The contribution of both sensors 1 and 2 to the uncertainty are represented respectively by a blue circle and a red circle. The frequency dependency \(\omega\) is here omitted.}
\caption{\label{fig:uncertainty_set_super_sensor}Uncertainty region of the super sensor dynamics in the complex plane (grey circle). The contribution of both sensors 1 and 2 to the total uncertainty are represented respectively by a blue circle and a red circle. The frequency dependency \(\omega\) is here omitted.}
\end{figure}
The super sensor dynamical uncertainty (i.e. the robustness of the fusion) clearly depends on the complementary filters norms.
For instance, the phase uncertainty \(\Delta\phi(\omega)\) added by the super sensor dynamics at frequency \(\omega\) can be found by drawing a tangent from the origin to the uncertainty circle of super sensor (Figure \ref{fig:uncertainty_set_super_sensor}) and is bounded by \eqref{eq:max_phase_uncertainty}.
For instance, the phase \(\Delta\phi(\omega)\) added by the super sensor dynamics at frequency \(\omega\) is bounded by \(\Delta\phi_{\text{max}}(\omega)\) which can be found by drawing a tangent from the origin to the uncertainty circle of the super sensor (Fig.~\ref{fig:uncertainty_set_super_sensor}).
Therefore, the phase uncertainty of the super sensor dynamics depends on the Complementary filters norms \eqref{eq:max_phase_uncertainty}.
\begin{equation}
\label{eq:max_phase_uncertainty}
\Delta\phi(\omega) < \arcsin\big( |w_1(j\omega) H_1(j\omega)| + |w_2(j\omega) H_2(j\omega)| \big)
\Delta\phi_\text{max}(\omega) = \arcsin\big( |w_1(j\omega) H_1(j\omega)| + |w_2(j\omega) H_2(j\omega)| \big)
\end{equation}
As it is generally desired to limit the maximum phase added by the super sensor, \(H_1(s)\) and \(H_2(s)\) should be designed such that \(\Delta \phi\) is bounded to acceptable values.
Typically, the norm of the complementary filter \(|H_i(j\omega)|\) should be made small when \(|w_i(j\omega)|\) is large, i.e., at frequencies where the sensor dynamics is uncertain.
\section{Complementary Filters Shaping}
\label{sec:org82bc276}
\label{sec:org25756ab}
\label{sec:hinf_method}
As shown in Section \ref{sec:requirements}, the noise and robustness of the ``super sensor'' are determined by the complementary filters norms.
As shown in Section~\ref{sec:requirements}, the noise and robustness of the super sensor are a function of the complementary filters norms.
Therefore, a complementary filters synthesis method that allows to shape their norms would be of great use.
In this section, such synthesis is proposed by expressing this problem as a \(\mathcal{H}_\infty\) norm optimization.
In this section, such synthesis is proposed by writing the synthesis objective as a standard \(\mathcal{H}_\infty\) optimization problem.
As weighting functions are used to represent the wanted complementary filters shapes during the synthesis, the proper design of weighting functions is discussed.
Finally, the synthesis method is validated on an simple example.
\subsection{Synthesis Objective}
\label{sec:orgceb5825}
\label{sec:org6c44b50}
\label{sec:synthesis_objective}
The synthesis objective is to shape the norm of two filters \(H_1(s)\) and \(H_2(s)\) while ensuring their complementary property \eqref{eq:comp_filter}.
This is equivalent as to finding proper and stable transfer functions \(H_1(s)\) and \(H_2(s)\) such that conditions \eqref{eq:comp_filter_problem_form} are satisfied.
This is equivalent as to finding proper and stable transfer functions \(H_1(s)\) and \(H_2(s)\) such that conditions \eqref{eq:hinf_cond_complementarity}, \eqref{eq:hinf_cond_h1} and \eqref{eq:hinf_cond_h2} are satisfied.
\begin{subequations}
\label{eq:comp_filter_problem_form}
@ -342,15 +311,15 @@ This is equivalent as to finding proper and stable transfer functions \(H_1(s)\)
& |H_2(j\omega)| \le \frac{1}{|W_2(j\omega)|} \quad \forall\omega \label{eq:hinf_cond_h2}
\end{align}
\end{subequations}
where \(W_1(s)\) and \(W_2(s)\) are two weighting transfer functions that are chosen to specify the maximum wanted norms of the complementary filters during the synthesis.
\(W_1(s)\) and \(W_2(s)\) are two weighting transfer functions that are carefully chosen to specify the maximum wanted norms of the complementary filters during the synthesis.
\subsection{Shaping of Complementary Filters using \(\mathcal{H}_\infty\) synthesis}
\label{sec:org79feac5}
\label{sec:org1538346}
\label{sec:hinf_synthesis}
In this section, it is shown that the synthesis objective can be easily expressed as a standard \(\mathcal{H}_\infty\) optimal control problem and therefore solved using convenient tools readily available.
In this section, it is shown that the synthesis objective can be easily expressed as a standard \(\mathcal{H}_\infty\) optimization problem and therefore solved using convenient tools readily available.
Consider the generalized plant \(P(s)\) shown in Figure \ref{fig:h_infinity_robust_fusion} and mathematically described by \eqref{eq:generalized_plant}.
Consider the generalized plant \(P(s)\) shown in Fig.~\ref{fig:h_infinity_robust_fusion_plant} and mathematically described by \eqref{eq:generalized_plant}.
\begin{equation}
\label{eq:generalized_plant}
@ -358,12 +327,24 @@ Consider the generalized plant \(P(s)\) shown in Figure \ref{fig:h_infinity_robu
\end{equation}
\begin{figure}[htbp]
\begin{subfigure}[t]{0.5\linewidth}
\centering
\includegraphics[scale=1]{figs/h_infinity_robust_fusion_plant.pdf}
\caption{\label{fig:h_infinity_robust_fusion_plant} Generalized plant}
\vfill
\end{subfigure}
\hfill
\begin{subfigure}[t]{0.5\linewidth}
\centering
\includegraphics[scale=1]{figs/h_infinity_robust_fusion_fb.pdf}
\caption{\label{fig:h_infinity_robust_fusion_fb} Generalized plant with the synthesized filter}
\end{subfigure}
\hfill
\caption{\label{fig:h_infinity_robust_fusion} Architecture for the $\mathcal{H}_\infty$ synthesis of complementary filters}
\centering
\includegraphics[scale=1,scale=1]{figs/h_infinity_robust_fusion.pdf}
\caption{\label{fig:h_infinity_robust_fusion}Generalized plant used for \(\mathcal{H}_\infty\) synthesis of complementary filters}
\end{figure}
Applying the standard \(\mathcal{H}_\infty\) synthesis on the generalized plant \(P(s)\) is then equivalent as finding a stable filter \(H_2(s)\) which based on \(v\), generates a signal \(u\) such that the \(\mathcal{H}_\infty\) norm from \(w\) to \([z_1, \ z_2]\) is less than one \eqref{eq:hinf_syn_obj}.
Applying the standard \(\mathcal{H}_\infty\) synthesis on the generalized plant \(P(s)\) is then equivalent as finding a stable filter \(H_2(s)\) which based on \(v\), generates a signal \(u\) such that the \(\mathcal{H}_\infty\) norm of the system in Fig.~\ref{fig:h_infinity_robust_fusion_fb} from \(w\) to \([z_1, \ z_2]\) is less than one \eqref{eq:hinf_syn_obj}.
\begin{equation}
\label{eq:hinf_syn_obj}
\left\|\begin{matrix} \left(1 - H_2(s)\right) W_1(s) \\ H_2(s) W_2(s) \end{matrix}\right\|_\infty \le 1
@ -383,22 +364,24 @@ By then defining \(H_1(s)\) to be the complementary of \(H_2(s)\) \eqref{eq:defi
Therefore, applying the \(\mathcal{H}_\infty\) synthesis on the standard plant \(P(s)\) \eqref{eq:generalized_plant} will generate two filters \(H_2(s)\) and \(H_1(s) \triangleq 1 - H_2(s)\) that are complementary \eqref{eq:comp_filter_problem_form} and such that there norms are bellow specified bounds \eqref{eq:hinf_cond_h1},\eqref{eq:hinf_cond_h2}.
The above optimization problem can be efficiently solved in Matlab \cite{matlab20} using the Robust Control Toolbox.
Note that there is not an equivalence between the \(\mathcal{H}_\infty\) norm condition~\eqref{eq:hinf_problem} and the initial synthesis objectives~\eqref{eq:hinf_cond_h1} and~\eqref{eq:hinf_cond_h2}, but only an implication.
Hence, the optimization may be a little bit conservative with respect to the ``set'' of filters on which it is performed.
There might be solutions were the objectives~\eqref{eq:hinf_cond_h1} and~\eqref{eq:hinf_cond_h2} are valid but where the \(\mathcal{H}_\infty\) norm~\eqref{eq:hinf_problem} is larger than one.
In practice, this is however not an found to be an issue.
\subsection{Weighting Functions Design}
\label{sec:orgd27beed}
\label{sec:org11b4246}
\label{sec:hinf_weighting_func}
Weighting functions are used during the synthesis to specify what is the maximum allowed norms of the complementary filters.
The proper design of these weighting functions is of primary importance for the success of the presented complementary filters \(\mathcal{H}_\infty\) synthesis.
Weighting functions are used during the synthesis to specify the maximum allowed norms of the complementary filters.
The proper design of these weighting functions is of primary importance for the success of the presented \(\mathcal{H}_\infty\) synthesis of complementary filters.
First, only proper and stable transfer functions should be used.
Second, the order of the weighting functions should stay reasonably small in order to reduce the computational costs associated with the solving of the optimization problem and for the physical implementation of the filters (the order of the synthesized filters being equal to the sum of the weighting functions order).
Second, the order of the weighting functions should stay reasonably small in order to reduce the computational costs associated with the solving of the optimization problem and for the physical implementation of the filters (the order of the synthesized filters being equal to the sum of the weighting functions orders).
Third, one should not forget the fundamental limitations imposed by the complementary property \eqref{eq:comp_filter}.
This implies for instance that \(|H_1(j\omega)|\) and \(|H_2(j\omega)|\) cannot be made small at the same frequency.
When designing complementary filters, it is usually desired to specify its slope, its crossover frequency and its maximum gain at low and high frequency.
When designing complementary filters, it is usually desired to specify their slopes, their ``blending'' frequency and their maximum gains at low and high frequency.
To help with the design of the weighting functions such that the above specification can be easily expressed, the formula \eqref{eq:weight_formula} is proposed.
\begin{equation}
@ -414,8 +397,8 @@ The parameters in formula \eqref{eq:weight_formula} are:
\begin{itemize}
\item \(G_0 = lim_{\omega \to 0} |W(j\omega)|\): the low frequency gain
\item \(G_\infty = lim_{\omega \to \infty} |W(j\omega)|\): the high frequency gain
\item \(G_c = |W(j\omega_c)|\): the gain at \(\omega_c\)
\item \(n\): the slope between high and low frequency. It is also the order of the weighting function.
\item \(G_c = |W(j\omega_c)|\): the gain at a specific frequency \(\omega_c\) in \(\si{rad/s}\).
\item \(n\): the slope between high and low frequency. It also corresponds to the order of the weighting function.
\end{itemize}
The parameters \(G_0\), \(G_c\) and \(G_\infty\) should either satisfy condition \eqref{eq:cond_formula_1} or \eqref{eq:cond_formula_2}.
@ -427,34 +410,34 @@ The parameters \(G_0\), \(G_c\) and \(G_\infty\) should either satisfy condition
\end{align}
\end{subequations}
The typical shape of a weighting function generated using \eqref{eq:weight_formula} is shown in Figure \ref{fig:weight_formula}.
An example of the obtained magnitude of a weighting function generated using \eqref{eq:weight_formula} is shown in Fig.~\ref{fig:weight_formula}.
\begin{figure}[htbp]
\centering
\includegraphics[scale=1,scale=1]{figs/weight_formula.pdf}
\includegraphics[scale=1]{figs/weight_formula.pdf}
\caption{\label{fig:weight_formula}Magnitude of a weighting function generated using the proposed formula \eqref{eq:weight_formula}, \(G_0 = 1e^{-3}\), \(G_\infty = 10\), \(\omega_c = \SI{10}{Hz}\), \(G_c = 2\), \(n = 3\)}
\end{figure}
\subsection{Validation of the proposed synthesis method}
\label{sec:orgc8f3eb3}
\label{sec:orgb9a4dc3}
\label{sec:hinf_example}
The proposed methodology for the design of complementary filters is now applied on a simple example where two complementary filters \(H_1(s)\) and \(H_2(s)\) have to be designed such that:
\begin{itemize}
\item the merging frequency is around \(\SI{10}{Hz}\)
\item the slope of \(|H_1(j\omega)|\) is \(-2\) above \(\SI{10}{Hz}\)
\item the slope of \(|H_2(j\omega)|\) is \(+3\) below \(\SI{10}{Hz}\)
\item the maximum gain of both filters is \(10^{-3}\) away from the merging frequency
\item the blending frequency is around \(\SI{10}{Hz}\)
\item the slope of \(|H_1(j\omega)|\) is \(+2\) below \(\SI{10}{Hz}\), its low frequency gain is \(10^{-3}\)
\item the slope of \(|H_2(j\omega)|\) is \(-3\) above \(\SI{10}{Hz}\), its high frequency gain is \(10^{-3}\)
\end{itemize}
The first step is to design weighting functions that translate the above requirements.
They are here designed using \eqref{eq:weight_formula} with parameters summarized in table \ref{tab:weights_params}.
The magnitudes of the weighting functions are shown by dashed lines in Figure \ref{fig:hinf_filters_results}.
The first step is to translate the above requirements into the design of the weighting functions.
The proposed formula \eqref{eq:weight_formula} is here used for such purpose.
Parameters used are summarized in Table~\ref{tab:weights_params}.
The inverse magnitudes of the designed weighting functions, which are representing the maximum allowed norms of the complementary filters, are shown by the dashed lines in Fig.~\ref{fig:weights_W1_W2}.
\begin{table}[htbp]
\caption{\label{tab:weights_params}Parameters used for weighting functions \(W_1(s)\) and \(W_2(s)\) using \eqref{eq:weight_formula}}
\begin{figure}
\begin{minipage}[b]{0.49\linewidth}
\centering
\begin{tabularx}{0.29\linewidth}{ccc}
\begin{tabularx}{0.65\linewidth}{ccc}
\toprule
Parameters & \(W_1(s)\) & \(W_2(s)\)\\
\midrule
@ -465,17 +448,28 @@ Parameters & \(W_1(s)\) & \(W_2(s)\)\\
\(n\) & \(2\) & \(3\)\\
\bottomrule
\end{tabularx}
\end{table}
\captionof{table}{\label{tab:weights_params}Parameters used for weighting functions \(W_1(s)\) and \(W_2(s)\) using \eqref{eq:weight_formula}}
\end{minipage}
\hfill
\begin{minipage}[b]{0.49\linewidth}
\centering
\includegraphics[scale=1]{figs/weights_W1_W2.pdf}
\caption{\label{fig:weights_W1_W2}Inverse magnitude of the weighting functions}
\end{minipage}
\end{figure}
The \(\mathcal{H}_\infty\) synthesis is applied on the generalized plant of Figure \ref{fig:h_infinity_robust_fusion} using the Matlab \texttt{hinfsyn} command.
The synthesized filter \(H_2(s)\) is such that \(\mathcal{H}_\infty\) norm between \(w\) and \([z_1,\ z_2]^T\) is minimized and here found close to one \eqref{eq:hinf_synthesis_result}.
The \(\mathcal{H}_\infty\) synthesis is then applied to the generalized plant of Fig.~\ref{fig:h_infinity_robust_fusion_plant} on efficiently solved in Matlab~\cite{matlab20} using the Robust Control Toolbox.
The filter \(H_2(s)\) that minimizes the \(\mathcal{H}_\infty\) norm between \(w\) and \([z_1,\ z_2]^T\) is obtained.
The \(\mathcal{H}_\infty\) norm is here found to be close to one \eqref{eq:hinf_synthesis_result} which indicates that the synthesis is successful: the complementary filters norms are below the maximum specified upper bounds.
This is confirmed by the bode plots of the obtained complementary filters in Fig.~\ref{fig:hinf_filters_results}.
\begin{equation}
\label{eq:hinf_synthesis_result}
\left\|\begin{matrix} \left(1 - H_2(s)\right) W_1(s) \\ H_2(s) W_2(s) \end{matrix}\right\|_\infty \approx 1
\end{equation}
The bode plots of the obtained complementary filters are shown by solid lines in Figure \ref{fig:hinf_filters_results} and their transfer functions in the Laplace domain are given in \eqref{eq:hinf_synthesis_result_tf}.
Their transfer functions in the Laplace domain are given in \eqref{eq:hinf_synthesis_result_tf}.
As expected, the obtained filters are of order \(5\), that is the sum of the weighting functions orders.
\begin{subequations}
\label{eq:hinf_synthesis_result_tf}
@ -485,44 +479,44 @@ The bode plots of the obtained complementary filters are shown by solid lines in
\end{align}
\end{subequations}
The obtained transfer functions are of order \(5\) as expected (sum of the weighting functions orders), and their magnitudes are bellow the maximum specified ones as ensured by \eqref{eq:hinf_synthesis_result}.
\begin{figure}[htbp]
\centering
\includegraphics[scale=1]{figs/hinf_filters_results.pdf}
\caption{\label{fig:hinf_filters_results}Frequency response of the weighting functions and complementary filters obtained using \(\mathcal{H}_\infty\) synthesis}
\caption{\label{fig:hinf_filters_results}Bode plot of the obtained complementary filters}
\end{figure}
This simple example illustrates the fact that the proposed methodology for complementary filters shaping is quite easy to use and effective.
This simple example illustrates the fact that the proposed methodology for complementary filters shaping is easy to use and effective.
A more complex real life example is taken up in the next section.
\section{Application: Design of Complementary Filters used in the Active Vibration Isolation System at the LIGO}
\label{sec:org8cb3b2e}
\label{sec:org157e8c9}
\label{sec:application_ligo}
Sensor fusion using complementary filters are widely used in active vibration isolation systems in gravitational wave detectors such at the LIGO \cite{matichard15_seism_isolat_advan_ligo,hua05_low_ligo}, the VIRGO \cite{lucia18_low_frequen_optim_perfor_advan,heijningen18_low} and the KAGRA \cite[Chap. 5]{sekiguchi16_study_low_frequen_vibrat_isolat_system}.
Sensor fusion using complementary filters are widely used in active vibration isolation systems in gravitational wave detectors such at the LIGO~\cite{matichard15_seism_isolat_advan_ligo,hua05_low_ligo}, the VIRGO~\cite{lucia18_low_frequen_optim_perfor_advan,heijningen18_low} and the KAGRA \cite[Chap. 5]{sekiguchi16_study_low_frequen_vibrat_isolat_system}.
In the first isolation stage at the LIGO, two sets of complementary filters are used and included in a feedback loop \cite{hua04_low_ligo}.
In the first isolation stage at the LIGO, two sets of complementary filters are used and included in a feedback loop~\cite{hua04_low_ligo}.
A set of complementary filters (\(L_2,H_2\)) is first used to fuse a seismometer and a geophone.
Then, another set of complementary filters (\(L_1,H_1\)) is used to merge the output of the first ``inertial super sensor'' with a position sensor.
A simplified block diagram of the sensor fusion architecture is shown in Figure \ref{fig:ligo_super_sensor_architecture}.
A simplified block diagram of the sensor fusion architecture is shown in Fig.~\ref{fig:ligo_super_sensor_architecture}.
\begin{figure}[htbp]
\centering
\includegraphics[scale=1,scale=1]{figs/ligo_super_sensor_architecture.pdf}
\caption{\label{fig:ligo_super_sensor_architecture}Simplified block diagram of the sensor blending strategy for the first stage at the LIGO \cite{hua04_low_ligo}}
\includegraphics[scale=1]{figs/ligo_super_sensor_architecture.pdf}
\caption{\label{fig:ligo_super_sensor_architecture}Simplified block diagram of the sensor blending strategy for the first stage at the LIGO~\cite{hua04_low_ligo}}
\end{figure}
The fusion of the position sensor at low frequency with the ``inertial super sensor'' at high frequency using the complementary filters (\(L_1,H_1\)) is done for several reasons, first of which is to give the super sensor a DC sensibility and therefore allow the feedback loop to have authority at zero frequency.
The requirements on those filters are very tight and thus their design is complex and should be expressed as an optimization problem.
The approach used in \cite{hua05_low_ligo} is to use FIR complementary filters and to write the synthesis as a convex optimization problem.
The approach used in~\cite{hua05_low_ligo} is to use FIR complementary filters and to write the synthesis as a convex optimization problem.
After synthesis, the obtained FIR filters were found to be compliant with the requirements.
However they are of very high order so their implementation is quite complex.
In this section, the effectiveness of the proposed complementary filter synthesis strategy is demonstrated on the same set of requirements.
\subsection{Complementary Filters Specifications}
\label{sec:orgb603be6}
\label{sec:org872bc34}
\label{sec:ligo_specifications}
The specifications for the set of complementary filters (\(L_1,H_1\)) used at the LIGO are summarized below (for further details, refer to \cite{hua04_polyp_fir_compl_filter_contr_system}):
The specifications for the set of complementary filters (\(L_1,H_1\)) used at the LIGO are summarized below (for further details, refer to~\cite{hua04_polyp_fir_compl_filter_contr_system}):
\begin{itemize}
\item From \(0\) to \(\SI{0.008}{Hz}\), the magnitude \(|L_1(j\omega)|\) should be less or equal to \(8 \times 10^{-4}\)
\item Between \(\SI{0.008}{Hz}\) to \(\SI{0.04}{Hz}\), the filter \(L_1(s)\) should attenuate the input signal proportional to frequency cubed
@ -531,92 +525,92 @@ The specifications for the set of complementary filters (\(L_1,H_1\)) used at th
\end{itemize}
These specifications are therefore upper bounds on the complementary filters' magnitudes.
They are physically represented in Figure \ref{fig:fir_filter_ligo} as well as the obtained magnitude of the FIR filters in \cite{hua05_low_ligo}.
They are physically represented in Fig.~\ref{fig:fir_filter_ligo} as well as the obtained magnitude of the FIR filters in~\cite{hua05_low_ligo}.
\begin{figure}[htbp]
\centering
\includegraphics[scale=1,scale=1]{figs/fir_filter_ligo.pdf}
\caption{\label{fig:fir_filter_ligo}Specifications and Bode plot of the obtained FIR filters in \cite{hua05_low_ligo}}
\includegraphics[scale=1]{figs/fir_filter_ligo.pdf}
\caption{\label{fig:fir_filter_ligo}Specifications and Bode plot of the obtained FIR complementary filters in~\cite{hua05_low_ligo}. The filters are here obtained using the SeDuMi Matlab toolbox~\cite{sturm99_using_sedum}}
\end{figure}
\subsection{Weighting Functions Design}
\label{sec:orgd94a6e5}
\label{sec:org97ac8c9}
\label{sec:ligo_weights}
The weighting functions should be designed such that their inverse magnitude is as close as possible to the specifications in order to not over-constrain the synthesis problem.
However, the order of each weight should stay reasonably small in order to reduce the computational costs of the optimization problem as well as for the physical implementation of the filters.
A Type I Chebyshev filter of order \(20\) is used as the weighting transfer function \(w_L(s)\) corresponding to the low pass filter.
For the one corresponding to the high pass filter \(w_H(s)\), a \(7^{\text{th}}\) order transfer function is designed.
The magnitudes of the weighting functions are shown in Fig. \ref{fig:ligo_weights}.
A Type I Chebyshev filter of order \(20\) is used for the weighting transfer function \(W_L(s)\) corresponding to the low pass filter.
For the one corresponding to the high pass filter \(W_H(s)\), a \(7^{\text{th}}\) order transfer function is designed.
The magnitudes of the weighting functions are shown in Fig.~\ref{fig:ligo_weights}.
\begin{figure}[htbp]
\centering
\includegraphics[scale=1,scale=1]{figs/ligo_weights.pdf}
\caption{\label{fig:ligo_weights}Specifications and weighting functions magnitudes}
\includegraphics[scale=1]{figs/ligo_weights.pdf}
\caption{\label{fig:ligo_weights}Specifications and weighting functions inverse magnitudes}
\end{figure}
\subsection{\(\mathcal{H}_\infty\) Synthesis}
\label{sec:org1f03af8}
\subsection{\(\mathcal{H}_\infty\) Synthesis of the complementary filters}
\label{sec:orgf148a21}
\label{sec:ligo_results}
\(\mathcal{H}_\infty\) synthesis is performed using the architecture shown in Fig. \ref{eq:generalized_plant}.
The complementary filters obtained are of order \(27\).
In Fig. \ref{fig:comp_fir_ligo_hinf}, their bode plot is compared with the FIR filters of order 512 obtained in \cite{hua05_low_ligo}.
They are found to be very close to each other and this shows the effectiveness of the proposed synthesis method.
The proposed \(\mathcal{H}_\infty\) synthesis is performed on the generalized plant shown in Fig.~\ref{fig:h_infinity_robust_fusion_plant}.
After optimization, the \(\mathcal{H}_\infty\) norm from \(w\) to \([z_1,\ z_2]^T\) is found close to one indication successful synthesis.
In Fig.~\ref{fig:comp_fir_ligo_hinf}, the bode plot of the obtained complementary filters are compared with the FIR filters of order 512 obtained in~\cite{hua05_low_ligo}.
Even though the complementary filters using the \(\mathcal{H}_\infty\) synthesis are of much lower order (order 27), they are found to be very close to the FIR filters.
This confirms the effectiveness of the proposed synthesis method even when the complementary filters are subject to complex requirements.
\begin{figure}[htbp]
\centering
\includegraphics[scale=1,scale=1]{figs/comp_fir_ligo_hinf.pdf}
\caption{\label{fig:comp_fir_ligo_hinf}Comparison of the FIR filters (solid) designed in \cite{hua05_low_ligo} with the filters obtained with \(\mathcal{H}_\infty\) synthesis (dashed)}
\includegraphics[scale=1]{figs/comp_fir_ligo_hinf.pdf}
\caption{\label{fig:comp_fir_ligo_hinf}Comparison of the FIR filters (dashed) designed in~\cite{hua05_low_ligo} with the filters obtained with \(\mathcal{H}_\infty\) synthesis (solid)}
\end{figure}
\section{Discussion}
\label{sec:org013b9e6}
\label{sec:orgbc3d67c}
\label{sec:discussion}
\subsection{``Closed-Loop'' complementary filters}
\label{sec:orga1ea439}
\label{sec:org84a8225}
\label{sec:closed_loop_complementary_filters}
It is possible to use the fundamental properties of a feedback architecture to generate complementary filters.
It has been proposed by:
\begin{itemize}
\item \cite{plummer06_optim_compl_filter_their_applic_motion_measur} use H-Infinity to optimize complementary filters (flatten the super sensor noise spectral density)
\item \cite{jensen13_basic_uas} design of complementary filters with classical control theory, PID
\item Maybe also cite \cite{mahony05_compl_filter_desig_special_orthog}
\end{itemize}
Consider the feedback architecture of Figure \ref{fig:feedback_sensor_fusion}, with two inputs \(\hat{x}_1\) and \(\hat{x}_2\), and one output \(\hat{x}\).
An alternative way to implement complementary filters is by using a fundamental property of the classical feedback architecture shown in Fig.~\ref{fig:feedback_sensor_fusion}.
This is for instance presented in \cite{mahony05_compl_filter_desig_special_orthog,plummer06_optim_compl_filter_their_applic_motion_measur,jensen13_basic_uas}.
\begin{figure}[htbp]
\centering
\includegraphics[scale=1,scale=1]{figs/feedback_sensor_fusion.pdf}
\includegraphics[scale=1]{figs/feedback_sensor_fusion.pdf}
\caption{\label{fig:feedback_sensor_fusion}``Closed-Loop'' complementary filters}
\end{figure}
The output \(\hat{x}\) is described by \eqref{eq:closed_loop_complementary_filters}.
Consider the feedback architecture of Fig.~\ref{fig:feedback_sensor_fusion}, with two inputs \(\hat{x}_1\) and \(\hat{x}_2\), and one output \(\hat{x}\).
The output \(\hat{x}\) is linked to the inputs by \eqref{eq:closed_loop_complementary_filters}.
\begin{equation}
\label{eq:closed_loop_complementary_filters}
\hat{x} = \underbrace{\frac{1}{1 + L(s)}}_{S(s)} \hat{x}_1 + \underbrace{\frac{L(s)}{1 + L(s)}}_{T(s)} \hat{x}_2
\end{equation}
with the famous relationship
As for any classical feedback architecture, we have that the sum of the sensitivity and complementary sensitivity transfer function is equal to one~\eqref{eq:sensitivity_sum}.
\begin{equation}
\label{eq:sensitivity_sum}
T(s) + S(s) = 1
\end{equation}
Provided that the closed-loop system is stable, this indeed forms two complementary filters.
Therefore, two filters can be merged as shown in Figure \ref{fig:feedback_sensor_fusion_arch}.
Therefore, provided that the closed-loop system is stable, the closed-loop system in Fig.~\ref{fig:feedback_sensor_fusion} is corresponding to two complementary filters.
Two sensors can then be merged as shown in Fig.~\ref{fig:feedback_sensor_fusion_arch}.
\begin{figure}[htbp]
\centering
\includegraphics[scale=1,scale=1]{figs/feedback_sensor_fusion_arch.pdf}
\caption{\label{fig:feedback_sensor_fusion_arch}Classical feedback architecture for sensor fusion}
\includegraphics[scale=1]{figs/feedback_sensor_fusion_arch.pdf}
\caption{\label{fig:feedback_sensor_fusion_arch}Classical feedback architecture used for sensor fusion}
\end{figure}
One of the main advantage of this configuration is that standard tools of the linear control theory can be applied.
If one want to shape both the transfer functions \(\frac{\hat{x}}{\hat{x}_1}(s) = S(s)\) and \(\frac{\hat{x}}{\hat{x}_2}(s) = T(s)\), this corresponds to the \(\mathcal{H}_\infty\) mixed-sensitivity synthesis.
One of the main advantage of implementing and designing complementary filters using the feedback architecture of Fig.~\ref{fig:feedback_sensor_fusion} is that all the tools of the linear control theory can be applied for the design of the filters.
If one want to shape both \(\frac{\hat{x}}{\hat{x}_1}(s) = S(s)\) and \(\frac{\hat{x}}{\hat{x}_2}(s) = T(s)\), the \(\mathcal{H}_\infty\) mixed-sensitivity synthesis can be easily applied.
The \(\mathcal{H}_\infty\) mixed-sensitivity synthesis can be perform by applying the \(\mathcal{H}_\infty\) synthesis to the generalized plant \(P_L(s)\) shown in Figure \ref{fig:feedback_synthesis_architecture_generalized_plant} and described by \eqref{eq:generalized_plant_mixed_sensitivity} where \(W_1(s)\) and \(W_2(s)\) are weighting functions used to respectively shape \(S(s)\) and \(T(s)\).
To do so, weighting functions \(W_1(s)\) and \(W_2(s)\) are added to respectively shape \(S(s)\) and \(T(s)\) (Fig.~\ref{fig:feedback_synthesis_architecture}).
Then the system is re-organized to form the generalized plant \(P_L(s)\) shown in Fig.~\ref{fig:feedback_synthesis_architecture_generalized_plant}.
The \(\mathcal{H}_\infty\) mixed-sensitivity synthesis can finally be performed by applying the \(\mathcal{H}_\infty\) synthesis to the generalized plant \(P_L(s)\) which is described by \eqref{eq:generalized_plant_mixed_sensitivity}.
\begin{equation}
\label{eq:generalized_plant_mixed_sensitivity}
@ -626,70 +620,65 @@ The \(\mathcal{H}_\infty\) mixed-sensitivity synthesis can be perform by applyin
\end{bmatrix}
\end{equation}
This is equivalent as to find a filter \(L(s)\) such that \eqref{eq:comp_filters_feedback_obj} is verified.
The output of the synthesis is a filter \(L(s)\) such that the ``closed-loop'' \(\mathcal{H}_\infty\) norm from \([w_1,\ w_2]\) to \(z\) is less than one \eqref{eq:comp_filters_feedback_obj}.
\begin{equation}
\label{eq:comp_filters_feedback_obj}
\left\|\begin{matrix} \frac{1}{1 + L(s)} W_1(s) \\ \frac{L(s)}{1 + L(s)} W_2(s) \end{matrix}\right\|_\infty \le 1
\left\| \begin{matrix} \frac{z}{w_1} \\ \frac{z}{w_2} \end{matrix} \right\|_\infty = \left\| \begin{matrix} \frac{1}{1 + L(s)} W_1(s) \\ \frac{L(s)}{1 + L(s)} W_2(s) \end{matrix} \right\|_\infty \le 1
\end{equation}
The sensor fusion can be implemented as shown in Figure \ref{fig:feedback_sensor_fusion_arch} using the feedback architecture or more classically as shown in Figure \ref{fig:sensor_fusion_overview} using \eqref{eq:comp_filters_feedback}.
If the synthesis is successful, two complementary filters are obtained with their magnitudes bounded by the inverse magnitudes of the weighting functions.
The sensor fusion can then be implemented as shown in Fig.~\ref{fig:feedback_sensor_fusion_arch} using the feedback architecture or more classically as shown in Fig.~\ref{fig:sensor_fusion_overview} by defining the two complementary filters as in~\eqref{eq:comp_filters_feedback}.
\begin{equation}
\label{eq:comp_filters_feedback}
H_1(s) = \frac{1}{1 + L(s)}; \quad H_2(s) = \frac{L(s)}{1 + L(s)}
\end{equation}
The two being equivalent considering only the inputs/outputs relationships.
The two architecture are equivalent regarding their inputs/outputs relationships.
\begin{figure}[htbp]
\begin{subfigure}[t]{0.6\linewidth}
\centering
\includegraphics[scale=1]{figs/feedback_synthesis_architecture.pdf}
\caption{\label{fig:feedback_synthesis_architecture} Feedback architecture with included weights}
\end{subfigure}
\hfill
\begin{subfigure}[t]{0.4\linewidth}
\centering
\includegraphics[scale=1]{figs/feedback_synthesis_architecture_generalized_plant.pdf}
\caption{\label{fig:feedback_synthesis_architecture_generalized_plant} Generalized plant}
\end{subfigure}
\hfill
\caption{\label{fig:h_inf_mixed_sensitivity_synthesis} $\mathcal{H}_\infty$ mixed-sensitivity synthesis}
\centering
\end{figure}
As an example, two ``closed-loop'' complementary filters are designed using the \(\mathcal{H}_\infty\) mixed-sensitivity synthesis.
The weighting functions are designed using formula~\eqref{eq:weight_formula} with parameters shown in Table~\ref{tab:weights_params}.
After synthesis, a filter \(L(s)\) is obtained, its magnitude is shown in Fig.~\ref{fig:hinf_filters_results_mixed_sensitivity} by the dashed line.
The ``closed-loop'' complementary filters are compared with the inverse magnitude of the weighting functions in Fig.~\ref{fig:hinf_filters_results_mixed_sensitivity} confirming that the synthesis is successful.
The obtained ``closed-loop'' complementary filters are indeed equal to the ones obtained in Section~\ref{sec:hinf_example}.
\begin{figure}[htbp]
\centering
\includegraphics[scale=1,scale=1]{figs/feedback_synthesis_architecture_generalized_plant.pdf}
\caption{\label{fig:feedback_synthesis_architecture_generalized_plant}Generalized plant for the \(\mathcal{H}_\infty\) mixed-sensitivity synthesis}
\includegraphics[scale=1]{figs/hinf_filters_results_mixed_sensitivity.pdf}
\caption{\label{fig:hinf_filters_results_mixed_sensitivity}Bode plot of the obtained complementary filters after \(\mathcal{H}_\infty\) mixed-sensitivity synthesis}
\end{figure}
Example: same weights as in \ref{tab:weights_params}.
Therefore, complementary filter design is very similar to mixed-sensitivity synthesis.
They are actually equivalent by taking
\begin{equation}
L = H_H^{-1} - 1
\end{equation}
(provided \(H_H\) is invertible, therefore bi-proper)
\subsection{Imposing zero at origin / roll-off}
\label{sec:org293cf77}
\label{sec:add_features_in_filters}
3 methods:
Link to literature about doing that with mixed sensitivity
\subsection{Synthesis of Three Complementary Filters}
\label{sec:orgd44eb72}
\subsection{Synthesis of more than two Complementary Filters}
\label{sec:org862b3a1}
\label{sec:hinf_three_comp_filters}
Some applications may require to merge more than two sensors.
For instance at the LIGO, three sensors (an LVDT, a seismometer and a geophone) are merged to form a super sensor (Figure \ref{fig:ligo_super_sensor_architecture}). \par
\begin{itemize}
\item[{$\square$}] \cite{becker15_compl_filter_desig_three_frequen_bands}
\end{itemize}
When merging \(n>2\) sensors using complementary filters, two architectures can be used as shown in Figure \ref{fig:sensor_fusion_three}.
Some applications may require to merge more than two sensors~\cite{stoten01_fusion_kinet_data_using_compos_filter,becker15_compl_filter_desig_three_frequen_bands}.
For instance at the LIGO~\cite{matichard15_seism_isolat_advan_ligo}, three sensors (an LVDT, a seismometer and a geophone) are merged to form a super sensor (Fig.~\ref{fig:ligo_super_sensor_architecture}).\par
The fusion can either be done in a ``sequential'' way where \(n-1\) sets of two complementary filters are used (Figure \ref{fig:sensor_fusion_three_sequential}), or in a ``parallel'' way where one set of \(n\) complementary filters is used (Figure \ref{fig:sensor_fusion_three_parallel}).
When merging \(n>2\) sensors using complementary filters, two architectures can be used as shown in Fig.~\ref{fig:sensor_fusion_three}.
The fusion can either be done in a ``sequential'' way where \(n-1\) sets of two complementary filters are used (Fig.~\ref{fig:sensor_fusion_three_sequential}), or in a ``parallel'' way where one set of \(n\) complementary filters is used (Fig.~\ref{fig:sensor_fusion_three_parallel}).
In the first case, typical sensor fusion synthesis techniques can be used.
However, when a parallel architecture is used, a new synthesis method for a set of more than two complementary filters is required.
Such synthesis method is presented in this section. \par
\begin{center}
\fbox{
\begin{minipage}[c]{.6\textwidth}
Say possible advantages of parallel architecture
\end{minipage}
}
\end{center}
However, when a parallel architecture is used, a new synthesis method for a set of more than two complementary filters is required as only simple analytical formulas have been proposed in the literature~\{\cite{stoten01_fusion_kinet_data_using_compos_filter,becker15_compl_filter_desig_three_frequen_bands}\}.
A generalization of the proposed synthesis method of complementary filters is presented in this section. \par
\begin{figure}[htbp]
\begin{subfigure}[b]{0.59\linewidth}
@ -703,10 +692,11 @@ Say possible advantages of parallel architecture
\includegraphics[scale=1]{figs/sensor_fusion_three_parallel.pdf}
\caption{\label{fig:sensor_fusion_three_parallel}Parallel fusion}
\end{subfigure}
\caption{\label{fig:sensor_fusion_three}Sensor fusion architecture with more than two sensors}
\caption{\label{fig:sensor_fusion_three}Possible sensor fusion architecture when more than two sensors are to be merged}
\centering
\end{figure}
The synthesis objective is to compute a set of \(n\) stable transfer functions \([H_1(s),\ H_2(s),\ \dots,\ H_n(s)]\) such that \eqref{eq:hinf_problem_gen} is satisfied.
The synthesis objective is to compute a set of \(n\) stable transfer functions \([H_1(s),\ H_2(s),\ \dots,\ H_n(s)]\) such that conditions~\eqref{eq:hinf_cond_compl_gen} and~\eqref{eq:hinf_cond_perf_gen} are satisfied.
\begin{subequations}
\label{eq:hinf_problem_gen}
\begin{align}
@ -716,8 +706,10 @@ The synthesis objective is to compute a set of \(n\) stable transfer functions \
\end{subequations}
where \([W_1(s),\ W_2(s),\ \dots,\ W_n(s)]\) are weighting transfer functions that are chosen to specify the maximum wanted norms of the complementary filters during the synthesis.
Such synthesis objective is very close to the one described in Section \ref{sec:synthesis_objective}, and indeed the proposed synthesis architecture is also very similar. \par
Consider the generalized plant \(P_3(s)\) shown in Figure \ref{fig:comp_filter_three_hinf} which is also described by \eqref{eq:generalized_plant_three_filters}.
Such synthesis objective is very close to the one described in Section~\ref{sec:synthesis_objective}, and indeed the proposed synthesis method is a generalization of the one presented in Section~\ref{sec:hinf_synthesis}. \par
Before presenting the generalized synthesis method, the case with three sensors is presented.
Consider the generalized plant \(P_3(s)\) shown in Fig.~\ref{fig:comp_filter_three_hinf_gen_plant} which is also described by~\eqref{eq:generalized_plant_three_filters}.
\begin{equation}
\label{eq:generalized_plant_three_filters}
@ -725,12 +717,24 @@ Consider the generalized plant \(P_3(s)\) shown in Figure \ref{fig:comp_filter_t
\end{equation}
\begin{figure}[htbp]
\begin{subfigure}[t]{0.5\linewidth}
\centering
\includegraphics[scale=1]{figs/comp_filter_three_hinf_gen_plant.pdf}
\caption{\label{fig:comp_filter_three_hinf_gen_plant} Generalized plant}
\vfill
\end{subfigure}
\hfill
\begin{subfigure}[t]{0.5\linewidth}
\centering
\includegraphics[scale=1]{figs/comp_filter_three_hinf_fb.pdf}
\caption{\label{fig:comp_filter_three_hinf_fb} Generalized plant with the synthesized filter}
\end{subfigure}
\hfill
\caption{\label{fig:comp_filter_three_hinf} Architecture for the $\mathcal{H}_\infty$ synthesis of three complementary filters}
\centering
\includegraphics[scale=1,scale=1]{figs/comp_filter_three_hinf.pdf}
\caption{\label{fig:comp_filter_three_hinf}Architecture for \(\mathcal{H}_\infty\) synthesis of three complementary filters}
\end{figure}
Applying the \(\mathcal{H}_\infty\) synthesis on the generalized plant \(P_3(s)\) is equivalent as to find two stable filters \([H_2(s),\ H_3(s)]\) (shown in Figure \ref{fig:comp_filter_three_hinf}) such that the \(\mathcal{H}_\infty\) norm of the transfer function from \(w\) to \([z_1,\ z_2, \ z_3]\) is less than one \eqref{eq:hinf_syn_obj_three}.
Applying the \(\mathcal{H}_\infty\) synthesis on the generalized plant \(P_3(s)\) is equivalent as to find two stable filters \([H_2(s),\ H_3(s)]\) (shown in Fig.~\ref{fig:comp_filter_three_hinf_fb}) such that the \(\mathcal{H}_\infty\) norm of the transfer function from \(w\) to \([z_1,\ z_2, \ z_3]\) is less than one~\eqref{eq:hinf_syn_obj_three}.
\begin{equation}
\label{eq:hinf_syn_obj_three}
@ -738,16 +742,20 @@ Applying the \(\mathcal{H}_\infty\) synthesis on the generalized plant \(P_3(s)\
\end{equation}
By defining \(H_1(s) \triangleq 1 - H_2(s) - H_3(s)\), the proposed \(\mathcal{H}_\infty\) synthesis solves the design problem \eqref{eq:hinf_problem_gen} with \(n=3\). \par
An example is given to validate the method where three sensors are used in different frequency bands (up to \(\SI{1}{Hz}\), from \(1\) to \(\SI{10}{Hz}\) and above \(\SI{10}{Hz}\) respectively).
Three weighting functions are designed using \eqref{eq:weight_formula} and shown by dashed curves in Fig. \ref{fig:three_complementary_filters_results}.
The bode plots of the obtained complementary filters are shown in Fig. \ref{fig:three_complementary_filters_results}. \par
An example is given to validate the method where three sensors are used in different frequency bands.
For instance a displacement sensor from DC up to \(\SI{1}{Hz}\), a geophone from \(1\) to \(\SI{10}{Hz}\) and an accelerometer above \(\SI{10}{Hz}\).
Three weighting functions are designed using formula~\eqref{eq:weight_formula} and their inverse magnitudes are shown in Fig.~\ref{fig:three_complementary_filters_results} (dashed curves).
The \(\mathcal{H}_\infty\) synthesis is performed on the generalized plant \(P_3(s)\) and the bode plot of the obtained complementary filters are shown in Fig.~\ref{fig:three_complementary_filters_results} (solid lines). \par
\begin{figure}[htbp]
\centering
\includegraphics[scale=1,scale=1]{figs/three_complementary_filters_results.pdf}
\caption{\label{fig:three_complementary_filters_results}Frequency response of the weighting functions and three complementary filters obtained using \(\mathcal{H}_\infty\) synthesis}
\includegraphics[scale=1]{figs/three_complementary_filters_results.pdf}
\caption{\label{fig:three_complementary_filters_results}Bode plot of the inverse weighting functions and of the three complementary filters obtained using the \(\mathcal{H}_\infty\) synthesis}
\end{figure}
Such synthesis method can be generalized to a set of \(n\) complementary filters, even though there might not be any practical application for \(n>3\).
Even though there might not be any practical application for a set of more than 3 complementary filters, it can still be designed using the same procedure.
A set of \(n\) complementary filters can be shaped using the generalized plant \(P_n(s)\) described by~\eqref{eq:generalized_plant_n_filters}.
\begin{equation}
\label{eq:generalized_plant_n_filters}
@ -763,16 +771,28 @@ Such synthesis method can be generalized to a set of \(n\) complementary filters
\end{equation}
\section{Conclusion}
\label{sec:orgc6071ad}
\label{sec:org35fb45f}
\label{sec:conclusion}
This paper has shown how complementary filters can be used to combine multiple sensors in order to obtain a super sensor.
Typical specification on the super sensor noise and on the robustness of the sensor fusion has been shown to be linked to the norm of the complementary filters.
Therefore, a synthesis method that permits the shaping of the complementary filters norms has been proposed and has been successfully applied for the design of complex filters.
Future work will aim at further developing this synthesis method for the robust and optimal synthesis of complementary filters used in sensor fusion.
Sensors measuring a physical quantities are always subject to limitations both in terms of bandwidth or accuracy.
Complementary filters can be used to fuse multiple sensors with different characteristics in order to combine their benefits and yield a better estimate of the measured physical quantity.
The sensor fusion robustness and the obtained super sensor noise can be linked to the magnitude of the complementary filters.
In this paper, a synthesis method that enables the shaping of the complementary filters norms has been proposed.
Several example were used to emphasize the simplicity and the effectiveness of the proposed method.
Links with ``closed-loop'' complementary filters where highlighted, and the proposed method was generalized for the design of a set of more than two complementary filters.
Future work will aim at developing a complementary filter synthesis method that minimizes the super sensor noise while ensuring the robustness of the fusion.
\section*{Acknowledgment}
\label{sec:org4efce57}
\label{sec:org7a5f4b3}
This research benefited from a FRIA grant from the French Community of Belgium.
This paper is based on a paper previously presented at the ICCMA conference~\cite{dehaeze19_compl_filter_shapin_using_synth}.
\section*{Data Availability}
\label{sec:org767c106}
Matlab \cite{matlab20} was used for this study.
The source code is available under a MIT License and archived in Zenodo~\cite{dehaeze21_new_method_desig_compl_filter_code}.
\bibliographystyle{elsarticle-num}
\bibliography{ref}

BIN
journal/figs/comp_filter_three_hinf_fb.pdf Normal file
View File

Binary file not shown.

BIN
journal/figs/comp_filter_three_hinf_fb.png Normal file
View File

Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 19 KiB

287
journal/figs/comp_filter_three_hinf_fb.svg Normal file
View File

@ -0,0 +1,287 @@
<?xml version="1.0" encoding="UTF-8"?>
<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="184.799pt" height="186.709pt" viewBox="0 0 184.799 186.709" version="1.2">
<defs>
<g>
<symbol overflow="visible" id="glyph0-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph0-1">
<path style="stroke:none;" d="M 7.5 -5.296875 C 7.5 -6.0625 6.84375 -6.796875 5.546875 -6.796875 L 2.3125 -6.796875 C 2.125 -6.796875 2.015625 -6.796875 2.015625 -6.609375 C 2.015625 -6.484375 2.109375 -6.484375 2.3125 -6.484375 C 2.4375 -6.484375 2.625 -6.484375 2.734375 -6.46875 C 2.890625 -6.453125 2.953125 -6.421875 2.953125 -6.3125 C 2.953125 -6.265625 2.953125 -6.234375 2.921875 -6.125 L 1.578125 -0.78125 C 1.484375 -0.390625 1.46875 -0.3125 0.671875 -0.3125 C 0.5 -0.3125 0.390625 -0.3125 0.390625 -0.125 C 0.390625 0 0.515625 0 0.546875 0 L 1.8125 -0.03125 L 2.453125 -0.015625 C 2.671875 -0.015625 2.890625 0 3.09375 0 C 3.171875 0 3.296875 0 3.296875 -0.203125 C 3.296875 -0.3125 3.203125 -0.3125 3.015625 -0.3125 C 2.640625 -0.3125 2.375 -0.3125 2.375 -0.484375 C 2.375 -0.546875 2.390625 -0.59375 2.40625 -0.65625 L 3.015625 -3.140625 L 4.703125 -3.140625 C 6.125 -3.140625 7.5 -4.1875 7.5 -5.296875 Z M 6.578125 -5.53125 C 6.578125 -5.140625 6.375 -4.296875 5.984375 -3.9375 C 5.5 -3.484375 4.890625 -3.40625 4.453125 -3.40625 L 3.0625 -3.40625 L 3.734375 -6.109375 C 3.828125 -6.453125 3.84375 -6.484375 4.265625 -6.484375 L 5.21875 -6.484375 C 6.046875 -6.484375 6.578125 -6.21875 6.578125 -5.53125 Z M 6.578125 -5.53125 "/>
</symbol>
<symbol overflow="visible" id="glyph0-2">
<path style="stroke:none;" d="M 3.9375 -1.546875 C 3.9375 -1.90625 3.75 -2.15625 3.640625 -2.265625 C 3.359375 -2.546875 3.078125 -2.59375 2.640625 -2.6875 C 2.28125 -2.765625 1.875 -2.84375 1.875 -3.28125 C 1.875 -3.578125 2.125 -4.1875 3 -4.1875 C 3.25 -4.1875 3.75 -4.109375 3.890625 -3.71875 C 3.609375 -3.71875 3.40625 -3.5 3.40625 -3.28125 C 3.40625 -3.140625 3.5 -2.984375 3.71875 -2.984375 C 3.9375 -2.984375 4.1875 -3.15625 4.1875 -3.546875 C 4.1875 -3.984375 3.75 -4.40625 3 -4.40625 C 1.6875 -4.40625 1.3125 -3.390625 1.3125 -2.953125 C 1.3125 -2.171875 2.046875 -2.015625 2.34375 -1.953125 C 2.859375 -1.859375 3.375 -1.75 3.375 -1.203125 C 3.375 -0.953125 3.140625 -0.109375 1.953125 -0.109375 C 1.8125 -0.109375 1.046875 -0.109375 0.8125 -0.640625 C 1.1875 -0.59375 1.4375 -0.890625 1.4375 -1.171875 C 1.4375 -1.390625 1.28125 -1.515625 1.078125 -1.515625 C 0.8125 -1.515625 0.515625 -1.296875 0.515625 -0.859375 C 0.515625 -0.28125 1.078125 0.109375 1.9375 0.109375 C 3.546875 0.109375 3.9375 -1.09375 3.9375 -1.546875 Z M 3.9375 -1.546875 "/>
</symbol>
<symbol overflow="visible" id="glyph0-3">
<path style="stroke:none;" d="M 10.4375 -6.6875 C 10.4375 -6.734375 10.390625 -6.796875 10.3125 -6.796875 C 10.078125 -6.796875 9.796875 -6.765625 9.546875 -6.765625 C 9.203125 -6.765625 8.84375 -6.796875 8.515625 -6.796875 C 8.453125 -6.796875 8.328125 -6.796875 8.328125 -6.609375 C 8.328125 -6.5 8.40625 -6.484375 8.46875 -6.484375 C 8.703125 -6.484375 9.0625 -6.40625 9.0625 -6.09375 C 9.0625 -5.984375 9 -5.90625 8.921875 -5.765625 L 6.234375 -1.078125 L 5.875 -6.015625 C 5.859375 -6.21875 5.84375 -6.484375 6.5625 -6.484375 C 6.734375 -6.484375 6.84375 -6.484375 6.84375 -6.6875 C 6.84375 -6.78125 6.734375 -6.796875 6.6875 -6.796875 C 6.296875 -6.796875 5.875 -6.765625 5.46875 -6.765625 C 5.25 -6.765625 4.65625 -6.796875 4.421875 -6.796875 C 4.375 -6.796875 4.234375 -6.796875 4.234375 -6.59375 C 4.234375 -6.484375 4.34375 -6.484375 4.484375 -6.484375 C 4.921875 -6.484375 4.984375 -6.421875 5 -6.234375 L 5.0625 -5.46875 L 2.546875 -1.078125 L 2.171875 -6.125 C 2.171875 -6.25 2.171875 -6.484375 2.921875 -6.484375 C 3.03125 -6.484375 3.140625 -6.484375 3.140625 -6.6875 C 3.140625 -6.796875 3 -6.796875 3 -6.796875 C 2.59375 -6.796875 2.1875 -6.765625 1.765625 -6.765625 C 1.421875 -6.765625 1.0625 -6.796875 0.734375 -6.796875 C 0.671875 -6.796875 0.546875 -6.796875 0.546875 -6.609375 C 0.546875 -6.484375 0.640625 -6.484375 0.796875 -6.484375 C 1.296875 -6.484375 1.296875 -6.40625 1.328125 -6.125 L 1.765625 -0.046875 C 1.78125 0.140625 1.796875 0.21875 1.9375 0.21875 C 2.046875 0.21875 2.078125 0.15625 2.171875 0.015625 L 5.09375 -5.0625 L 5.46875 -0.046875 C 5.484375 0.171875 5.5 0.21875 5.625 0.21875 C 5.75 0.21875 5.8125 0.125 5.859375 0.03125 L 9.125 -5.640625 C 9.375 -6.078125 9.609375 -6.4375 10.234375 -6.484375 C 10.328125 -6.5 10.4375 -6.5 10.4375 -6.6875 Z M 10.4375 -6.6875 "/>
</symbol>
<symbol overflow="visible" id="glyph0-4">
<path style="stroke:none;" d="M 7.140625 -0.203125 C 7.140625 -0.3125 7.0625 -0.3125 6.875 -0.3125 C 6.5 -0.3125 6.21875 -0.3125 6.21875 -0.484375 C 6.21875 -0.546875 6.234375 -0.59375 6.25 -0.65625 L 7.59375 -6.03125 C 7.6875 -6.390625 7.703125 -6.484375 8.4375 -6.484375 C 8.6875 -6.484375 8.765625 -6.484375 8.765625 -6.6875 C 8.765625 -6.796875 8.65625 -6.796875 8.625 -6.796875 L 7.359375 -6.765625 L 6.078125 -6.796875 C 5.984375 -6.796875 5.875 -6.796875 5.875 -6.59375 C 5.875 -6.484375 5.96875 -6.484375 6.15625 -6.484375 C 6.15625 -6.484375 6.375 -6.484375 6.546875 -6.46875 C 6.71875 -6.453125 6.8125 -6.4375 6.8125 -6.3125 C 6.8125 -6.265625 6.796875 -6.25 6.765625 -6.125 L 6.171875 -3.6875 L 3.140625 -3.6875 L 3.71875 -6.03125 C 3.8125 -6.390625 3.84375 -6.484375 4.5625 -6.484375 C 4.8125 -6.484375 4.890625 -6.484375 4.890625 -6.6875 C 4.890625 -6.796875 4.78125 -6.796875 4.75 -6.796875 L 3.484375 -6.765625 L 2.203125 -6.796875 C 2.125 -6.796875 2.015625 -6.796875 2.015625 -6.59375 C 2.015625 -6.484375 2.09375 -6.484375 2.296875 -6.484375 C 2.296875 -6.484375 2.5 -6.484375 2.671875 -6.46875 C 2.84375 -6.453125 2.9375 -6.4375 2.9375 -6.3125 C 2.9375 -6.265625 2.921875 -6.234375 2.890625 -6.125 L 1.5625 -0.78125 C 1.46875 -0.390625 1.4375 -0.3125 0.65625 -0.3125 C 0.484375 -0.3125 0.390625 -0.3125 0.390625 -0.109375 C 0.390625 0 0.53125 0 0.53125 0 L 1.796875 -0.03125 L 2.421875 -0.015625 C 2.640625 -0.015625 2.859375 0 3.078125 0 C 3.15625 0 3.28125 0 3.28125 -0.203125 C 3.28125 -0.3125 3.1875 -0.3125 3 -0.3125 C 2.625 -0.3125 2.34375 -0.3125 2.34375 -0.484375 C 2.34375 -0.546875 2.375 -0.59375 2.375 -0.65625 L 3.0625 -3.390625 L 6.09375 -3.390625 L 5.40625 -0.640625 C 5.3125 -0.3125 5.109375 -0.3125 4.5 -0.3125 C 4.34375 -0.3125 4.265625 -0.3125 4.265625 -0.109375 C 4.265625 0 4.40625 0 4.40625 0 L 5.65625 -0.03125 L 6.296875 -0.015625 C 6.515625 -0.015625 6.734375 0 6.953125 0 C 7.03125 0 7.140625 0 7.140625 -0.203125 Z M 7.140625 -0.203125 "/>
</symbol>
<symbol overflow="visible" id="glyph0-5">
<path style="stroke:none;" d="M 6.875 -3.703125 C 6.875 -4.25 6.609375 -4.40625 6.4375 -4.40625 C 6.1875 -4.40625 5.9375 -4.140625 5.9375 -3.921875 C 5.9375 -3.796875 6 -3.734375 6.09375 -3.65625 C 6.203125 -3.546875 6.453125 -3.28125 6.453125 -2.8125 C 6.453125 -2.46875 6.15625 -1.5 5.90625 -0.984375 C 5.640625 -0.453125 5.296875 -0.109375 4.8125 -0.109375 C 4.34375 -0.109375 4.078125 -0.40625 4.078125 -0.96875 C 4.078125 -1.25 4.140625 -1.5625 4.1875 -1.703125 L 4.59375 -3.375 C 4.640625 -3.59375 4.75 -3.96875 4.75 -4.015625 C 4.75 -4.203125 4.609375 -4.296875 4.453125 -4.296875 C 4.34375 -4.296875 4.15625 -4.203125 4.09375 -4.015625 C 4.0625 -3.9375 3.59375 -2.046875 3.515625 -1.78125 C 3.453125 -1.484375 3.4375 -1.296875 3.4375 -1.125 C 3.4375 -1.015625 3.4375 -1 3.4375 -0.953125 C 3.21875 -0.421875 2.921875 -0.109375 2.53125 -0.109375 C 1.734375 -0.109375 1.734375 -0.84375 1.734375 -1.015625 C 1.734375 -1.328125 1.78125 -1.71875 2.25 -2.953125 C 2.359375 -3.25 2.421875 -3.390625 2.421875 -3.578125 C 2.421875 -4.03125 2.09375 -4.40625 1.609375 -4.40625 C 0.65625 -4.40625 0.28125 -2.953125 0.28125 -2.859375 C 0.28125 -2.765625 0.40625 -2.765625 0.40625 -2.765625 C 0.5 -2.765625 0.515625 -2.78125 0.5625 -2.953125 C 0.84375 -3.875 1.21875 -4.1875 1.578125 -4.1875 C 1.65625 -4.1875 1.828125 -4.171875 1.828125 -3.859375 C 1.828125 -3.609375 1.71875 -3.328125 1.640625 -3.15625 C 1.203125 -1.984375 1.078125 -1.515625 1.078125 -1.140625 C 1.078125 -0.234375 1.75 0.109375 2.5 0.109375 C 2.671875 0.109375 3.140625 0.109375 3.53125 -0.59375 C 3.796875 0.046875 4.484375 0.109375 4.78125 0.109375 C 5.53125 0.109375 5.96875 -0.515625 6.21875 -1.109375 C 6.5625 -1.890625 6.875 -3.21875 6.875 -3.703125 Z M 6.875 -3.703125 "/>
</symbol>
<symbol overflow="visible" id="glyph0-6">
<path style="stroke:none;" d="M 4.328125 -1.421875 C 4.328125 -1.46875 4.296875 -1.515625 4.203125 -1.515625 C 4.125 -1.515625 4.09375 -1.46875 4.078125 -1.390625 C 3.84375 -0.75 3.203125 -0.5625 2.875 -0.5625 C 2.671875 -0.5625 2.5 -0.609375 2.296875 -0.6875 C 1.953125 -0.8125 1.796875 -0.859375 1.59375 -0.859375 C 1.59375 -0.859375 1.40625 -0.859375 1.328125 -0.828125 C 1.859375 -1.40625 2.15625 -1.65625 2.515625 -1.953125 C 2.515625 -1.953125 3.125 -2.5 3.484375 -2.859375 C 4.421875 -3.78125 4.640625 -4.265625 4.640625 -4.296875 C 4.640625 -4.40625 4.53125 -4.40625 4.53125 -4.40625 C 4.46875 -4.40625 4.4375 -4.375 4.390625 -4.296875 C 4.09375 -3.8125 3.875 -3.65625 3.640625 -3.65625 C 3.40625 -3.65625 3.28125 -3.796875 3.140625 -3.96875 C 2.953125 -4.203125 2.78125 -4.40625 2.453125 -4.40625 C 1.703125 -4.40625 1.25 -3.46875 1.25 -3.265625 C 1.25 -3.21875 1.28125 -3.15625 1.359375 -3.15625 C 1.453125 -3.15625 1.46875 -3.203125 1.5 -3.265625 C 1.6875 -3.71875 2.265625 -3.734375 2.34375 -3.734375 C 2.546875 -3.734375 2.734375 -3.65625 2.96875 -3.578125 C 3.359375 -3.4375 3.46875 -3.4375 3.734375 -3.4375 C 3.375 -3 2.53125 -2.296875 2.34375 -2.125 L 1.453125 -1.296875 C 0.78125 -0.625 0.421875 -0.0625 0.421875 0.015625 C 0.421875 0.109375 0.546875 0.109375 0.546875 0.109375 C 0.625 0.109375 0.640625 0.09375 0.703125 -0.015625 C 0.9375 -0.375 1.234375 -0.640625 1.546875 -0.640625 C 1.78125 -0.640625 1.875 -0.546875 2.125 -0.265625 C 2.296875 -0.046875 2.484375 0.109375 2.765625 0.109375 C 3.75 0.109375 4.328125 -1.15625 4.328125 -1.421875 Z M 4.328125 -1.421875 "/>
</symbol>
<symbol overflow="visible" id="glyph0-7">
<path style="stroke:none;" d="M 5.40625 -1.421875 C 5.40625 -1.515625 5.3125 -1.515625 5.28125 -1.515625 C 5.1875 -1.515625 5.171875 -1.484375 5.140625 -1.34375 C 5 -0.78125 4.8125 -0.109375 4.40625 -0.109375 C 4.1875 -0.109375 4.09375 -0.234375 4.09375 -0.5625 C 4.09375 -0.78125 4.203125 -1.25 4.296875 -1.609375 L 4.5625 -2.671875 C 4.59375 -2.828125 4.703125 -3.203125 4.734375 -3.359375 C 4.78125 -3.578125 4.890625 -3.96875 4.890625 -4.015625 C 4.890625 -4.203125 4.75 -4.296875 4.59375 -4.296875 C 4.546875 -4.296875 4.296875 -4.28125 4.203125 -3.9375 L 3.46875 -0.953125 C 3.453125 -0.90625 3.0625 -0.109375 2.328125 -0.109375 C 1.8125 -0.109375 1.71875 -0.5625 1.71875 -0.921875 C 1.71875 -1.484375 1.984375 -2.265625 2.25 -2.953125 C 2.375 -3.25 2.421875 -3.390625 2.421875 -3.578125 C 2.421875 -4.03125 2.09375 -4.40625 1.609375 -4.40625 C 0.65625 -4.40625 0.28125 -2.953125 0.28125 -2.859375 C 0.28125 -2.765625 0.40625 -2.765625 0.40625 -2.765625 C 0.5 -2.765625 0.515625 -2.78125 0.5625 -2.953125 C 0.8125 -3.8125 1.1875 -4.1875 1.578125 -4.1875 C 1.65625 -4.1875 1.828125 -4.171875 1.828125 -3.859375 C 1.828125 -3.609375 1.71875 -3.328125 1.65625 -3.171875 C 1.28125 -2.1875 1.078125 -1.578125 1.078125 -1.078125 C 1.078125 -0.140625 1.765625 0.109375 2.296875 0.109375 C 2.953125 0.109375 3.3125 -0.34375 3.484375 -0.5625 C 3.59375 -0.15625 3.9375 0.109375 4.375 0.109375 C 4.71875 0.109375 4.953125 -0.125 5.109375 -0.4375 C 5.28125 -0.796875 5.40625 -1.421875 5.40625 -1.421875 Z M 5.40625 -1.421875 "/>
</symbol>
<symbol overflow="visible" id="glyph0-8">
<path style="stroke:none;" d="M 4.65625 -3.703125 C 4.65625 -4.234375 4.40625 -4.40625 4.21875 -4.40625 C 3.96875 -4.40625 3.734375 -4.140625 3.734375 -3.921875 C 3.734375 -3.796875 3.78125 -3.734375 3.890625 -3.625 C 4.09375 -3.421875 4.234375 -3.171875 4.234375 -2.8125 C 4.234375 -2.390625 3.625 -0.109375 2.453125 -0.109375 C 1.953125 -0.109375 1.71875 -0.453125 1.71875 -0.96875 C 1.71875 -1.53125 1.984375 -2.265625 2.296875 -3.078125 C 2.375 -3.25 2.421875 -3.390625 2.421875 -3.578125 C 2.421875 -4.03125 2.09375 -4.40625 1.609375 -4.40625 C 0.671875 -4.40625 0.28125 -2.953125 0.28125 -2.859375 C 0.28125 -2.765625 0.40625 -2.765625 0.40625 -2.765625 C 0.5 -2.765625 0.515625 -2.78125 0.5625 -2.953125 C 0.859375 -3.953125 1.28125 -4.1875 1.578125 -4.1875 C 1.65625 -4.1875 1.828125 -4.1875 1.828125 -3.859375 C 1.828125 -3.609375 1.71875 -3.34375 1.65625 -3.171875 C 1.21875 -2.015625 1.078125 -1.546875 1.078125 -1.125 C 1.078125 -0.046875 1.953125 0.109375 2.421875 0.109375 C 4.09375 0.109375 4.65625 -3.1875 4.65625 -3.703125 Z M 4.65625 -3.703125 "/>
</symbol>
<symbol overflow="visible" id="glyph1-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph1-1">
<path style="stroke:none;" d="M 3.578125 -1.203125 C 3.578125 -1.75 3.140625 -2.296875 2.375 -2.453125 C 3.09375 -2.71875 3.359375 -3.234375 3.359375 -3.671875 C 3.359375 -4.21875 2.734375 -4.625 1.953125 -4.625 C 1.1875 -4.625 0.59375 -4.25 0.59375 -3.6875 C 0.59375 -3.453125 0.75 -3.328125 0.953125 -3.328125 C 1.171875 -3.328125 1.3125 -3.484375 1.3125 -3.671875 C 1.3125 -3.875 1.171875 -4.03125 0.953125 -4.046875 C 1.203125 -4.34375 1.671875 -4.421875 1.9375 -4.421875 C 2.25 -4.421875 2.6875 -4.265625 2.6875 -3.671875 C 2.6875 -3.375 2.59375 -3.046875 2.40625 -2.84375 C 2.1875 -2.578125 1.984375 -2.5625 1.640625 -2.53125 C 1.46875 -2.515625 1.453125 -2.515625 1.421875 -2.515625 C 1.421875 -2.515625 1.34375 -2.5 1.34375 -2.421875 C 1.34375 -2.328125 1.40625 -2.328125 1.53125 -2.328125 L 1.90625 -2.328125 C 2.453125 -2.328125 2.84375 -1.953125 2.84375 -1.203125 C 2.84375 -0.34375 2.328125 -0.078125 1.9375 -0.078125 C 1.65625 -0.078125 1.03125 -0.15625 0.75 -0.578125 C 1.078125 -0.578125 1.15625 -0.8125 1.15625 -0.96875 C 1.15625 -1.1875 0.984375 -1.34375 0.765625 -1.34375 C 0.578125 -1.34375 0.375 -1.21875 0.375 -0.9375 C 0.375 -0.28125 1.09375 0.140625 1.9375 0.140625 C 2.90625 0.140625 3.578125 -0.515625 3.578125 -1.203125 Z M 3.578125 -1.203125 "/>
</symbol>
<symbol overflow="visible" id="glyph1-2">
<path style="stroke:none;" d="M 3.296875 0 L 3.296875 -0.25 L 3.03125 -0.25 C 2.328125 -0.25 2.328125 -0.34375 2.328125 -0.578125 L 2.328125 -4.4375 C 2.328125 -4.625 2.328125 -4.625 2.125 -4.625 C 1.671875 -4.1875 1.046875 -4.1875 0.765625 -4.1875 L 0.765625 -3.9375 C 0.921875 -3.9375 1.390625 -3.9375 1.765625 -4.125 L 1.765625 -0.578125 C 1.765625 -0.34375 1.765625 -0.25 1.078125 -0.25 L 0.8125 -0.25 L 0.8125 0 L 2.046875 -0.03125 Z M 3.296875 0 "/>
</symbol>
<symbol overflow="visible" id="glyph1-3">
<path style="stroke:none;" d="M 3.515625 -1.265625 L 3.28125 -1.265625 C 3.265625 -1.109375 3.1875 -0.703125 3.09375 -0.640625 C 3.046875 -0.59375 2.515625 -0.59375 2.40625 -0.59375 L 1.125 -0.59375 C 1.859375 -1.234375 2.109375 -1.4375 2.515625 -1.765625 C 3.03125 -2.171875 3.515625 -2.609375 3.515625 -3.265625 C 3.515625 -4.109375 2.78125 -4.625 1.890625 -4.625 C 1.03125 -4.625 0.4375 -4.015625 0.4375 -3.375 C 0.4375 -3.03125 0.734375 -2.984375 0.8125 -2.984375 C 0.96875 -2.984375 1.171875 -3.109375 1.171875 -3.359375 C 1.171875 -3.484375 1.125 -3.734375 0.765625 -3.734375 C 0.984375 -4.21875 1.453125 -4.375 1.78125 -4.375 C 2.484375 -4.375 2.84375 -3.828125 2.84375 -3.265625 C 2.84375 -2.65625 2.40625 -2.1875 2.1875 -1.9375 L 0.515625 -0.265625 C 0.4375 -0.203125 0.4375 -0.1875 0.4375 0 L 3.3125 0 Z M 3.515625 -1.265625 "/>
</symbol>
<symbol overflow="visible" id="glyph2-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph2-1">
<path style="stroke:none;" d="M 3.296875 2.390625 C 3.296875 2.359375 3.296875 2.34375 3.125 2.171875 C 1.875 0.921875 1.5625 -0.96875 1.5625 -2.484375 C 1.5625 -4.21875 1.9375 -5.953125 3.171875 -7.203125 C 3.296875 -7.3125 3.296875 -7.328125 3.296875 -7.359375 C 3.296875 -7.4375 3.25 -7.46875 3.1875 -7.46875 C 3.09375 -7.46875 2.203125 -6.78125 1.609375 -5.53125 C 1.109375 -4.421875 0.984375 -3.328125 0.984375 -2.484375 C 0.984375 -1.71875 1.09375 -0.5 1.640625 0.609375 C 2.234375 1.84375 3.09375 2.484375 3.1875 2.484375 C 3.25 2.484375 3.296875 2.453125 3.296875 2.390625 Z M 3.296875 2.390625 "/>
</symbol>
<symbol overflow="visible" id="glyph2-2">
<path style="stroke:none;" d="M 2.875 -2.484375 C 2.875 -3.265625 2.765625 -4.46875 2.21875 -5.59375 C 1.625 -6.8125 0.765625 -7.46875 0.671875 -7.46875 C 0.609375 -7.46875 0.5625 -7.421875 0.5625 -7.359375 C 0.5625 -7.328125 0.5625 -7.3125 0.75 -7.140625 C 1.734375 -6.15625 2.296875 -4.5625 2.296875 -2.484375 C 2.296875 -0.78125 1.9375 0.96875 0.703125 2.21875 C 0.5625 2.34375 0.5625 2.359375 0.5625 2.390625 C 0.5625 2.453125 0.609375 2.484375 0.671875 2.484375 C 0.765625 2.484375 1.65625 1.8125 2.25 0.546875 C 2.75 -0.546875 2.875 -1.65625 2.875 -2.484375 Z M 2.875 -2.484375 "/>
</symbol>
<symbol overflow="visible" id="glyph3-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph3-1">
<path style="stroke:none;" d="M 8.09375 -3 C 8.09375 -3.21875 7.90625 -3.21875 7.71875 -3.21875 L 4.53125 -3.21875 L 4.53125 -6.40625 C 4.53125 -6.5625 4.53125 -6.78125 4.328125 -6.78125 C 4.109375 -6.78125 4.109375 -6.578125 4.109375 -6.40625 L 4.109375 -3.21875 L 0.921875 -3.21875 C 0.765625 -3.21875 0.546875 -3.21875 0.546875 -3.015625 C 0.546875 -2.796875 0.75 -2.796875 0.921875 -2.796875 L 4.109375 -2.796875 L 4.109375 0.390625 C 4.109375 0.5625 4.109375 0.765625 4.3125 0.765625 C 4.53125 0.765625 4.53125 0.578125 4.53125 0.390625 L 4.53125 -2.796875 L 7.71875 -2.796875 C 7.890625 -2.796875 8.09375 -2.796875 8.09375 -3 Z M 8.09375 -3 "/>
</symbol>
<symbol overflow="visible" id="glyph4-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph4-1">
<path style="stroke:none;" d="M 6.90625 -2.484375 C 6.90625 -2.6875 6.71875 -2.6875 6.578125 -2.6875 L 1.15625 -2.6875 C 1.015625 -2.6875 0.828125 -2.6875 0.828125 -2.484375 C 0.828125 -2.296875 1.015625 -2.296875 1.15625 -2.296875 L 6.578125 -2.296875 C 6.71875 -2.296875 6.90625 -2.296875 6.90625 -2.484375 Z M 6.90625 -2.484375 "/>
</symbol>
<symbol overflow="visible" id="glyph5-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph5-1">
<path style="stroke:none;" d="M 5.109375 23.484375 L 5.109375 22.9375 L 3.03125 22.9375 L 3.03125 0.15625 L 5.109375 0.15625 L 5.109375 -0.390625 L 2.484375 -0.390625 L 2.484375 23.484375 Z M 5.109375 23.484375 "/>
</symbol>
<symbol overflow="visible" id="glyph5-2">
<path style="stroke:none;" d="M 2.75 23.484375 L 2.75 -0.390625 L 0.140625 -0.390625 L 0.140625 0.15625 L 2.203125 0.15625 L 2.203125 22.9375 L 0.140625 22.9375 L 0.140625 23.484375 Z M 2.75 23.484375 "/>
</symbol>
</g>
<clipPath id="clip1">
<path d="M 0.0429688 42 L 45 42 L 45 44 L 0.0429688 44 Z M 0.0429688 42 "/>
</clipPath>
<clipPath id="clip2">
<path d="M 33 42 L 184.554688 42 L 184.554688 171 L 33 171 Z M 33 42 "/>
</clipPath>
<clipPath id="clip3">
<path d="M 0.0429688 74 L 106 74 L 106 165 L 0.0429688 165 Z M 0.0429688 74 "/>
</clipPath>
<clipPath id="clip4">
<path d="M 163 27 L 184.554688 27 L 184.554688 59 L 163 59 Z M 163 27 "/>
</clipPath>
<clipPath id="clip5">
<path d="M 163 59 L 184.554688 59 L 184.554688 92 L 163 92 Z M 163 59 "/>
</clipPath>
<clipPath id="clip6">
<path d="M 163 91 L 184.554688 91 L 184.554688 124 L 163 124 Z M 163 91 "/>
</clipPath>
</defs>
<g id="surface1">
<path style="fill-rule:nonzero;fill:rgb(89.99939%,89.99939%,89.99939%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -70.86535 -63.781263 L 70.866926 -63.781263 L 70.866926 63.780524 L -70.86535 63.780524 Z M -70.86535 -63.781263 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-1" x="80.86947" y="11.275426"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-1" x="87.255517" y="12.767102"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-1" x="91.718561" y="11.275426"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-2" x="95.586533" y="11.275426"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-2" x="100.249266" y="11.275426"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(100%,100%,100%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 18.085571 27.227548 L 51.023547 27.227548 L 51.023547 49.907372 L 18.085571 49.907372 Z M 18.085571 27.227548 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-3" x="113.866043" y="45.731724"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-2" x="123.260401" y="47.224397"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-1" x="127.723445" y="45.731724"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-2" x="131.591416" y="45.731724"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-2" x="136.254149" y="45.731724"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(100%,100%,100%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 18.085571 -4.91231 L 51.023547 -4.91231 L 51.023547 17.767514 L 18.085571 17.767514 Z M 18.085571 -4.91231 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-3" x="113.866043" y="77.820711"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-3" x="123.260401" y="79.312386"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-1" x="127.723445" y="77.820711"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-2" x="131.591416" y="77.820711"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-2" x="136.254149" y="77.820711"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(100%,100%,100%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 18.085571 -37.048256 L 51.023547 -37.048256 L 51.023547 -14.372344 L 18.085571 -14.372344 Z M 18.085571 -37.048256 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-3" x="113.866043" y="109.909697"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-1" x="123.260401" y="111.401373"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-1" x="127.723445" y="109.909697"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-2" x="131.591416" y="109.909697"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-2" x="136.254149" y="109.909697"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(100%,100%,100%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 0.577858 38.569416 C 0.577858 42.419157 -2.544188 45.541203 -6.393929 45.541203 C -10.247582 45.541203 -13.369628 42.419157 -13.369628 38.569416 C -13.369628 34.715763 -10.247582 31.593717 -6.393929 31.593717 C -2.544188 31.593717 0.577858 34.715763 0.577858 38.569416 Z M 0.577858 38.569416 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph3-1" x="81.778054" y="46.157061"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph4-1" x="76.382464" y="57.999603"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(100%,100%,100%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -30.376562 38.569416 C -30.376562 42.419157 -33.498608 45.541203 -37.352261 45.541203 C -41.202001 45.541203 -44.324047 42.419157 -44.324047 38.569416 C -44.324047 34.715763 -41.202001 31.593717 -37.352261 31.593717 C -33.498608 31.593717 -30.376562 34.715763 -30.376562 38.569416 Z M -30.376562 38.569416 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph3-1" x="50.871225" y="46.157061"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph4-1" x="45.475634" y="57.999603"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(100%,100%,100%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -21.163005 -103.831869 L 21.164582 -103.831869 L 21.164582 -73.28042 L -21.163005 -73.28042 Z M -21.163005 -103.831869 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph5-1" x="74.679119" y="158.630762"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-4" x="79.928936" y="166.58836"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-3" x="88.19705" y="168.081033"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-1" x="92.659095" y="166.58836"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-2" x="96.528065" y="166.58836"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-2" x="101.190798" y="166.58836"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-4" x="79.928936" y="178.525754"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-1" x="88.19705" y="180.017429"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-1" x="92.659095" y="178.525754"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-2" x="96.528065" y="178.525754"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-2" x="101.190798" y="178.525754"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph5-2" x="105.05877" y="158.630762"/>
</g>
<g clip-path="url(#clip1)" clip-rule="nonzero">
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -91.209508 38.569416 L -48.960168 38.569416 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.052524 0.0019463 L 1.608108 1.684252 L 3.086972 0.0019463 L 1.608108 -1.684271 Z M 6.052524 0.0019463 " transform="matrix(0.998444,0,0,-0.998444,40.777207,43.244131)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-5" x="5.241343" y="39.432541"/>
</g>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -6.393929 -25.7103 L -6.393929 26.961509 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.052039 -0.000801368 L 1.607622 1.681504 L 3.086486 -0.000801368 L 1.607622 -1.683107 Z M 6.052039 -0.000801368 " transform="matrix(0,-0.998444,-0.998444,0,86.108575,57.667621)"/>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 88.097656 107.421875 C 88.097656 106.324219 87.207031 105.433594 86.109375 105.433594 C 85.011719 105.433594 84.117188 106.324219 84.117188 107.421875 C 84.117188 108.519531 85.011719 109.410156 86.109375 109.410156 C 87.207031 109.410156 88.097656 108.519531 88.097656 107.421875 Z M 88.097656 107.421875 "/>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -37.352261 6.429558 L -37.352261 26.961509 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.052039 0.00148068 L 1.607622 1.683786 L 3.086486 0.00148068 L 1.607622 -1.684737 Z M 6.052039 0.00148068 " transform="matrix(0,-0.998444,-0.998444,0,55.200697,57.667621)"/>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 57.191406 75.332031 C 57.191406 74.234375 56.300781 73.34375 55.199219 73.34375 C 54.101562 73.34375 53.210938 74.234375 53.210938 75.332031 C 53.210938 76.433594 54.101562 77.324219 55.199219 77.324219 C 56.300781 77.324219 57.191406 76.433594 57.191406 75.332031 Z M 57.191406 75.332031 "/>
<g clip-path="url(#clip2)" clip-rule="nonzero">
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -58.498449 38.569416 L -58.498449 -51.422187 L 91.207172 -51.422187 L 91.207172 -88.558101 L 26.293657 -88.558101 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.052273 0.00137065 L 1.607857 1.683676 L 3.08672 0.00137065 L 1.607857 -1.684847 Z M 6.052273 0.00137065 " transform="matrix(-0.998444,0,0,0.998444,121.581917,170.170506)"/>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 36.074219 43.242188 C 36.074219 42.144531 35.183594 41.253906 34.085938 41.253906 C 32.988281 41.253906 32.097656 42.144531 32.097656 43.242188 C 32.097656 44.34375 32.988281 45.234375 34.085938 45.234375 C 35.183594 45.234375 36.074219 44.34375 36.074219 43.242188 Z M 36.074219 43.242188 "/>
<g clip-path="url(#clip3)" clip-rule="nonzero">
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -21.659872 -82.247499 L -91.209508 -82.247499 L -91.209508 6.429558 L 12.952584 6.429558 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.053196 0.00131812 L 1.608779 1.683624 L 3.087643 0.00131812 L 1.608779 -1.6849 Z M 6.053196 0.00131812 " transform="matrix(0.998444,0,0,-0.998444,102.592943,75.333347)"/>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -21.659872 -94.86479 L -79.86764 -94.86479 L -79.86764 -25.7103 L 12.952584 -25.7103 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.053196 0.000679939 L 1.608779 1.682985 L 3.087643 0.000679939 L 1.608779 -1.681625 Z M 6.053196 0.000679939 " transform="matrix(0.998444,0,0,-0.998444,102.592943,107.422554)"/>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -30.376562 38.569416 L -18.001836 38.569416 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.054806 0.0019463 L 1.61039 1.684252 L 3.089254 0.0019463 L 1.61039 -1.684271 Z M 6.054806 0.0019463 " transform="matrix(0.998444,0,0,-0.998444,71.685085,43.244131)"/>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 0.577858 38.569416 L 12.952584 38.569416 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.053196 0.0019463 L 1.608779 1.684252 L 3.087643 0.0019463 L 1.608779 -1.684271 Z M 6.053196 0.0019463 " transform="matrix(0.998444,0,0,-0.998444,102.592943,43.244131)"/>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 51.524326 38.569416 L 86.574964 38.569416 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 182.144531 43.242188 L 177.703125 41.5625 L 179.179688 43.242188 L 177.703125 44.925781 Z M 182.144531 43.242188 "/>
<g clip-path="url(#clip4)" clip-rule="nonzero">
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.055146 0.0019463 L 1.606817 1.684252 L 3.085681 0.0019463 L 1.606817 -1.684271 Z M 6.055146 0.0019463 " transform="matrix(0.998444,0,0,-0.998444,176.098808,43.244131)"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-6" x="170.657529" y="37.939868"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-2" x="175.283319" y="39.432541"/>
</g>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 51.524326 6.429558 L 86.574964 6.429558 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 182.144531 75.332031 L 177.703125 73.652344 L 179.179688 75.332031 L 177.703125 77.015625 Z M 182.144531 75.332031 "/>
<g clip-path="url(#clip5)" clip-rule="nonzero">
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.055146 0.00131812 L 1.606817 1.683624 L 3.085681 0.00131812 L 1.606817 -1.6849 Z M 6.055146 0.00131812 " transform="matrix(0.998444,0,0,-0.998444,176.098808,75.333347)"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-6" x="170.657529" y="70.028855"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-3" x="175.283319" y="71.521528"/>
</g>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 51.524326 -25.7103 L 86.574964 -25.7103 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 182.144531 107.421875 L 177.703125 105.742188 L 179.179688 107.421875 L 177.703125 109.101562 Z M 182.144531 107.421875 "/>
<g clip-path="url(#clip6)" clip-rule="nonzero">
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.055146 0.000679939 L 1.606817 1.682985 L 3.085681 0.000679939 L 1.606817 -1.681625 Z M 6.055146 0.000679939 " transform="matrix(0.998444,0,0,-0.998444,176.098808,107.422554)"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-6" x="170.657529" y="102.117842"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-1" x="175.283319" y="103.609517"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-7" x="3.857499" y="70.028855"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-2" x="9.552623" y="71.521528"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-7" x="3.857499" y="102.117842"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-3" x="9.552623" y="103.609517"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-8" x="172.612482" y="129.280507"/>
</g>
</g>
</svg>
Side by Side

After

Width:  |  Height:  |  Size: 36 KiB

BIN
journal/figs/comp_filter_three_hinf_gen_plant.pdf Normal file
View File

Binary file not shown.

BIN
journal/figs/comp_filter_three_hinf_gen_plant.png Normal file
View File

Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 16 KiB

244
journal/figs/comp_filter_three_hinf_gen_plant.svg Normal file
View File

@ -0,0 +1,244 @@
<?xml version="1.0" encoding="UTF-8"?>
<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="184.799pt" height="186.709pt" viewBox="0 0 184.799 186.709" version="1.2">
<defs>
<g>
<symbol overflow="visible" id="glyph0-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph0-1">
<path style="stroke:none;" d="M 7.5 -5.296875 C 7.5 -6.0625 6.84375 -6.796875 5.546875 -6.796875 L 2.3125 -6.796875 C 2.125 -6.796875 2.015625 -6.796875 2.015625 -6.609375 C 2.015625 -6.484375 2.109375 -6.484375 2.3125 -6.484375 C 2.4375 -6.484375 2.625 -6.484375 2.734375 -6.46875 C 2.890625 -6.453125 2.953125 -6.421875 2.953125 -6.3125 C 2.953125 -6.265625 2.953125 -6.234375 2.921875 -6.125 L 1.578125 -0.78125 C 1.484375 -0.390625 1.46875 -0.3125 0.671875 -0.3125 C 0.5 -0.3125 0.390625 -0.3125 0.390625 -0.125 C 0.390625 0 0.515625 0 0.546875 0 L 1.8125 -0.03125 L 2.453125 -0.015625 C 2.671875 -0.015625 2.890625 0 3.09375 0 C 3.171875 0 3.296875 0 3.296875 -0.203125 C 3.296875 -0.3125 3.203125 -0.3125 3.015625 -0.3125 C 2.640625 -0.3125 2.375 -0.3125 2.375 -0.484375 C 2.375 -0.546875 2.390625 -0.59375 2.40625 -0.65625 L 3.015625 -3.140625 L 4.703125 -3.140625 C 6.125 -3.140625 7.5 -4.1875 7.5 -5.296875 Z M 6.578125 -5.53125 C 6.578125 -5.140625 6.375 -4.296875 5.984375 -3.9375 C 5.5 -3.484375 4.890625 -3.40625 4.453125 -3.40625 L 3.0625 -3.40625 L 3.734375 -6.109375 C 3.828125 -6.453125 3.84375 -6.484375 4.265625 -6.484375 L 5.21875 -6.484375 C 6.046875 -6.484375 6.578125 -6.21875 6.578125 -5.53125 Z M 6.578125 -5.53125 "/>
</symbol>
<symbol overflow="visible" id="glyph0-2">
<path style="stroke:none;" d="M 3.9375 -1.546875 C 3.9375 -1.90625 3.75 -2.15625 3.640625 -2.265625 C 3.359375 -2.546875 3.078125 -2.59375 2.640625 -2.6875 C 2.28125 -2.765625 1.875 -2.84375 1.875 -3.28125 C 1.875 -3.578125 2.125 -4.1875 3 -4.1875 C 3.25 -4.1875 3.75 -4.109375 3.890625 -3.71875 C 3.609375 -3.71875 3.40625 -3.5 3.40625 -3.28125 C 3.40625 -3.140625 3.5 -2.984375 3.71875 -2.984375 C 3.9375 -2.984375 4.1875 -3.15625 4.1875 -3.546875 C 4.1875 -3.984375 3.75 -4.40625 3 -4.40625 C 1.6875 -4.40625 1.3125 -3.390625 1.3125 -2.953125 C 1.3125 -2.171875 2.046875 -2.015625 2.34375 -1.953125 C 2.859375 -1.859375 3.375 -1.75 3.375 -1.203125 C 3.375 -0.953125 3.140625 -0.109375 1.953125 -0.109375 C 1.8125 -0.109375 1.046875 -0.109375 0.8125 -0.640625 C 1.1875 -0.59375 1.4375 -0.890625 1.4375 -1.171875 C 1.4375 -1.390625 1.28125 -1.515625 1.078125 -1.515625 C 0.8125 -1.515625 0.515625 -1.296875 0.515625 -0.859375 C 0.515625 -0.28125 1.078125 0.109375 1.9375 0.109375 C 3.546875 0.109375 3.9375 -1.09375 3.9375 -1.546875 Z M 3.9375 -1.546875 "/>
</symbol>
<symbol overflow="visible" id="glyph0-3">
<path style="stroke:none;" d="M 10.4375 -6.6875 C 10.4375 -6.734375 10.390625 -6.796875 10.3125 -6.796875 C 10.078125 -6.796875 9.796875 -6.765625 9.546875 -6.765625 C 9.203125 -6.765625 8.84375 -6.796875 8.515625 -6.796875 C 8.453125 -6.796875 8.328125 -6.796875 8.328125 -6.609375 C 8.328125 -6.5 8.40625 -6.484375 8.46875 -6.484375 C 8.703125 -6.484375 9.0625 -6.40625 9.0625 -6.09375 C 9.0625 -5.984375 9 -5.90625 8.921875 -5.765625 L 6.234375 -1.078125 L 5.875 -6.015625 C 5.859375 -6.21875 5.84375 -6.484375 6.5625 -6.484375 C 6.734375 -6.484375 6.84375 -6.484375 6.84375 -6.6875 C 6.84375 -6.78125 6.734375 -6.796875 6.6875 -6.796875 C 6.296875 -6.796875 5.875 -6.765625 5.46875 -6.765625 C 5.25 -6.765625 4.65625 -6.796875 4.421875 -6.796875 C 4.375 -6.796875 4.234375 -6.796875 4.234375 -6.59375 C 4.234375 -6.484375 4.34375 -6.484375 4.484375 -6.484375 C 4.921875 -6.484375 4.984375 -6.421875 5 -6.234375 L 5.0625 -5.46875 L 2.546875 -1.078125 L 2.171875 -6.125 C 2.171875 -6.25 2.171875 -6.484375 2.921875 -6.484375 C 3.03125 -6.484375 3.140625 -6.484375 3.140625 -6.6875 C 3.140625 -6.796875 3 -6.796875 3 -6.796875 C 2.59375 -6.796875 2.1875 -6.765625 1.765625 -6.765625 C 1.421875 -6.765625 1.0625 -6.796875 0.734375 -6.796875 C 0.671875 -6.796875 0.546875 -6.796875 0.546875 -6.609375 C 0.546875 -6.484375 0.640625 -6.484375 0.796875 -6.484375 C 1.296875 -6.484375 1.296875 -6.40625 1.328125 -6.125 L 1.765625 -0.046875 C 1.78125 0.140625 1.796875 0.21875 1.9375 0.21875 C 2.046875 0.21875 2.078125 0.15625 2.171875 0.015625 L 5.09375 -5.0625 L 5.46875 -0.046875 C 5.484375 0.171875 5.5 0.21875 5.625 0.21875 C 5.75 0.21875 5.8125 0.125 5.859375 0.03125 L 9.125 -5.640625 C 9.375 -6.078125 9.609375 -6.4375 10.234375 -6.484375 C 10.328125 -6.5 10.4375 -6.5 10.4375 -6.6875 Z M 10.4375 -6.6875 "/>
</symbol>
<symbol overflow="visible" id="glyph0-4">
<path style="stroke:none;" d="M 6.875 -3.703125 C 6.875 -4.25 6.609375 -4.40625 6.4375 -4.40625 C 6.1875 -4.40625 5.9375 -4.140625 5.9375 -3.921875 C 5.9375 -3.796875 6 -3.734375 6.09375 -3.65625 C 6.203125 -3.546875 6.453125 -3.28125 6.453125 -2.8125 C 6.453125 -2.46875 6.15625 -1.5 5.90625 -0.984375 C 5.640625 -0.453125 5.296875 -0.109375 4.8125 -0.109375 C 4.34375 -0.109375 4.078125 -0.40625 4.078125 -0.96875 C 4.078125 -1.25 4.140625 -1.5625 4.1875 -1.703125 L 4.59375 -3.375 C 4.640625 -3.59375 4.75 -3.96875 4.75 -4.015625 C 4.75 -4.203125 4.609375 -4.296875 4.453125 -4.296875 C 4.34375 -4.296875 4.15625 -4.203125 4.09375 -4.015625 C 4.0625 -3.9375 3.59375 -2.046875 3.515625 -1.78125 C 3.453125 -1.484375 3.4375 -1.296875 3.4375 -1.125 C 3.4375 -1.015625 3.4375 -1 3.4375 -0.953125 C 3.21875 -0.421875 2.921875 -0.109375 2.53125 -0.109375 C 1.734375 -0.109375 1.734375 -0.84375 1.734375 -1.015625 C 1.734375 -1.328125 1.78125 -1.71875 2.25 -2.953125 C 2.359375 -3.25 2.421875 -3.390625 2.421875 -3.578125 C 2.421875 -4.03125 2.09375 -4.40625 1.609375 -4.40625 C 0.65625 -4.40625 0.28125 -2.953125 0.28125 -2.859375 C 0.28125 -2.765625 0.40625 -2.765625 0.40625 -2.765625 C 0.5 -2.765625 0.515625 -2.78125 0.5625 -2.953125 C 0.84375 -3.875 1.21875 -4.1875 1.578125 -4.1875 C 1.65625 -4.1875 1.828125 -4.171875 1.828125 -3.859375 C 1.828125 -3.609375 1.71875 -3.328125 1.640625 -3.15625 C 1.203125 -1.984375 1.078125 -1.515625 1.078125 -1.140625 C 1.078125 -0.234375 1.75 0.109375 2.5 0.109375 C 2.671875 0.109375 3.140625 0.109375 3.53125 -0.59375 C 3.796875 0.046875 4.484375 0.109375 4.78125 0.109375 C 5.53125 0.109375 5.96875 -0.515625 6.21875 -1.109375 C 6.5625 -1.890625 6.875 -3.21875 6.875 -3.703125 Z M 6.875 -3.703125 "/>
</symbol>
<symbol overflow="visible" id="glyph0-5">
<path style="stroke:none;" d="M 4.328125 -1.421875 C 4.328125 -1.46875 4.296875 -1.515625 4.203125 -1.515625 C 4.125 -1.515625 4.09375 -1.46875 4.078125 -1.390625 C 3.84375 -0.75 3.203125 -0.5625 2.875 -0.5625 C 2.671875 -0.5625 2.5 -0.609375 2.296875 -0.6875 C 1.953125 -0.8125 1.796875 -0.859375 1.59375 -0.859375 C 1.59375 -0.859375 1.40625 -0.859375 1.328125 -0.828125 C 1.859375 -1.40625 2.15625 -1.65625 2.515625 -1.953125 C 2.515625 -1.953125 3.125 -2.5 3.484375 -2.859375 C 4.421875 -3.78125 4.640625 -4.265625 4.640625 -4.296875 C 4.640625 -4.40625 4.53125 -4.40625 4.53125 -4.40625 C 4.46875 -4.40625 4.4375 -4.375 4.390625 -4.296875 C 4.09375 -3.8125 3.875 -3.65625 3.640625 -3.65625 C 3.40625 -3.65625 3.28125 -3.796875 3.140625 -3.96875 C 2.953125 -4.203125 2.78125 -4.40625 2.453125 -4.40625 C 1.703125 -4.40625 1.25 -3.46875 1.25 -3.265625 C 1.25 -3.21875 1.28125 -3.15625 1.359375 -3.15625 C 1.453125 -3.15625 1.46875 -3.203125 1.5 -3.265625 C 1.6875 -3.71875 2.265625 -3.734375 2.34375 -3.734375 C 2.546875 -3.734375 2.734375 -3.65625 2.96875 -3.578125 C 3.359375 -3.4375 3.46875 -3.4375 3.734375 -3.4375 C 3.375 -3 2.53125 -2.296875 2.34375 -2.125 L 1.453125 -1.296875 C 0.78125 -0.625 0.421875 -0.0625 0.421875 0.015625 C 0.421875 0.109375 0.546875 0.109375 0.546875 0.109375 C 0.625 0.109375 0.640625 0.09375 0.703125 -0.015625 C 0.9375 -0.375 1.234375 -0.640625 1.546875 -0.640625 C 1.78125 -0.640625 1.875 -0.546875 2.125 -0.265625 C 2.296875 -0.046875 2.484375 0.109375 2.765625 0.109375 C 3.75 0.109375 4.328125 -1.15625 4.328125 -1.421875 Z M 4.328125 -1.421875 "/>
</symbol>
<symbol overflow="visible" id="glyph0-6">
<path style="stroke:none;" d="M 5.40625 -1.421875 C 5.40625 -1.515625 5.3125 -1.515625 5.28125 -1.515625 C 5.1875 -1.515625 5.171875 -1.484375 5.140625 -1.34375 C 5 -0.78125 4.8125 -0.109375 4.40625 -0.109375 C 4.1875 -0.109375 4.09375 -0.234375 4.09375 -0.5625 C 4.09375 -0.78125 4.203125 -1.25 4.296875 -1.609375 L 4.5625 -2.671875 C 4.59375 -2.828125 4.703125 -3.203125 4.734375 -3.359375 C 4.78125 -3.578125 4.890625 -3.96875 4.890625 -4.015625 C 4.890625 -4.203125 4.75 -4.296875 4.59375 -4.296875 C 4.546875 -4.296875 4.296875 -4.28125 4.203125 -3.9375 L 3.46875 -0.953125 C 3.453125 -0.90625 3.0625 -0.109375 2.328125 -0.109375 C 1.8125 -0.109375 1.71875 -0.5625 1.71875 -0.921875 C 1.71875 -1.484375 1.984375 -2.265625 2.25 -2.953125 C 2.375 -3.25 2.421875 -3.390625 2.421875 -3.578125 C 2.421875 -4.03125 2.09375 -4.40625 1.609375 -4.40625 C 0.65625 -4.40625 0.28125 -2.953125 0.28125 -2.859375 C 0.28125 -2.765625 0.40625 -2.765625 0.40625 -2.765625 C 0.5 -2.765625 0.515625 -2.78125 0.5625 -2.953125 C 0.8125 -3.8125 1.1875 -4.1875 1.578125 -4.1875 C 1.65625 -4.1875 1.828125 -4.171875 1.828125 -3.859375 C 1.828125 -3.609375 1.71875 -3.328125 1.65625 -3.171875 C 1.28125 -2.1875 1.078125 -1.578125 1.078125 -1.078125 C 1.078125 -0.140625 1.765625 0.109375 2.296875 0.109375 C 2.953125 0.109375 3.3125 -0.34375 3.484375 -0.5625 C 3.59375 -0.15625 3.9375 0.109375 4.375 0.109375 C 4.71875 0.109375 4.953125 -0.125 5.109375 -0.4375 C 5.28125 -0.796875 5.40625 -1.421875 5.40625 -1.421875 Z M 5.40625 -1.421875 "/>
</symbol>
<symbol overflow="visible" id="glyph0-7">
<path style="stroke:none;" d="M 4.65625 -3.703125 C 4.65625 -4.234375 4.40625 -4.40625 4.21875 -4.40625 C 3.96875 -4.40625 3.734375 -4.140625 3.734375 -3.921875 C 3.734375 -3.796875 3.78125 -3.734375 3.890625 -3.625 C 4.09375 -3.421875 4.234375 -3.171875 4.234375 -2.8125 C 4.234375 -2.390625 3.625 -0.109375 2.453125 -0.109375 C 1.953125 -0.109375 1.71875 -0.453125 1.71875 -0.96875 C 1.71875 -1.53125 1.984375 -2.265625 2.296875 -3.078125 C 2.375 -3.25 2.421875 -3.390625 2.421875 -3.578125 C 2.421875 -4.03125 2.09375 -4.40625 1.609375 -4.40625 C 0.671875 -4.40625 0.28125 -2.953125 0.28125 -2.859375 C 0.28125 -2.765625 0.40625 -2.765625 0.40625 -2.765625 C 0.5 -2.765625 0.515625 -2.78125 0.5625 -2.953125 C 0.859375 -3.953125 1.28125 -4.1875 1.578125 -4.1875 C 1.65625 -4.1875 1.828125 -4.1875 1.828125 -3.859375 C 1.828125 -3.609375 1.71875 -3.34375 1.65625 -3.171875 C 1.21875 -2.015625 1.078125 -1.546875 1.078125 -1.125 C 1.078125 -0.046875 1.953125 0.109375 2.421875 0.109375 C 4.09375 0.109375 4.65625 -3.1875 4.65625 -3.703125 Z M 4.65625 -3.703125 "/>
</symbol>
<symbol overflow="visible" id="glyph1-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph1-1">
<path style="stroke:none;" d="M 3.578125 -1.203125 C 3.578125 -1.75 3.140625 -2.296875 2.375 -2.453125 C 3.09375 -2.71875 3.359375 -3.234375 3.359375 -3.671875 C 3.359375 -4.21875 2.734375 -4.625 1.953125 -4.625 C 1.1875 -4.625 0.59375 -4.25 0.59375 -3.6875 C 0.59375 -3.453125 0.75 -3.328125 0.953125 -3.328125 C 1.171875 -3.328125 1.3125 -3.484375 1.3125 -3.671875 C 1.3125 -3.875 1.171875 -4.03125 0.953125 -4.046875 C 1.203125 -4.34375 1.671875 -4.421875 1.9375 -4.421875 C 2.25 -4.421875 2.6875 -4.265625 2.6875 -3.671875 C 2.6875 -3.375 2.59375 -3.046875 2.40625 -2.84375 C 2.1875 -2.578125 1.984375 -2.5625 1.640625 -2.53125 C 1.46875 -2.515625 1.453125 -2.515625 1.421875 -2.515625 C 1.421875 -2.515625 1.34375 -2.5 1.34375 -2.421875 C 1.34375 -2.328125 1.40625 -2.328125 1.53125 -2.328125 L 1.90625 -2.328125 C 2.453125 -2.328125 2.84375 -1.953125 2.84375 -1.203125 C 2.84375 -0.34375 2.328125 -0.078125 1.9375 -0.078125 C 1.65625 -0.078125 1.03125 -0.15625 0.75 -0.578125 C 1.078125 -0.578125 1.15625 -0.8125 1.15625 -0.96875 C 1.15625 -1.1875 0.984375 -1.34375 0.765625 -1.34375 C 0.578125 -1.34375 0.375 -1.21875 0.375 -0.9375 C 0.375 -0.28125 1.09375 0.140625 1.9375 0.140625 C 2.90625 0.140625 3.578125 -0.515625 3.578125 -1.203125 Z M 3.578125 -1.203125 "/>
</symbol>
<symbol overflow="visible" id="glyph1-2">
<path style="stroke:none;" d="M 3.296875 0 L 3.296875 -0.25 L 3.03125 -0.25 C 2.328125 -0.25 2.328125 -0.34375 2.328125 -0.578125 L 2.328125 -4.4375 C 2.328125 -4.625 2.328125 -4.625 2.125 -4.625 C 1.671875 -4.1875 1.046875 -4.1875 0.765625 -4.1875 L 0.765625 -3.9375 C 0.921875 -3.9375 1.390625 -3.9375 1.765625 -4.125 L 1.765625 -0.578125 C 1.765625 -0.34375 1.765625 -0.25 1.078125 -0.25 L 0.8125 -0.25 L 0.8125 0 L 2.046875 -0.03125 Z M 3.296875 0 "/>
</symbol>
<symbol overflow="visible" id="glyph1-3">
<path style="stroke:none;" d="M 3.515625 -1.265625 L 3.28125 -1.265625 C 3.265625 -1.109375 3.1875 -0.703125 3.09375 -0.640625 C 3.046875 -0.59375 2.515625 -0.59375 2.40625 -0.59375 L 1.125 -0.59375 C 1.859375 -1.234375 2.109375 -1.4375 2.515625 -1.765625 C 3.03125 -2.171875 3.515625 -2.609375 3.515625 -3.265625 C 3.515625 -4.109375 2.78125 -4.625 1.890625 -4.625 C 1.03125 -4.625 0.4375 -4.015625 0.4375 -3.375 C 0.4375 -3.03125 0.734375 -2.984375 0.8125 -2.984375 C 0.96875 -2.984375 1.171875 -3.109375 1.171875 -3.359375 C 1.171875 -3.484375 1.125 -3.734375 0.765625 -3.734375 C 0.984375 -4.21875 1.453125 -4.375 1.78125 -4.375 C 2.484375 -4.375 2.84375 -3.828125 2.84375 -3.265625 C 2.84375 -2.65625 2.40625 -2.1875 2.1875 -1.9375 L 0.515625 -0.265625 C 0.4375 -0.203125 0.4375 -0.1875 0.4375 0 L 3.3125 0 Z M 3.515625 -1.265625 "/>
</symbol>
<symbol overflow="visible" id="glyph2-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph2-1">
<path style="stroke:none;" d="M 3.296875 2.390625 C 3.296875 2.359375 3.296875 2.34375 3.125 2.171875 C 1.875 0.921875 1.5625 -0.96875 1.5625 -2.484375 C 1.5625 -4.21875 1.9375 -5.953125 3.171875 -7.203125 C 3.296875 -7.3125 3.296875 -7.328125 3.296875 -7.359375 C 3.296875 -7.4375 3.25 -7.46875 3.1875 -7.46875 C 3.09375 -7.46875 2.203125 -6.78125 1.609375 -5.53125 C 1.109375 -4.421875 0.984375 -3.328125 0.984375 -2.484375 C 0.984375 -1.71875 1.09375 -0.5 1.640625 0.609375 C 2.234375 1.84375 3.09375 2.484375 3.1875 2.484375 C 3.25 2.484375 3.296875 2.453125 3.296875 2.390625 Z M 3.296875 2.390625 "/>
</symbol>
<symbol overflow="visible" id="glyph2-2">
<path style="stroke:none;" d="M 2.875 -2.484375 C 2.875 -3.265625 2.765625 -4.46875 2.21875 -5.59375 C 1.625 -6.8125 0.765625 -7.46875 0.671875 -7.46875 C 0.609375 -7.46875 0.5625 -7.421875 0.5625 -7.359375 C 0.5625 -7.328125 0.5625 -7.3125 0.75 -7.140625 C 1.734375 -6.15625 2.296875 -4.5625 2.296875 -2.484375 C 2.296875 -0.78125 1.9375 0.96875 0.703125 2.21875 C 0.5625 2.34375 0.5625 2.359375 0.5625 2.390625 C 0.5625 2.453125 0.609375 2.484375 0.671875 2.484375 C 0.765625 2.484375 1.65625 1.8125 2.25 0.546875 C 2.75 -0.546875 2.875 -1.65625 2.875 -2.484375 Z M 2.875 -2.484375 "/>
</symbol>
<symbol overflow="visible" id="glyph3-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph3-1">
<path style="stroke:none;" d="M 8.09375 -3 C 8.09375 -3.21875 7.90625 -3.21875 7.71875 -3.21875 L 4.53125 -3.21875 L 4.53125 -6.40625 C 4.53125 -6.5625 4.53125 -6.78125 4.328125 -6.78125 C 4.109375 -6.78125 4.109375 -6.578125 4.109375 -6.40625 L 4.109375 -3.21875 L 0.921875 -3.21875 C 0.765625 -3.21875 0.546875 -3.21875 0.546875 -3.015625 C 0.546875 -2.796875 0.75 -2.796875 0.921875 -2.796875 L 4.109375 -2.796875 L 4.109375 0.390625 C 4.109375 0.5625 4.109375 0.765625 4.3125 0.765625 C 4.53125 0.765625 4.53125 0.578125 4.53125 0.390625 L 4.53125 -2.796875 L 7.71875 -2.796875 C 7.890625 -2.796875 8.09375 -2.796875 8.09375 -3 Z M 8.09375 -3 "/>
</symbol>
<symbol overflow="visible" id="glyph4-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph4-1">
<path style="stroke:none;" d="M 6.90625 -2.484375 C 6.90625 -2.6875 6.71875 -2.6875 6.578125 -2.6875 L 1.15625 -2.6875 C 1.015625 -2.6875 0.828125 -2.6875 0.828125 -2.484375 C 0.828125 -2.296875 1.015625 -2.296875 1.15625 -2.296875 L 6.578125 -2.296875 C 6.71875 -2.296875 6.90625 -2.296875 6.90625 -2.484375 Z M 6.90625 -2.484375 "/>
</symbol>
</g>
<clipPath id="clip1">
<path d="M 0.0429688 42 L 45 42 L 45 44 L 0.0429688 44 Z M 0.0429688 42 "/>
</clipPath>
<clipPath id="clip2">
<path d="M 163 117 L 184.554688 117 L 184.554688 149 L 163 149 Z M 163 117 "/>
</clipPath>
<clipPath id="clip3">
<path d="M 0.0429688 74 L 106 74 L 106 76 L 0.0429688 76 Z M 0.0429688 74 "/>
</clipPath>
<clipPath id="clip4">
<path d="M 0.0429688 106 L 106 106 L 106 108 L 0.0429688 108 Z M 0.0429688 106 "/>
</clipPath>
<clipPath id="clip5">
<path d="M 163 27 L 184.554688 27 L 184.554688 59 L 163 59 Z M 163 27 "/>
</clipPath>
<clipPath id="clip6">
<path d="M 163 59 L 184.554688 59 L 184.554688 92 L 163 92 Z M 163 59 "/>
</clipPath>
<clipPath id="clip7">
<path d="M 163 91 L 184.554688 91 L 184.554688 124 L 163 124 Z M 163 91 "/>
</clipPath>
</defs>
<g id="surface1">
<path style="fill-rule:nonzero;fill:rgb(89.99939%,89.99939%,89.99939%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -70.86535 -63.781263 L 70.866926 -63.781263 L 70.866926 63.780524 L -70.86535 63.780524 Z M -70.86535 -63.781263 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-1" x="80.86947" y="11.275426"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-1" x="87.255517" y="12.767102"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-1" x="91.718561" y="11.275426"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-2" x="95.586533" y="11.275426"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-2" x="100.249266" y="11.275426"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(100%,100%,100%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 18.085571 27.227548 L 51.023547 27.227548 L 51.023547 49.907372 L 18.085571 49.907372 Z M 18.085571 27.227548 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-3" x="113.866043" y="45.731724"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-2" x="123.260401" y="47.224397"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-1" x="127.723445" y="45.731724"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-2" x="131.591416" y="45.731724"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-2" x="136.254149" y="45.731724"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(100%,100%,100%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 18.085571 -4.91231 L 51.023547 -4.91231 L 51.023547 17.767514 L 18.085571 17.767514 Z M 18.085571 -4.91231 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-3" x="113.866043" y="77.820711"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-3" x="123.260401" y="79.312386"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-1" x="127.723445" y="77.820711"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-2" x="131.591416" y="77.820711"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-2" x="136.254149" y="77.820711"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(100%,100%,100%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 18.085571 -37.048256 L 51.023547 -37.048256 L 51.023547 -14.372344 L 18.085571 -14.372344 Z M 18.085571 -37.048256 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-3" x="113.866043" y="109.909697"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-1" x="123.260401" y="111.401373"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-1" x="127.723445" y="109.909697"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-2" x="131.591416" y="109.909697"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-2" x="136.254149" y="109.909697"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(100%,100%,100%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 0.577858 38.569416 C 0.577858 42.419157 -2.544188 45.541203 -6.393929 45.541203 C -10.247582 45.541203 -13.369628 42.419157 -13.369628 38.569416 C -13.369628 34.715763 -10.247582 31.593717 -6.393929 31.593717 C -2.544188 31.593717 0.577858 34.715763 0.577858 38.569416 Z M 0.577858 38.569416 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph3-1" x="81.778054" y="46.157061"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph4-1" x="76.382464" y="57.999603"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(100%,100%,100%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -30.376562 38.569416 C -30.376562 42.419157 -33.498608 45.541203 -37.352261 45.541203 C -41.202001 45.541203 -44.324047 42.419157 -44.324047 38.569416 C -44.324047 34.715763 -41.202001 31.593717 -37.352261 31.593717 C -33.498608 31.593717 -30.376562 34.715763 -30.376562 38.569416 Z M -30.376562 38.569416 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph3-1" x="50.871225" y="46.157061"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph4-1" x="45.475634" y="57.999603"/>
</g>
<g clip-path="url(#clip1)" clip-rule="nonzero">
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -91.209508 38.569416 L -48.960168 38.569416 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.052524 0.0019463 L 1.608108 1.684252 L 3.086972 0.0019463 L 1.608108 -1.684271 Z M 6.052524 0.0019463 " transform="matrix(0.998444,0,0,-0.998444,40.777207,43.244131)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-4" x="5.241343" y="39.432541"/>
</g>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -6.393929 -25.7103 L -6.393929 26.961509 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.052039 -0.000801368 L 1.607622 1.681504 L 3.086486 -0.000801368 L 1.607622 -1.683107 Z M 6.052039 -0.000801368 " transform="matrix(0,-0.998444,-0.998444,0,86.108575,57.667621)"/>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 88.097656 107.421875 C 88.097656 106.324219 87.207031 105.433594 86.109375 105.433594 C 85.011719 105.433594 84.117188 106.324219 84.117188 107.421875 C 84.117188 108.519531 85.011719 109.410156 86.109375 109.410156 C 87.207031 109.410156 88.097656 108.519531 88.097656 107.421875 Z M 88.097656 107.421875 "/>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -37.352261 6.429558 L -37.352261 26.961509 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.052039 0.00148068 L 1.607622 1.683786 L 3.086486 0.00148068 L 1.607622 -1.684737 Z M 6.052039 0.00148068 " transform="matrix(0,-0.998444,-0.998444,0,55.200697,57.667621)"/>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 57.191406 75.332031 C 57.191406 74.234375 56.300781 73.34375 55.199219 73.34375 C 54.101562 73.34375 53.210938 74.234375 53.210938 75.332031 C 53.210938 76.433594 54.101562 77.324219 55.199219 77.324219 C 56.300781 77.324219 57.191406 76.433594 57.191406 75.332031 Z M 57.191406 75.332031 "/>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -58.498449 38.569416 L -58.498449 -51.422187 L 86.574964 -51.422187 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 182.144531 133.09375 L 177.703125 131.414062 L 179.179688 133.09375 L 177.703125 134.773438 Z M 182.144531 133.09375 "/>
<g clip-path="url(#clip2)" clip-rule="nonzero">
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.055146 -0.000186606 L 1.606817 1.682119 L 3.085681 -0.000186606 L 1.606817 -1.682492 Z M 6.055146 -0.000186606 " transform="matrix(0.998444,0,0,-0.998444,176.098808,133.093564)"/>
</g>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 36.074219 43.242188 C 36.074219 42.144531 35.183594 41.253906 34.085938 41.253906 C 32.988281 41.253906 32.097656 42.144531 32.097656 43.242188 C 32.097656 44.34375 32.988281 45.234375 34.085938 45.234375 C 35.183594 45.234375 36.074219 44.34375 36.074219 43.242188 Z M 36.074219 43.242188 "/>
<g clip-path="url(#clip3)" clip-rule="nonzero">
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -91.209508 6.429558 L 12.952584 6.429558 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.053196 0.00131812 L 1.608779 1.683624 L 3.087643 0.00131812 L 1.608779 -1.6849 Z M 6.053196 0.00131812 " transform="matrix(0.998444,0,0,-0.998444,102.592943,75.333347)"/>
<g clip-path="url(#clip4)" clip-rule="nonzero">
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -91.209508 -25.7103 L 12.952584 -25.7103 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.053196 0.000679939 L 1.608779 1.682985 L 3.087643 0.000679939 L 1.608779 -1.681625 Z M 6.053196 0.000679939 " transform="matrix(0.998444,0,0,-0.998444,102.592943,107.422554)"/>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -30.376562 38.569416 L -18.001836 38.569416 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.054806 0.0019463 L 1.61039 1.684252 L 3.089254 0.0019463 L 1.61039 -1.684271 Z M 6.054806 0.0019463 " transform="matrix(0.998444,0,0,-0.998444,71.685085,43.244131)"/>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 0.577858 38.569416 L 12.952584 38.569416 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.053196 0.0019463 L 1.608779 1.684252 L 3.087643 0.0019463 L 1.608779 -1.684271 Z M 6.053196 0.0019463 " transform="matrix(0.998444,0,0,-0.998444,102.592943,43.244131)"/>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 51.524326 38.569416 L 86.574964 38.569416 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 182.144531 43.242188 L 177.703125 41.5625 L 179.179688 43.242188 L 177.703125 44.925781 Z M 182.144531 43.242188 "/>
<g clip-path="url(#clip5)" clip-rule="nonzero">
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.055146 0.0019463 L 1.606817 1.684252 L 3.085681 0.0019463 L 1.606817 -1.684271 Z M 6.055146 0.0019463 " transform="matrix(0.998444,0,0,-0.998444,176.098808,43.244131)"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-5" x="170.657529" y="37.939868"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-2" x="175.283319" y="39.432541"/>
</g>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 51.524326 6.429558 L 86.574964 6.429558 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 182.144531 75.332031 L 177.703125 73.652344 L 179.179688 75.332031 L 177.703125 77.015625 Z M 182.144531 75.332031 "/>
<g clip-path="url(#clip6)" clip-rule="nonzero">
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.055146 0.00131812 L 1.606817 1.683624 L 3.085681 0.00131812 L 1.606817 -1.6849 Z M 6.055146 0.00131812 " transform="matrix(0.998444,0,0,-0.998444,176.098808,75.333347)"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-5" x="170.657529" y="70.028855"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-3" x="175.283319" y="71.521528"/>
</g>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 51.524326 -25.7103 L 86.574964 -25.7103 " transform="matrix(0.998444,0,0,-0.998444,92.493354,81.751584)"/>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 182.144531 107.421875 L 177.703125 105.742188 L 179.179688 107.421875 L 177.703125 109.101562 Z M 182.144531 107.421875 "/>
<g clip-path="url(#clip7)" clip-rule="nonzero">
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.055146 0.000679939 L 1.606817 1.682985 L 3.085681 0.000679939 L 1.606817 -1.681625 Z M 6.055146 0.000679939 " transform="matrix(0.998444,0,0,-0.998444,176.098808,107.422554)"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-5" x="170.657529" y="102.117842"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-1" x="175.283319" y="103.609517"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-6" x="3.857499" y="70.028855"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-2" x="9.552623" y="71.521528"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-6" x="3.857499" y="102.117842"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-3" x="9.552623" y="103.609517"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-7" x="172.612482" y="129.280507"/>
</g>
</g>
</svg>
Side by Side

After

Width:  |  Height:  |  Size: 32 KiB

BIN
journal/figs/feedback_synthesis_architecture.pdf
View File

Binary file not shown.

BIN
journal/figs/feedback_synthesis_architecture.png
View File

Binary file not shown.
Side by Side Swipe Overlay

Before

Width:  |  Height:  |  Size: 8.1 KiB

After

Width:  |  Height:  |  Size: 9.3 KiB

158
journal/figs/feedback_synthesis_architecture.svg
View File

@ -1,157 +1,179 @@
<?xml version="1.0" encoding="UTF-8"?>
<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="235.208pt" height="106.344pt" viewBox="0 0 235.208 106.344" version="1.2">
<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="239.8pt" height="100.675pt" viewBox="0 0 239.8 100.675" version="1.2">
<defs>
<g>
<symbol overflow="visible" id="glyph0-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph0-1">
<path style="stroke:none;" d="M 10.375 -6.65625 C 10.375 -6.703125 10.34375 -6.765625 10.265625 -6.765625 C 10.03125 -6.765625 9.75 -6.734375 9.5 -6.734375 C 9.15625 -6.734375 8.796875 -6.765625 8.46875 -6.765625 C 8.40625 -6.765625 8.28125 -6.765625 8.28125 -6.578125 C 8.28125 -6.46875 8.375 -6.453125 8.4375 -6.453125 C 8.671875 -6.453125 9.015625 -6.375 9.015625 -6.078125 C 9.015625 -5.96875 8.96875 -5.890625 8.890625 -5.75 L 6.21875 -1.078125 L 5.84375 -6 C 5.828125 -6.1875 5.8125 -6.453125 6.53125 -6.453125 C 6.703125 -6.453125 6.8125 -6.453125 6.8125 -6.65625 C 6.8125 -6.75 6.703125 -6.765625 6.65625 -6.765625 C 6.265625 -6.765625 5.84375 -6.734375 5.453125 -6.734375 C 5.21875 -6.734375 4.640625 -6.765625 4.40625 -6.765625 C 4.34375 -6.765625 4.21875 -6.765625 4.21875 -6.5625 C 4.21875 -6.453125 4.3125 -6.453125 4.453125 -6.453125 C 4.890625 -6.453125 4.96875 -6.40625 4.984375 -6.21875 L 5.046875 -5.453125 L 2.53125 -1.078125 L 2.15625 -6.109375 C 2.15625 -6.21875 2.15625 -6.453125 2.90625 -6.453125 C 3.015625 -6.453125 3.125 -6.453125 3.125 -6.65625 C 3.125 -6.765625 2.984375 -6.765625 2.984375 -6.765625 C 2.578125 -6.765625 2.171875 -6.734375 1.765625 -6.734375 C 1.421875 -6.734375 1.0625 -6.765625 0.71875 -6.765625 C 0.671875 -6.765625 0.546875 -6.765625 0.546875 -6.578125 C 0.546875 -6.453125 0.640625 -6.453125 0.796875 -6.453125 C 1.28125 -6.453125 1.296875 -6.375 1.3125 -6.09375 L 1.765625 -0.046875 C 1.765625 0.140625 1.78125 0.21875 1.921875 0.21875 C 2.046875 0.21875 2.078125 0.15625 2.15625 0.015625 L 5.078125 -5.046875 L 5.4375 -0.046875 C 5.453125 0.171875 5.484375 0.21875 5.59375 0.21875 C 5.71875 0.21875 5.78125 0.125 5.828125 0.03125 L 9.078125 -5.609375 C 9.328125 -6.046875 9.5625 -6.40625 10.1875 -6.453125 C 10.28125 -6.46875 10.375 -6.46875 10.375 -6.65625 Z M 10.375 -6.65625 "/>
<path style="stroke:none;" d="M 10.421875 -6.671875 C 10.421875 -6.71875 10.375 -6.78125 10.296875 -6.78125 C 10.0625 -6.78125 9.78125 -6.75 9.53125 -6.75 C 9.1875 -6.75 8.828125 -6.78125 8.5 -6.78125 C 8.4375 -6.78125 8.3125 -6.78125 8.3125 -6.59375 C 8.3125 -6.484375 8.390625 -6.484375 8.453125 -6.484375 C 8.703125 -6.46875 9.046875 -6.390625 9.046875 -6.09375 C 9.046875 -5.984375 9 -5.90625 8.90625 -5.765625 L 6.234375 -1.078125 L 5.859375 -6.015625 C 5.859375 -6.21875 5.828125 -6.46875 6.5625 -6.484375 C 6.734375 -6.484375 6.828125 -6.484375 6.828125 -6.671875 C 6.828125 -6.78125 6.71875 -6.78125 6.671875 -6.78125 C 6.28125 -6.78125 5.859375 -6.75 5.46875 -6.75 C 5.234375 -6.75 4.65625 -6.78125 4.421875 -6.78125 C 4.359375 -6.78125 4.234375 -6.78125 4.234375 -6.59375 C 4.234375 -6.484375 4.328125 -6.484375 4.46875 -6.484375 C 4.90625 -6.484375 4.984375 -6.421875 5 -6.234375 L 5.0625 -5.46875 L 2.546875 -1.078125 L 2.171875 -6.125 C 2.171875 -6.234375 2.171875 -6.46875 2.921875 -6.484375 C 3.015625 -6.484375 3.125 -6.484375 3.125 -6.671875 C 3.125 -6.78125 2.984375 -6.78125 2.984375 -6.78125 C 2.59375 -6.78125 2.171875 -6.75 1.765625 -6.75 C 1.421875 -6.75 1.0625 -6.78125 0.71875 -6.78125 C 0.671875 -6.78125 0.546875 -6.78125 0.546875 -6.59375 C 0.546875 -6.484375 0.640625 -6.484375 0.796875 -6.484375 C 1.296875 -6.484375 1.296875 -6.390625 1.328125 -6.109375 L 1.765625 -0.046875 C 1.78125 0.140625 1.78125 0.21875 1.921875 0.21875 C 2.046875 0.21875 2.078125 0.15625 2.171875 0.015625 L 5.09375 -5.0625 L 5.453125 -0.046875 C 5.46875 0.171875 5.5 0.21875 5.609375 0.21875 C 5.75 0.21875 5.796875 0.125 5.859375 0.03125 L 9.109375 -5.640625 C 9.359375 -6.078125 9.59375 -6.421875 10.21875 -6.484375 C 10.3125 -6.484375 10.421875 -6.484375 10.421875 -6.671875 Z M 10.421875 -6.671875 "/>
</symbol>
<symbol overflow="visible" id="glyph0-2">
<path style="stroke:none;" d="M 6.84375 -3.6875 C 6.84375 -4.234375 6.578125 -4.375 6.40625 -4.375 C 6.15625 -4.375 5.90625 -4.125 5.90625 -3.90625 C 5.90625 -3.78125 5.96875 -3.71875 6.0625 -3.640625 C 6.171875 -3.53125 6.421875 -3.265625 6.421875 -2.796875 C 6.421875 -2.453125 6.125 -1.484375 5.875 -0.984375 C 5.609375 -0.453125 5.265625 -0.109375 4.78125 -0.109375 C 4.3125 -0.109375 4.046875 -0.40625 4.046875 -0.96875 C 4.046875 -1.25 4.125 -1.5625 4.15625 -1.6875 L 4.578125 -3.359375 C 4.625 -3.578125 4.71875 -3.9375 4.71875 -4 C 4.71875 -4.1875 4.59375 -4.265625 4.4375 -4.265625 C 4.3125 -4.265625 4.140625 -4.1875 4.078125 -4 C 4.046875 -3.921875 3.578125 -2.03125 3.5 -1.765625 C 3.4375 -1.46875 3.421875 -1.296875 3.421875 -1.125 C 3.421875 -1.015625 3.421875 -0.984375 3.421875 -0.9375 C 3.203125 -0.421875 2.90625 -0.109375 2.515625 -0.109375 C 1.71875 -0.109375 1.71875 -0.84375 1.71875 -1.015625 C 1.71875 -1.328125 1.765625 -1.71875 2.234375 -2.9375 C 2.34375 -3.234375 2.40625 -3.375 2.40625 -3.5625 C 2.40625 -4.015625 2.078125 -4.375 1.59375 -4.375 C 0.65625 -4.375 0.28125 -2.9375 0.28125 -2.859375 C 0.28125 -2.75 0.40625 -2.75 0.40625 -2.75 C 0.5 -2.75 0.515625 -2.78125 0.5625 -2.9375 C 0.828125 -3.859375 1.21875 -4.15625 1.5625 -4.15625 C 1.65625 -4.15625 1.8125 -4.15625 1.8125 -3.828125 C 1.8125 -3.59375 1.703125 -3.3125 1.640625 -3.140625 C 1.203125 -1.96875 1.078125 -1.515625 1.078125 -1.140625 C 1.078125 -0.234375 1.75 0.109375 2.484375 0.109375 C 2.65625 0.109375 3.125 0.109375 3.515625 -0.578125 C 3.78125 0.046875 4.453125 0.109375 4.75 0.109375 C 5.5 0.109375 5.9375 -0.515625 6.1875 -1.109375 C 6.53125 -1.875 6.84375 -3.203125 6.84375 -3.6875 Z M 6.84375 -3.6875 "/>
<path style="stroke:none;" d="M 3.9375 -1.546875 C 3.9375 -1.890625 3.75 -2.140625 3.625 -2.265625 C 3.359375 -2.546875 3.078125 -2.59375 2.640625 -2.6875 C 2.28125 -2.765625 1.875 -2.828125 1.875 -3.28125 C 1.875 -3.5625 2.109375 -4.171875 2.984375 -4.171875 C 3.234375 -4.171875 3.734375 -4.109375 3.890625 -3.71875 C 3.609375 -3.703125 3.40625 -3.484375 3.40625 -3.265625 C 3.40625 -3.125 3.5 -2.984375 3.71875 -2.984375 C 3.9375 -2.984375 4.171875 -3.15625 4.171875 -3.53125 C 4.171875 -3.984375 3.75 -4.390625 2.984375 -4.390625 C 1.671875 -4.390625 1.3125 -3.375 1.3125 -2.9375 C 1.3125 -2.171875 2.046875 -2.015625 2.328125 -1.953125 C 2.859375 -1.859375 3.375 -1.75 3.375 -1.203125 C 3.375 -0.9375 3.140625 -0.109375 1.953125 -0.109375 C 1.8125 -0.109375 1.046875 -0.109375 0.8125 -0.640625 C 1.1875 -0.59375 1.4375 -0.890625 1.4375 -1.15625 C 1.4375 -1.390625 1.28125 -1.515625 1.078125 -1.515625 C 0.8125 -1.515625 0.515625 -1.296875 0.515625 -0.859375 C 0.515625 -0.28125 1.078125 0.109375 1.9375 0.109375 C 3.546875 0.109375 3.9375 -1.09375 3.9375 -1.546875 Z M 3.9375 -1.546875 "/>
</symbol>
<symbol overflow="visible" id="glyph0-3">
<path style="stroke:none;" d="M 4.640625 -3.6875 C 4.640625 -4.21875 4.375 -4.375 4.203125 -4.375 C 3.953125 -4.375 3.71875 -4.125 3.71875 -3.90625 C 3.71875 -3.78125 3.765625 -3.71875 3.875 -3.609375 C 4.078125 -3.40625 4.203125 -3.15625 4.203125 -2.796875 C 4.203125 -2.375 3.609375 -0.109375 2.453125 -0.109375 C 1.9375 -0.109375 1.71875 -0.453125 1.71875 -0.96875 C 1.71875 -1.53125 1.984375 -2.25 2.28125 -3.078125 C 2.359375 -3.234375 2.40625 -3.375 2.40625 -3.5625 C 2.40625 -4.015625 2.09375 -4.375 1.59375 -4.375 C 0.65625 -4.375 0.28125 -2.9375 0.28125 -2.859375 C 0.28125 -2.75 0.40625 -2.75 0.40625 -2.75 C 0.5 -2.75 0.515625 -2.78125 0.5625 -2.9375 C 0.859375 -3.9375 1.28125 -4.15625 1.5625 -4.15625 C 1.640625 -4.15625 1.8125 -4.15625 1.8125 -3.84375 C 1.8125 -3.59375 1.71875 -3.328125 1.640625 -3.15625 C 1.203125 -2 1.078125 -1.546875 1.078125 -1.125 C 1.078125 -0.046875 1.953125 0.109375 2.40625 0.109375 C 4.078125 0.109375 4.640625 -3.171875 4.640625 -3.6875 Z M 4.640625 -3.6875 "/>
<path style="stroke:none;" d="M 6.859375 -3.703125 C 6.859375 -4.25 6.59375 -4.390625 6.421875 -4.390625 C 6.1875 -4.390625 5.9375 -4.140625 5.9375 -3.921875 C 5.9375 -3.78125 6 -3.734375 6.078125 -3.640625 C 6.1875 -3.53125 6.4375 -3.28125 6.4375 -2.796875 C 6.4375 -2.46875 6.15625 -1.484375 5.890625 -0.984375 C 5.640625 -0.453125 5.28125 -0.109375 4.796875 -0.109375 C 4.328125 -0.109375 4.0625 -0.40625 4.0625 -0.96875 C 4.0625 -1.25 4.140625 -1.5625 4.171875 -1.703125 L 4.59375 -3.375 C 4.640625 -3.59375 4.734375 -3.953125 4.734375 -4.015625 C 4.734375 -4.1875 4.59375 -4.28125 4.453125 -4.28125 C 4.328125 -4.28125 4.15625 -4.203125 4.078125 -4 C 4.046875 -3.9375 3.59375 -2.03125 3.515625 -1.78125 C 3.453125 -1.484375 3.421875 -1.296875 3.421875 -1.125 C 3.421875 -1.015625 3.421875 -1 3.4375 -0.9375 C 3.203125 -0.421875 2.90625 -0.109375 2.53125 -0.109375 C 1.734375 -0.109375 1.734375 -0.84375 1.734375 -1.015625 C 1.734375 -1.328125 1.78125 -1.71875 2.25 -2.9375 C 2.359375 -3.234375 2.421875 -3.375 2.421875 -3.578125 C 2.421875 -4.03125 2.09375 -4.390625 1.59375 -4.390625 C 0.65625 -4.390625 0.28125 -2.953125 0.28125 -2.859375 C 0.28125 -2.765625 0.40625 -2.765625 0.40625 -2.765625 C 0.5 -2.765625 0.515625 -2.78125 0.5625 -2.9375 C 0.828125 -3.859375 1.21875 -4.171875 1.5625 -4.171875 C 1.65625 -4.171875 1.8125 -4.15625 1.8125 -3.84375 C 1.8125 -3.59375 1.703125 -3.3125 1.640625 -3.15625 C 1.203125 -1.984375 1.078125 -1.515625 1.078125 -1.140625 C 1.078125 -0.234375 1.75 0.109375 2.5 0.109375 C 2.65625 0.109375 3.125 0.109375 3.53125 -0.59375 C 3.78125 0.046875 4.46875 0.109375 4.765625 0.109375 C 5.515625 0.109375 5.953125 -0.515625 6.21875 -1.109375 C 6.546875 -1.890625 6.859375 -3.21875 6.859375 -3.703125 Z M 6.859375 -3.703125 "/>
</symbol>
<symbol overflow="visible" id="glyph0-4">
<path style="stroke:none;" d="M 5.375 -1.421875 C 5.375 -1.515625 5.296875 -1.515625 5.265625 -1.515625 C 5.15625 -1.515625 5.15625 -1.46875 5.125 -1.34375 C 4.984375 -0.78125 4.796875 -0.109375 4.375 -0.109375 C 4.171875 -0.109375 4.078125 -0.234375 4.078125 -0.5625 C 4.078125 -0.78125 4.1875 -1.25 4.265625 -1.59375 L 4.546875 -2.671875 C 4.578125 -2.8125 4.671875 -3.1875 4.71875 -3.34375 C 4.765625 -3.5625 4.859375 -3.9375 4.859375 -4 C 4.859375 -4.1875 4.71875 -4.265625 4.578125 -4.265625 C 4.53125 -4.265625 4.265625 -4.265625 4.1875 -3.921875 L 3.453125 -0.9375 C 3.4375 -0.90625 3.046875 -0.109375 2.3125 -0.109375 C 1.796875 -0.109375 1.703125 -0.5625 1.703125 -0.921875 C 1.703125 -1.46875 1.984375 -2.265625 2.234375 -2.9375 C 2.359375 -3.234375 2.40625 -3.375 2.40625 -3.5625 C 2.40625 -4.015625 2.09375 -4.375 1.59375 -4.375 C 0.65625 -4.375 0.28125 -2.9375 0.28125 -2.859375 C 0.28125 -2.75 0.40625 -2.75 0.40625 -2.75 C 0.5 -2.75 0.515625 -2.78125 0.5625 -2.9375 C 0.8125 -3.796875 1.1875 -4.15625 1.5625 -4.15625 C 1.65625 -4.15625 1.8125 -4.15625 1.8125 -3.828125 C 1.8125 -3.59375 1.703125 -3.3125 1.640625 -3.15625 C 1.28125 -2.171875 1.0625 -1.5625 1.0625 -1.078125 C 1.0625 -0.140625 1.75 0.109375 2.28125 0.109375 C 2.9375 0.109375 3.296875 -0.34375 3.46875 -0.5625 C 3.578125 -0.15625 3.921875 0.109375 4.34375 0.109375 C 4.703125 0.109375 4.921875 -0.125 5.078125 -0.4375 C 5.25 -0.796875 5.375 -1.421875 5.375 -1.421875 Z M 5.375 -1.421875 "/>
<path style="stroke:none;" d="M 4.65625 -3.703125 C 4.65625 -4.234375 4.390625 -4.390625 4.21875 -4.390625 C 3.96875 -4.390625 3.734375 -4.140625 3.734375 -3.921875 C 3.734375 -3.78125 3.78125 -3.734375 3.890625 -3.625 C 4.09375 -3.421875 4.21875 -3.15625 4.21875 -2.796875 C 4.21875 -2.390625 3.625 -0.109375 2.453125 -0.109375 C 1.953125 -0.109375 1.71875 -0.453125 1.71875 -0.96875 C 1.71875 -1.53125 1.984375 -2.25 2.296875 -3.078125 C 2.359375 -3.25 2.421875 -3.390625 2.421875 -3.578125 C 2.421875 -4.03125 2.09375 -4.390625 1.59375 -4.390625 C 0.671875 -4.390625 0.28125 -2.953125 0.28125 -2.859375 C 0.28125 -2.765625 0.40625 -2.765625 0.40625 -2.765625 C 0.5 -2.765625 0.515625 -2.78125 0.5625 -2.9375 C 0.859375 -3.9375 1.28125 -4.171875 1.5625 -4.171875 C 1.65625 -4.171875 1.8125 -4.171875 1.8125 -3.859375 C 1.8125 -3.609375 1.71875 -3.34375 1.65625 -3.15625 C 1.21875 -2 1.078125 -1.546875 1.078125 -1.125 C 1.078125 -0.046875 1.953125 0.109375 2.421875 0.109375 C 4.078125 0.109375 4.65625 -3.1875 4.65625 -3.703125 Z M 4.65625 -3.703125 "/>
</symbol>
<symbol overflow="visible" id="glyph0-5">
<path style="stroke:none;" d="M 4.3125 -1.421875 C 4.3125 -1.46875 4.265625 -1.515625 4.1875 -1.515625 C 4.09375 -1.515625 4.078125 -1.453125 4.046875 -1.390625 C 3.828125 -0.75 3.1875 -0.5625 2.859375 -0.5625 C 2.671875 -0.5625 2.484375 -0.609375 2.28125 -0.6875 C 1.9375 -0.8125 1.796875 -0.859375 1.578125 -0.859375 C 1.578125 -0.859375 1.40625 -0.859375 1.3125 -0.828125 C 1.859375 -1.390625 2.140625 -1.640625 2.5 -1.953125 C 2.5 -1.953125 3.109375 -2.484375 3.46875 -2.84375 C 4.40625 -3.765625 4.625 -4.234375 4.625 -4.28125 C 4.625 -4.375 4.515625 -4.375 4.515625 -4.375 C 4.453125 -4.375 4.421875 -4.359375 4.375 -4.265625 C 4.078125 -3.796875 3.859375 -3.640625 3.625 -3.640625 C 3.390625 -3.640625 3.265625 -3.78125 3.125 -3.953125 C 2.9375 -4.1875 2.765625 -4.375 2.4375 -4.375 C 1.6875 -4.375 1.234375 -3.453125 1.234375 -3.25 C 1.234375 -3.203125 1.265625 -3.140625 1.359375 -3.140625 C 1.453125 -3.140625 1.46875 -3.1875 1.484375 -3.25 C 1.671875 -3.703125 2.25 -3.71875 2.328125 -3.71875 C 2.53125 -3.71875 2.71875 -3.640625 2.953125 -3.5625 C 3.34375 -3.421875 3.453125 -3.421875 3.71875 -3.421875 C 3.359375 -2.984375 2.53125 -2.28125 2.34375 -2.125 L 1.453125 -1.28125 C 0.765625 -0.625 0.421875 -0.0625 0.421875 0.015625 C 0.421875 0.109375 0.546875 0.109375 0.546875 0.109375 C 0.625 0.109375 0.640625 0.09375 0.703125 -0.015625 C 0.9375 -0.359375 1.234375 -0.640625 1.546875 -0.640625 C 1.765625 -0.640625 1.875 -0.546875 2.125 -0.25 C 2.28125 -0.046875 2.46875 0.109375 2.75 0.109375 C 3.734375 0.109375 4.3125 -1.15625 4.3125 -1.421875 Z M 4.3125 -1.421875 "/>
<path style="stroke:none;" d="M 5.390625 -1.421875 C 5.390625 -1.515625 5.3125 -1.515625 5.28125 -1.515625 C 5.171875 -1.515625 5.171875 -1.484375 5.140625 -1.34375 C 5 -0.78125 4.8125 -0.109375 4.390625 -0.109375 C 4.1875 -0.109375 4.078125 -0.234375 4.078125 -0.5625 C 4.078125 -0.78125 4.203125 -1.25 4.28125 -1.59375 L 4.5625 -2.671875 C 4.59375 -2.828125 4.6875 -3.203125 4.734375 -3.34375 C 4.78125 -3.578125 4.875 -3.953125 4.875 -4.015625 C 4.875 -4.1875 4.734375 -4.28125 4.59375 -4.28125 C 4.546875 -4.28125 4.28125 -4.265625 4.203125 -3.9375 L 3.453125 -0.9375 C 3.453125 -0.90625 3.046875 -0.109375 2.328125 -0.109375 C 1.8125 -0.109375 1.703125 -0.5625 1.703125 -0.921875 C 1.703125 -1.484375 1.984375 -2.265625 2.25 -2.953125 C 2.359375 -3.25 2.421875 -3.390625 2.421875 -3.578125 C 2.421875 -4.03125 2.09375 -4.390625 1.59375 -4.390625 C 0.65625 -4.390625 0.28125 -2.953125 0.28125 -2.859375 C 0.28125 -2.765625 0.40625 -2.765625 0.40625 -2.765625 C 0.5 -2.765625 0.515625 -2.78125 0.5625 -2.9375 C 0.8125 -3.8125 1.1875 -4.171875 1.5625 -4.171875 C 1.65625 -4.171875 1.8125 -4.15625 1.8125 -3.84375 C 1.8125 -3.609375 1.703125 -3.3125 1.65625 -3.171875 C 1.28125 -2.1875 1.078125 -1.5625 1.078125 -1.078125 C 1.078125 -0.140625 1.765625 0.109375 2.296875 0.109375 C 2.953125 0.109375 3.3125 -0.34375 3.484375 -0.5625 C 3.59375 -0.15625 3.9375 0.109375 4.359375 0.109375 C 4.703125 0.109375 4.9375 -0.125 5.09375 -0.4375 C 5.265625 -0.796875 5.390625 -1.421875 5.390625 -1.421875 Z M 5.390625 -1.421875 "/>
</symbol>
<symbol overflow="visible" id="glyph0-6">
<path style="stroke:none;" d="M 4.328125 -1.421875 C 4.328125 -1.46875 4.28125 -1.515625 4.203125 -1.515625 C 4.109375 -1.515625 4.09375 -1.453125 4.0625 -1.390625 C 3.828125 -0.75 3.203125 -0.5625 2.875 -0.5625 C 2.671875 -0.5625 2.5 -0.609375 2.28125 -0.6875 C 1.953125 -0.8125 1.796875 -0.859375 1.59375 -0.859375 C 1.59375 -0.859375 1.40625 -0.859375 1.328125 -0.828125 C 1.859375 -1.40625 2.140625 -1.65625 2.5 -1.953125 C 2.5 -1.953125 3.125 -2.5 3.484375 -2.859375 C 4.421875 -3.78125 4.640625 -4.25 4.640625 -4.296875 C 4.640625 -4.390625 4.53125 -4.390625 4.53125 -4.390625 C 4.46875 -4.390625 4.4375 -4.375 4.375 -4.28125 C 4.078125 -3.8125 3.875 -3.640625 3.640625 -3.640625 C 3.40625 -3.640625 3.28125 -3.796875 3.125 -3.96875 C 2.9375 -4.1875 2.765625 -4.390625 2.4375 -4.390625 C 1.703125 -4.390625 1.25 -3.46875 1.25 -3.265625 C 1.25 -3.203125 1.265625 -3.15625 1.359375 -3.15625 C 1.453125 -3.15625 1.46875 -3.203125 1.484375 -3.265625 C 1.671875 -3.71875 2.25 -3.734375 2.328125 -3.734375 C 2.546875 -3.734375 2.734375 -3.65625 2.96875 -3.578125 C 3.359375 -3.421875 3.46875 -3.421875 3.734375 -3.421875 C 3.375 -3 2.53125 -2.28125 2.34375 -2.125 L 1.453125 -1.296875 C 0.78125 -0.625 0.421875 -0.0625 0.421875 0.015625 C 0.421875 0.109375 0.546875 0.109375 0.546875 0.109375 C 0.625 0.109375 0.640625 0.09375 0.703125 -0.015625 C 0.9375 -0.375 1.234375 -0.640625 1.546875 -0.640625 C 1.78125 -0.640625 1.875 -0.546875 2.125 -0.265625 C 2.296875 -0.046875 2.46875 0.109375 2.765625 0.109375 C 3.75 0.109375 4.328125 -1.15625 4.328125 -1.421875 Z M 4.328125 -1.421875 "/>
</symbol>
<symbol overflow="visible" id="glyph1-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph1-1">
<path style="stroke:none;" d="M 3.5 -1.265625 L 3.265625 -1.265625 C 3.25 -1.109375 3.171875 -0.703125 3.09375 -0.625 C 3.03125 -0.59375 2.5 -0.59375 2.40625 -0.59375 L 1.125 -0.59375 C 1.859375 -1.234375 2.09375 -1.421875 2.515625 -1.75 C 3.03125 -2.171875 3.5 -2.59375 3.5 -3.25 C 3.5 -4.09375 2.765625 -4.609375 1.875 -4.609375 C 1.015625 -4.609375 0.4375 -4 0.4375 -3.359375 C 0.4375 -3.015625 0.734375 -2.96875 0.8125 -2.96875 C 0.96875 -2.96875 1.171875 -3.09375 1.171875 -3.34375 C 1.171875 -3.46875 1.125 -3.71875 0.765625 -3.71875 C 0.984375 -4.203125 1.453125 -4.359375 1.78125 -4.359375 C 2.46875 -4.359375 2.828125 -3.8125 2.828125 -3.25 C 2.828125 -2.65625 2.40625 -2.171875 2.171875 -1.921875 L 0.5 -0.265625 C 0.4375 -0.203125 0.4375 -0.1875 0.4375 0 L 3.296875 0 Z M 3.5 -1.265625 "/>
<path style="stroke:none;" d="M 3.515625 -1.265625 L 3.28125 -1.265625 C 3.25 -1.109375 3.1875 -0.703125 3.09375 -0.625 C 3.03125 -0.59375 2.5 -0.59375 2.40625 -0.59375 L 1.125 -0.59375 C 1.859375 -1.234375 2.09375 -1.4375 2.515625 -1.765625 C 3.03125 -2.171875 3.515625 -2.59375 3.515625 -3.265625 C 3.515625 -4.109375 2.78125 -4.609375 1.890625 -4.609375 C 1.015625 -4.609375 0.4375 -4.015625 0.4375 -3.375 C 0.4375 -3.015625 0.734375 -2.984375 0.8125 -2.984375 C 0.96875 -2.984375 1.171875 -3.09375 1.171875 -3.34375 C 1.171875 -3.484375 1.125 -3.71875 0.765625 -3.71875 C 0.984375 -4.21875 1.453125 -4.359375 1.78125 -4.359375 C 2.46875 -4.359375 2.84375 -3.828125 2.84375 -3.265625 C 2.84375 -2.65625 2.40625 -2.171875 2.1875 -1.921875 L 0.5 -0.265625 C 0.4375 -0.203125 0.4375 -0.1875 0.4375 0 L 3.296875 0 Z M 3.515625 -1.265625 "/>
</symbol>
<symbol overflow="visible" id="glyph1-2">
<path style="stroke:none;" d="M 3.28125 0 L 3.28125 -0.25 L 3.015625 -0.25 C 2.328125 -0.25 2.328125 -0.34375 2.328125 -0.5625 L 2.328125 -4.40625 C 2.328125 -4.59375 2.3125 -4.609375 2.109375 -4.609375 C 1.671875 -4.171875 1.046875 -4.15625 0.75 -4.15625 L 0.75 -3.90625 C 0.921875 -3.90625 1.375 -3.90625 1.765625 -4.109375 L 1.765625 -0.5625 C 1.765625 -0.34375 1.765625 -0.25 1.0625 -0.25 L 0.8125 -0.25 L 0.8125 0 L 2.046875 -0.03125 Z M 3.28125 0 "/>
<path style="stroke:none;" d="M 3.28125 0 L 3.28125 -0.25 L 3.03125 -0.25 C 2.328125 -0.25 2.328125 -0.34375 2.328125 -0.5625 L 2.328125 -4.421875 C 2.328125 -4.609375 2.3125 -4.609375 2.125 -4.609375 C 1.671875 -4.171875 1.046875 -4.171875 0.765625 -4.171875 L 0.765625 -3.921875 C 0.921875 -3.921875 1.390625 -3.921875 1.765625 -4.109375 L 1.765625 -0.5625 C 1.765625 -0.34375 1.765625 -0.25 1.078125 -0.25 L 0.8125 -0.25 L 0.8125 0 L 2.046875 -0.03125 Z M 3.28125 0 "/>
</symbol>
<symbol overflow="visible" id="glyph2-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph2-1">
<path style="stroke:none;" d="M 8.0625 -2.984375 C 8.0625 -3.203125 7.875 -3.203125 7.6875 -3.203125 L 4.515625 -3.203125 L 4.515625 -6.375 C 4.515625 -6.546875 4.515625 -6.75 4.3125 -6.75 C 4.09375 -6.75 4.09375 -6.5625 4.09375 -6.375 L 4.09375 -3.203125 L 0.921875 -3.203125 C 0.75 -3.203125 0.546875 -3.203125 0.546875 -3 C 0.546875 -2.78125 0.75 -2.78125 0.921875 -2.78125 L 4.09375 -2.78125 L 4.09375 0.390625 C 4.09375 0.5625 4.09375 0.765625 4.296875 0.765625 C 4.515625 0.765625 4.515625 0.578125 4.515625 0.390625 L 4.515625 -2.78125 L 7.6875 -2.78125 C 7.859375 -2.78125 8.0625 -2.78125 8.0625 -2.984375 Z M 8.0625 -2.984375 "/>
<path style="stroke:none;" d="M 3.296875 2.390625 C 3.296875 2.359375 3.296875 2.328125 3.125 2.171875 C 1.875 0.921875 1.5625 -0.96875 1.5625 -2.484375 C 1.5625 -4.21875 1.9375 -5.9375 3.15625 -7.1875 C 3.296875 -7.296875 3.296875 -7.328125 3.296875 -7.359375 C 3.296875 -7.421875 3.25 -7.453125 3.1875 -7.453125 C 3.09375 -7.453125 2.203125 -6.78125 1.609375 -5.515625 C 1.109375 -4.421875 0.984375 -3.3125 0.984375 -2.484375 C 0.984375 -1.703125 1.09375 -0.5 1.640625 0.609375 C 2.234375 1.84375 3.09375 2.484375 3.1875 2.484375 C 3.25 2.484375 3.296875 2.453125 3.296875 2.390625 Z M 3.296875 2.390625 "/>
</symbol>
<symbol overflow="visible" id="glyph2-2">
<path style="stroke:none;" d="M 2.875 -2.484375 C 2.875 -3.265625 2.765625 -4.46875 2.21875 -5.578125 C 1.625 -6.8125 0.765625 -7.453125 0.671875 -7.453125 C 0.609375 -7.453125 0.5625 -7.40625 0.5625 -7.359375 C 0.5625 -7.328125 0.5625 -7.296875 0.75 -7.125 C 1.734375 -6.140625 2.296875 -4.5625 2.296875 -2.484375 C 2.296875 -0.78125 1.921875 0.96875 0.703125 2.21875 C 0.5625 2.328125 0.5625 2.359375 0.5625 2.390625 C 0.5625 2.4375 0.609375 2.484375 0.671875 2.484375 C 0.765625 2.484375 1.65625 1.8125 2.25 0.546875 C 2.75 -0.546875 2.875 -1.65625 2.875 -2.484375 Z M 2.875 -2.484375 "/>
</symbol>
<symbol overflow="visible" id="glyph2-3">
<path style="stroke:none;" d="M 4.140625 -6.3125 L 3.984375 -6.46875 C 3.984375 -6.46875 3.609375 -6 3.171875 -6 C 2.9375 -6 2.6875 -6.140625 2.53125 -6.234375 C 2.25 -6.40625 2.078125 -6.46875 1.90625 -6.46875 C 1.53125 -6.46875 1.34375 -6.25 0.828125 -5.6875 L 1 -5.546875 C 1 -5.546875 1.375 -6.015625 1.8125 -6.015625 C 2.03125 -6.015625 2.28125 -5.875 2.453125 -5.78125 C 2.71875 -5.609375 2.90625 -5.546875 3.078125 -5.546875 C 3.453125 -5.546875 3.640625 -5.765625 4.140625 -6.3125 Z M 4.140625 -6.3125 "/>
</symbol>
<symbol overflow="visible" id="glyph3-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph3-1">
<path style="stroke:none;" d="M 6.875 -2.484375 C 6.875 -2.671875 6.6875 -2.671875 6.546875 -2.671875 L 1.15625 -2.671875 C 1.015625 -2.671875 0.828125 -2.671875 0.828125 -2.484375 C 0.828125 -2.28125 1.015625 -2.28125 1.15625 -2.28125 L 6.546875 -2.28125 C 6.6875 -2.28125 6.875 -2.28125 6.875 -2.484375 Z M 6.875 -2.484375 "/>
<path style="stroke:none;" d="M 8.09375 -2.984375 C 8.09375 -3.203125 7.890625 -3.203125 7.71875 -3.203125 L 4.53125 -3.203125 L 4.53125 -6.390625 C 4.53125 -6.5625 4.53125 -6.765625 4.328125 -6.765625 C 4.109375 -6.765625 4.109375 -6.578125 4.109375 -6.390625 L 4.109375 -3.203125 L 0.921875 -3.203125 C 0.75 -3.203125 0.546875 -3.203125 0.546875 -3 C 0.546875 -2.78125 0.75 -2.78125 0.921875 -2.78125 L 4.109375 -2.78125 L 4.109375 0.390625 C 4.109375 0.5625 4.109375 0.765625 4.3125 0.765625 C 4.53125 0.765625 4.53125 0.578125 4.53125 0.390625 L 4.53125 -2.78125 L 7.71875 -2.78125 C 7.875 -2.78125 8.09375 -2.78125 8.09375 -2.984375 Z M 8.09375 -2.984375 "/>
</symbol>
<symbol overflow="visible" id="glyph4-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph4-1">
<path style="stroke:none;" d="M 4.125 -6.296875 L 3.96875 -6.453125 C 3.96875 -6.453125 3.59375 -5.96875 3.15625 -5.96875 C 2.9375 -5.96875 2.6875 -6.125 2.515625 -6.21875 C 2.25 -6.375 2.078125 -6.453125 1.90625 -6.453125 C 1.53125 -6.453125 1.34375 -6.234375 0.828125 -5.671875 L 0.984375 -5.53125 C 0.984375 -5.53125 1.359375 -6 1.796875 -6 C 2.03125 -6 2.28125 -5.859375 2.453125 -5.75 C 2.71875 -5.59375 2.890625 -5.53125 3.0625 -5.53125 C 3.4375 -5.53125 3.625 -5.75 4.125 -6.296875 Z M 4.125 -6.296875 "/>
<path style="stroke:none;" d="M 6.890625 -2.484375 C 6.890625 -2.6875 6.703125 -2.6875 6.5625 -2.6875 L 1.15625 -2.6875 C 1.015625 -2.6875 0.828125 -2.6875 0.828125 -2.484375 C 0.828125 -2.28125 1.015625 -2.28125 1.15625 -2.28125 L 6.5625 -2.28125 C 6.703125 -2.28125 6.890625 -2.28125 6.890625 -2.484375 Z M 6.890625 -2.484375 "/>
</symbol>
</g>
<clipPath id="clip1">
<path d="M 0.328125 81 L 26 81 L 26 83 L 0.328125 83 Z M 0.328125 81 "/>
<path d="M 0.285156 76 L 20 76 L 20 78 L 0.285156 78 Z M 0.285156 76 "/>
</clipPath>
<clipPath id="clip2">
<path d="M 213 66 L 234.09375 66 L 234.09375 98 L 213 98 Z M 213 66 "/>
<path d="M 218 61 L 239.316406 61 L 239.316406 93 L 218 93 Z M 218 61 "/>
</clipPath>
<clipPath id="clip3">
<path d="M 87 81 L 209 81 L 209 105.691406 L 87 105.691406 Z M 87 81 "/>
</clipPath>
<clipPath id="clip4">
<path d="M 72 76 L 104 76 L 104 105.691406 L 72 105.691406 Z M 72 76 "/>
<path d="M 71 71 L 103 71 L 103 100.351562 L 71 100.351562 Z M 71 71 "/>
</clipPath>
</defs>
<g id="surface1">
<path style="fill-rule:nonzero;fill:rgb(100%,100%,100%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -11.33719 -11.337023 L 11.336978 -11.337023 L 11.336978 11.337144 L -11.33719 11.337144 Z M -11.33719 -11.337023 " transform="matrix(0.993869,0,0,-0.993869,41.255965,82.162169)"/>
<path style="fill-rule:nonzero;fill:rgb(100%,100%,100%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -16.468537 -11.338946 L 16.469483 -11.338946 L 16.469483 11.339497 L -16.468537 11.339497 Z M -16.468537 -11.338946 " transform="matrix(0.996782,0,0,-0.996782,40.798357,76.752228)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-1" x="34.359507" y="84.801886"/>
<use xlink:href="#glyph0-1" x="27.691668" y="79.234215"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-1" x="43.710822" y="86.28772"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(100%,100%,100%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 54.130588 -0.00190468 C 54.130588 3.849835 51.009893 6.97446 47.158154 6.97446 C 43.306415 6.97446 40.181789 3.849835 40.181789 -0.00190468 C 40.181789 -3.849714 43.306415 -6.974339 47.158154 -6.974339 C 51.009893 -6.974339 54.130588 -3.849714 54.130588 -0.00190468 Z M 54.130588 -0.00190468 " transform="matrix(0.993869,0,0,-0.993869,41.255965,82.162169)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-1" x="83.813442" y="85.060292"/>
<use xlink:href="#glyph1-1" x="37.070391" y="80.724405"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph3-1" x="78.442574" y="96.849568"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(100%,100%,100%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 153.344321 -0.00190468 C 153.344321 3.849835 150.223626 6.97446 146.371886 6.97446 C 142.520147 6.97446 139.399452 3.849835 139.399452 -0.00190468 C 139.399452 -3.849714 142.520147 -6.974339 146.371886 -6.974339 C 150.223626 -6.974339 153.344321 -3.849714 153.344321 -0.00190468 Z M 153.344321 -0.00190468 " transform="matrix(0.993869,0,0,-0.993869,41.255965,82.162169)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-1" x="182.417189" y="85.060292"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(100%,100%,100%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 135.032837 30.147782 L 157.710936 30.147782 L 157.710936 52.82588 L 135.032837 52.82588 Z M 135.032837 30.147782 " transform="matrix(0.993869,0,0,-0.993869,41.255965,82.162169)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-1" x="179.832136" y="43.569236"/>
<use xlink:href="#glyph2-1" x="41.525011" y="79.234215"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-2" x="189.183451" y="45.054077"/>
<use xlink:href="#glyph0-2" x="45.387542" y="79.234215"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-2" x="50.042515" y="79.234215"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(100%,100%,100%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 53.592846 -0.00168395 C 53.592846 3.850556 50.469514 6.973887 46.617274 6.973887 C 42.768954 6.973887 39.645622 3.850556 39.645622 -0.00168395 C 39.645622 -3.850005 42.768954 -6.973336 46.617274 -6.973336 C 50.469514 -6.973336 53.592846 -3.850005 53.592846 -0.00168395 Z M 53.592846 -0.00168395 " transform="matrix(0.996782,0,0,-0.996782,40.798357,76.752228)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph3-1" x="82.943304" y="79.658845"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph4-1" x="77.556693" y="91.482675"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(100%,100%,100%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 158.47722 -0.00168395 C 158.47722 3.850556 155.353888 6.973887 151.501649 6.973887 C 147.649409 6.973887 144.529996 3.850556 144.529996 -0.00168395 C 144.529996 -3.850005 147.649409 -6.973336 151.501649 -6.973336 C 155.353888 -6.973336 158.47722 -3.850005 158.47722 -0.00168395 Z M 158.47722 -0.00168395 " transform="matrix(0.996782,0,0,-0.996782,40.798357,76.752228)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph3-1" x="187.486816" y="79.658845"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(100%,100%,100%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 135.034598 30.150026 L 167.972618 30.150026 L 167.972618 52.82847 L 135.034598 52.82847 Z M 135.034598 30.150026 " transform="matrix(0.996782,0,0,-0.996782,40.798357,76.752228)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-1" x="178.704168" y="37.880713"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-2" x="188.082891" y="39.369906"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-1" x="192.538508" y="37.880713"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-2" x="196.400042" y="37.880713"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-2" x="201.055015" y="37.880713"/>
</g>
<g clip-path="url(#clip1)" clip-rule="nonzero">
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -16.470223 -0.00190468 L -40.182002 -0.00190468 " transform="matrix(0.993869,0,0,-0.993869,41.255965,82.162169)"/>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -21.602244 -0.00168395 L -39.644676 -0.00168395 " transform="matrix(0.996782,0,0,-0.996782,40.798357,76.752228)"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.053781 -0.00190468 L 1.608559 1.684214 L 3.086369 -0.00190468 L 1.608559 -1.684093 Z M 6.053781 -0.00190468 " transform="matrix(0.993869,0,0,-0.993869,22.065365,82.162169)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.05474 -0.00168395 L 1.606834 1.683426 L 3.088163 -0.00168395 L 1.606834 -1.682875 Z M 6.05474 -0.00168395 " transform="matrix(0.996782,0,0,-0.996782,16.437399,76.752228)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-2" x="5.114907" y="76.881743"/>
<use xlink:href="#glyph0-3" x="5.087638" y="71.456324"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-1" x="12.203182" y="78.366583"/>
<use xlink:href="#glyph1-1" x="12.196689" y="72.945517"/>
</g>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 11.836132 -0.00190468 L 35.551841 -0.00190468 " transform="matrix(0.993869,0,0,-0.993869,41.255965,82.162169)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.055245 -0.00190468 L 1.610023 1.684214 L 3.087833 -0.00190468 L 1.610023 -1.684093 Z M 6.055245 -0.00190468 " transform="matrix(0.993869,0,0,-0.993869,73.767035,82.162169)"/>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 16.967178 -0.00168395 L 35.009611 -0.00168395 " transform="matrix(0.996782,0,0,-0.996782,40.798357,76.752228)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.054025 -0.00168395 L 1.610037 1.683426 L 3.087448 -0.00168395 L 1.610037 -1.682875 Z M 6.054025 -0.00168395 " transform="matrix(0.996782,0,0,-0.996782,72.8678,76.752228)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph4-1" x="67.89464" y="76.881743"/>
<use xlink:href="#glyph2-3" x="66.977844" y="71.456324"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-2" x="65.867147" y="76.881743"/>
<use xlink:href="#glyph0-3" x="64.944408" y="71.456324"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-1" x="72.956416" y="78.366583"/>
<use xlink:href="#glyph1-1" x="72.053459" y="72.945517"/>
</g>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 54.130588 -0.00190468 L 77.846298 -0.00190468 " transform="matrix(0.993869,0,0,-0.993869,41.255965,82.162169)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.055081 -0.00190468 L 1.60986 1.684214 L 3.08767 -0.00190468 L 1.60986 -1.684093 Z M 6.055081 -0.00190468 " transform="matrix(0.993869,0,0,-0.993869,115.802354,82.162169)"/>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 53.592846 -0.00168395 L 77.305869 -0.00168395 " transform="matrix(0.996782,0,0,-0.996782,40.798357,76.752228)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.051744 -0.00168395 L 1.607756 1.683426 L 3.089086 -0.00168395 L 1.607756 -1.682875 Z M 6.051744 -0.00168395 " transform="matrix(0.996782,0,0,-0.996782,115.026323,76.752228)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-3" x="114.27752" y="78.366583"/>
<use xlink:href="#glyph0-4" x="113.496671" y="72.945517"/>
</g>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 134.765574 -0.00190468 L 111.049864 -0.00190468 " transform="matrix(0.993869,0,0,-0.993869,41.255965,82.162169)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.055147 -0.00190468 L 1.609925 1.684214 L 3.087736 -0.00190468 L 1.609925 -1.684093 Z M 6.055147 -0.00190468 " transform="matrix(0.993869,0,0,-0.993869,172.372601,82.162169)"/>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 139.893985 -0.00168395 L 116.180962 -0.00168395 " transform="matrix(0.996782,0,0,-0.996782,40.798357,76.752228)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.055279 -0.00168395 L 1.607373 1.683426 L 3.088702 -0.00168395 L 1.607373 -1.682875 Z M 6.055279 -0.00168395 " transform="matrix(0.996782,0,0,-0.996782,177.413425,76.752228)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-4" x="155.419728" y="78.366583"/>
<use xlink:href="#glyph0-5" x="160.410225" y="72.945517"/>
</g>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 153.344321 -0.00190468 L 188.395149 -0.00190468 " transform="matrix(0.993869,0,0,-0.993869,41.255965,82.162169)"/>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 231.691406 82.164062 L 227.277344 80.488281 L 228.746094 82.164062 L 227.277344 83.835938 Z M 231.691406 82.164062 "/>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 158.47722 -0.00168395 L 193.527505 -0.00168395 " transform="matrix(0.996782,0,0,-0.996782,40.798357,76.752228)"/>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 236.90625 76.753906 L 232.476562 75.074219 L 233.953125 76.753906 L 232.476562 78.429688 Z M 236.90625 76.753906 "/>
<g clip-path="url(#clip2)" clip-rule="nonzero">
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.051563 -0.00190468 L 1.610271 1.684214 L 3.088082 -0.00190468 L 1.610271 -1.684093 Z M 6.051563 -0.00190468 " transform="matrix(0.993869,0,0,-0.993869,225.676945,82.162169)"/>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.051731 -0.00168395 L 1.607743 1.683426 L 3.089072 -0.00168395 L 1.607743 -1.682875 Z M 6.051731 -0.00168395 " transform="matrix(0.996782,0,0,-0.996782,230.873993,76.752228)"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-5" x="224.266038" y="78.366583"/>
<use xlink:href="#glyph0-6" x="229.458323" y="72.945517"/>
</g>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 146.371886 29.652558 L 146.371886 11.608338 " transform="matrix(0.993869,0,0,-0.993869,41.255965,82.162169)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.052665 0.000476446 L 1.607443 1.682665 L 3.089184 0.000476446 L 1.607443 -1.681712 Z M 6.052665 0.000476446 " transform="matrix(0,0.993869,0.993869,0,186.729995,67.804756)"/>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 151.501649 29.652331 L 151.501649 11.60598 " transform="matrix(0.996782,0,0,-0.996782,40.798357,76.752228)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.053465 -0.000411378 L 1.609478 1.684699 L 3.086888 -0.000411378 L 1.609478 -1.681602 Z M 6.053465 -0.000411378 " transform="matrix(0,0.996782,0.996782,0,191.81291,62.352732)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph4-1" x="192.550679" y="69.950499"/>
<use xlink:href="#glyph2-3" x="197.650007" y="64.504765"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-2" x="190.523186" y="69.950499"/>
<use xlink:href="#glyph0-3" x="195.616571" y="64.504765"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-2" x="197.612455" y="71.43534"/>
<use xlink:href="#glyph1-2" x="202.726618" y="65.993958"/>
</g>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 146.371886 57.958913 L 146.371886 81.670692 " transform="matrix(0.993869,0,0,-0.993869,41.255965,82.162169)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.053411 0.000476446 L 1.608189 1.682665 L 3.085999 0.000476446 L 1.608189 -1.681712 Z M 6.053411 0.000476446 " transform="matrix(0,0.993869,0.993869,0,186.729995,21.737608)"/>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 151.501649 57.958258 L 151.501649 76.00461 " transform="matrix(0.996782,0,0,-0.996782,40.798357,76.752228)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.052507 -0.000411378 L 1.60852 1.684699 L 3.08593 -0.000411378 L 1.60852 -1.681602 Z M 6.052507 -0.000411378 " transform="matrix(0,0.996782,0.996782,0,191.81291,16.150563)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-2" x="190.523186" y="9.050173"/>
<use xlink:href="#glyph0-3" x="195.616571" y="9.076699"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-2" x="197.612455" y="10.535013"/>
<use xlink:href="#glyph1-2" x="202.726618" y="10.565891"/>
</g>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 172.647818 -0.00168395 L 172.647818 -22.676209 L 46.617274 -22.676209 L 46.617274 -11.609348 " transform="matrix(0.996782,0,0,-0.996782,40.798357,76.752228)"/>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 87.265625 85.117188 L 85.589844 89.546875 L 87.265625 88.074219 L 88.945312 89.546875 Z M 87.265625 85.117188 "/>
<g clip-path="url(#clip3)" clip-rule="nonzero">
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 167.51715 -0.00190468 L 167.51715 -22.676073 L 47.158154 -22.676073 L 47.158154 -11.608217 " transform="matrix(0.993869,0,0,-0.993869,41.255965,82.162169)"/>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.054006 0.0016656 L 1.610019 1.682857 L 3.087429 0.0016656 L 1.610019 -1.683444 Z M 6.054006 0.0016656 " transform="matrix(0,-0.996782,-0.996782,0,87.267285,91.151713)"/>
</g>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 88.125 90.503906 L 86.453125 94.921875 L 88.125 93.449219 L 89.796875 94.921875 Z M 88.125 90.503906 "/>
<g clip-path="url(#clip4)" clip-rule="nonzero">
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.052776 -0.000554066 L 1.607554 1.681634 L 3.089295 -0.000554066 L 1.607554 -1.682742 Z M 6.052776 -0.000554066 " transform="matrix(0,-0.993869,-0.993869,0,88.124449,96.519574)"/>
</g>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 209.726562 82.164062 C 209.726562 81.070312 208.839844 80.183594 207.746094 80.183594 C 206.652344 80.183594 205.765625 81.070312 205.765625 82.164062 C 205.765625 83.257812 206.652344 84.140625 207.746094 84.140625 C 208.839844 84.140625 209.726562 83.257812 209.726562 82.164062 Z M 209.726562 82.164062 "/>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 214.878906 76.753906 C 214.878906 75.65625 213.988281 74.765625 212.890625 74.765625 C 211.796875 74.765625 210.90625 75.65625 210.90625 76.753906 C 210.90625 77.847656 211.796875 78.738281 212.890625 78.738281 C 213.988281 78.738281 214.878906 77.847656 214.878906 76.753906 Z M 214.878906 76.753906 "/>
</g>
</svg>

Side by Side Swipe Overlay

Before

Width:  |  Height:  |  Size: 22 KiB

After

Width:  |  Height:  |  Size: 25 KiB

BIN
journal/figs/h_infinity_robust_fusion.png
View File

Binary file not shown.
Side by Side

Before

Width:  |  Height:  |  Size: 9.7 KiB

180
journal/figs/h_infinity_robust_fusion.svg
View File

@ -1,180 +0,0 @@
<?xml version="1.0" encoding="UTF-8"?>
<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="156.063pt" height="104.137pt" viewBox="0 0 156.063 104.137" version="1.2">
<defs>
<g>
<symbol overflow="visible" id="glyph0-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph0-1">
<path style="stroke:none;" d="M 7.453125 -5.25 C 7.453125 -6.015625 6.796875 -6.75 5.5 -6.75 L 2.296875 -6.75 C 2.109375 -6.75 2 -6.75 2 -6.5625 C 2 -6.4375 2.09375 -6.4375 2.296875 -6.4375 C 2.421875 -6.4375 2.59375 -6.421875 2.71875 -6.421875 C 2.875 -6.40625 2.9375 -6.375 2.9375 -6.265625 C 2.9375 -6.21875 2.921875 -6.1875 2.890625 -6.078125 L 1.5625 -0.765625 C 1.46875 -0.390625 1.453125 -0.3125 0.671875 -0.3125 C 0.5 -0.3125 0.390625 -0.3125 0.390625 -0.125 C 0.390625 0 0.515625 0 0.546875 0 L 1.796875 -0.03125 L 2.421875 -0.015625 C 2.640625 -0.015625 2.859375 0 3.078125 0 C 3.140625 0 3.265625 0 3.265625 -0.203125 C 3.265625 -0.3125 3.1875 -0.3125 2.984375 -0.3125 C 2.625 -0.3125 2.34375 -0.3125 2.34375 -0.484375 C 2.34375 -0.546875 2.375 -0.59375 2.375 -0.65625 L 2.984375 -3.125 L 4.671875 -3.125 C 6.078125 -3.125 7.453125 -4.140625 7.453125 -5.25 Z M 6.53125 -5.484375 C 6.53125 -5.109375 6.328125 -4.25 5.9375 -3.90625 C 5.453125 -3.453125 4.859375 -3.375 4.421875 -3.375 L 3.03125 -3.375 L 3.703125 -6.0625 C 3.796875 -6.40625 3.8125 -6.4375 4.234375 -6.4375 L 5.1875 -6.4375 C 6 -6.4375 6.53125 -6.171875 6.53125 -5.484375 Z M 6.53125 -5.484375 "/>
</symbol>
<symbol overflow="visible" id="glyph0-2">
<path style="stroke:none;" d="M 3.90625 -1.53125 C 3.90625 -1.890625 3.71875 -2.140625 3.609375 -2.25 C 3.34375 -2.53125 3.046875 -2.578125 2.609375 -2.671875 C 2.265625 -2.75 1.859375 -2.8125 1.859375 -3.265625 C 1.859375 -3.546875 2.109375 -4.140625 2.96875 -4.140625 C 3.21875 -4.140625 3.71875 -4.078125 3.859375 -3.6875 C 3.578125 -3.6875 3.390625 -3.46875 3.390625 -3.25 C 3.390625 -3.109375 3.46875 -2.96875 3.6875 -2.96875 C 3.90625 -2.96875 4.140625 -3.125 4.140625 -3.515625 C 4.140625 -3.953125 3.71875 -4.359375 2.96875 -4.359375 C 1.671875 -4.359375 1.296875 -3.359375 1.296875 -2.921875 C 1.296875 -2.15625 2.03125 -2 2.328125 -1.953125 C 2.828125 -1.84375 3.34375 -1.734375 3.34375 -1.1875 C 3.34375 -0.9375 3.125 -0.109375 1.9375 -0.109375 C 1.796875 -0.109375 1.03125 -0.109375 0.8125 -0.625 C 1.1875 -0.578125 1.4375 -0.875 1.4375 -1.15625 C 1.4375 -1.375 1.28125 -1.5 1.0625 -1.5 C 0.8125 -1.5 0.515625 -1.296875 0.515625 -0.84375 C 0.515625 -0.28125 1.078125 0.109375 1.921875 0.109375 C 3.53125 0.109375 3.90625 -1.09375 3.90625 -1.53125 Z M 3.90625 -1.53125 "/>
</symbol>
<symbol overflow="visible" id="glyph0-3">
<path style="stroke:none;" d="M 10.34375 -6.640625 C 10.34375 -6.671875 10.3125 -6.75 10.234375 -6.75 C 10 -6.75 9.71875 -6.71875 9.46875 -6.71875 C 9.140625 -6.71875 8.765625 -6.75 8.4375 -6.75 C 8.390625 -6.75 8.25 -6.75 8.25 -6.5625 C 8.25 -6.453125 8.34375 -6.4375 8.40625 -6.4375 C 8.640625 -6.421875 8.984375 -6.34375 8.984375 -6.046875 C 8.984375 -5.9375 8.9375 -5.859375 8.859375 -5.734375 L 6.1875 -1.078125 L 5.828125 -5.96875 C 5.8125 -6.171875 5.796875 -6.421875 6.515625 -6.4375 C 6.6875 -6.4375 6.78125 -6.4375 6.78125 -6.640625 C 6.78125 -6.734375 6.671875 -6.75 6.640625 -6.75 C 6.234375 -6.75 5.828125 -6.71875 5.4375 -6.71875 C 5.203125 -6.71875 4.625 -6.75 4.390625 -6.75 C 4.328125 -6.75 4.203125 -6.75 4.203125 -6.546875 C 4.203125 -6.4375 4.3125 -6.4375 4.4375 -6.4375 C 4.875 -6.4375 4.953125 -6.375 4.96875 -6.1875 L 5.03125 -5.4375 L 2.53125 -1.078125 L 2.15625 -6.078125 C 2.15625 -6.203125 2.15625 -6.421875 2.90625 -6.4375 C 3 -6.4375 3.109375 -6.4375 3.109375 -6.640625 C 3.109375 -6.75 2.96875 -6.75 2.96875 -6.75 C 2.578125 -6.75 2.15625 -6.71875 1.75 -6.71875 C 1.40625 -6.71875 1.0625 -6.75 0.71875 -6.75 C 0.671875 -6.75 0.546875 -6.75 0.546875 -6.5625 C 0.546875 -6.4375 0.625 -6.4375 0.796875 -6.4375 C 1.28125 -6.4375 1.296875 -6.34375 1.3125 -6.078125 L 1.75 -0.046875 C 1.765625 0.140625 1.78125 0.21875 1.921875 0.21875 C 2.03125 0.21875 2.0625 0.15625 2.15625 0.015625 L 5.0625 -5.03125 L 5.421875 -0.046875 C 5.4375 0.171875 5.453125 0.21875 5.578125 0.21875 C 5.703125 0.21875 5.765625 0.125 5.8125 0.03125 L 9.0625 -5.59375 C 9.296875 -6.03125 9.546875 -6.390625 10.15625 -6.4375 C 10.25 -6.453125 10.34375 -6.453125 10.34375 -6.640625 Z M 10.34375 -6.640625 "/>
</symbol>
<symbol overflow="visible" id="glyph0-4">
<path style="stroke:none;" d="M 6.828125 -3.671875 C 6.828125 -4.21875 6.5625 -4.359375 6.390625 -4.359375 C 6.140625 -4.359375 5.890625 -4.109375 5.890625 -3.890625 C 5.890625 -3.765625 5.953125 -3.703125 6.046875 -3.625 C 6.15625 -3.515625 6.40625 -3.265625 6.40625 -2.78125 C 6.40625 -2.453125 6.109375 -1.484375 5.859375 -0.984375 C 5.59375 -0.453125 5.25 -0.109375 4.765625 -0.109375 C 4.3125 -0.109375 4.03125 -0.40625 4.03125 -0.96875 C 4.03125 -1.25 4.109375 -1.546875 4.140625 -1.6875 L 4.5625 -3.34375 C 4.609375 -3.5625 4.703125 -3.9375 4.703125 -3.984375 C 4.703125 -4.171875 4.578125 -4.25 4.421875 -4.25 C 4.3125 -4.25 4.125 -4.171875 4.0625 -3.984375 C 4.03125 -3.90625 3.5625 -2.03125 3.5 -1.765625 C 3.421875 -1.46875 3.40625 -1.296875 3.40625 -1.109375 C 3.40625 -1 3.40625 -0.984375 3.421875 -0.9375 C 3.1875 -0.421875 2.890625 -0.109375 2.515625 -0.109375 C 1.71875 -0.109375 1.71875 -0.84375 1.71875 -1 C 1.71875 -1.328125 1.765625 -1.703125 2.234375 -2.921875 C 2.34375 -3.21875 2.40625 -3.359375 2.40625 -3.5625 C 2.40625 -4 2.078125 -4.359375 1.59375 -4.359375 C 0.65625 -4.359375 0.28125 -2.9375 0.28125 -2.84375 C 0.28125 -2.75 0.40625 -2.75 0.40625 -2.75 C 0.5 -2.75 0.515625 -2.765625 0.5625 -2.921875 C 0.828125 -3.84375 1.21875 -4.140625 1.5625 -4.140625 C 1.65625 -4.140625 1.8125 -4.140625 1.8125 -3.828125 C 1.8125 -3.578125 1.703125 -3.296875 1.625 -3.125 C 1.1875 -1.96875 1.078125 -1.515625 1.078125 -1.140625 C 1.078125 -0.234375 1.734375 0.109375 2.484375 0.109375 C 2.640625 0.109375 3.109375 0.109375 3.5 -0.578125 C 3.765625 0.046875 4.4375 0.109375 4.734375 0.109375 C 5.484375 0.109375 5.921875 -0.515625 6.171875 -1.109375 C 6.5 -1.875 6.828125 -3.203125 6.828125 -3.671875 Z M 6.828125 -3.671875 "/>
</symbol>
<symbol overflow="visible" id="glyph0-5">
<path style="stroke:none;" d="M 5.359375 -1.40625 C 5.359375 -1.515625 5.265625 -1.515625 5.25 -1.515625 C 5.140625 -1.515625 5.140625 -1.46875 5.109375 -1.328125 C 4.96875 -0.78125 4.78125 -0.109375 4.359375 -0.109375 C 4.15625 -0.109375 4.0625 -0.234375 4.0625 -0.5625 C 4.0625 -0.78125 4.171875 -1.25 4.25 -1.59375 L 4.53125 -2.65625 C 4.5625 -2.796875 4.65625 -3.1875 4.703125 -3.328125 C 4.75 -3.5625 4.84375 -3.9375 4.84375 -3.984375 C 4.84375 -4.171875 4.703125 -4.25 4.5625 -4.25 C 4.515625 -4.25 4.25 -4.25 4.171875 -3.90625 L 3.4375 -0.9375 C 3.421875 -0.90625 3.03125 -0.109375 2.3125 -0.109375 C 1.796875 -0.109375 1.703125 -0.546875 1.703125 -0.921875 C 1.703125 -1.46875 1.96875 -2.25 2.234375 -2.9375 C 2.34375 -3.234375 2.40625 -3.375 2.40625 -3.5625 C 2.40625 -4 2.078125 -4.359375 1.59375 -4.359375 C 0.65625 -4.359375 0.28125 -2.9375 0.28125 -2.84375 C 0.28125 -2.75 0.40625 -2.75 0.40625 -2.75 C 0.5 -2.75 0.515625 -2.765625 0.5625 -2.921875 C 0.8125 -3.78125 1.1875 -4.140625 1.5625 -4.140625 C 1.65625 -4.140625 1.8125 -4.140625 1.8125 -3.828125 C 1.8125 -3.578125 1.703125 -3.296875 1.640625 -3.15625 C 1.28125 -2.171875 1.0625 -1.5625 1.0625 -1.078125 C 1.0625 -0.140625 1.75 0.109375 2.28125 0.109375 C 2.9375 0.109375 3.28125 -0.328125 3.453125 -0.546875 C 3.5625 -0.140625 3.90625 0.109375 4.328125 0.109375 C 4.6875 0.109375 4.90625 -0.125 5.0625 -0.4375 C 5.234375 -0.796875 5.359375 -1.40625 5.359375 -1.40625 Z M 5.359375 -1.40625 "/>
</symbol>
<symbol overflow="visible" id="glyph0-6">
<path style="stroke:none;" d="M 4.625 -3.671875 C 4.625 -4.203125 4.359375 -4.359375 4.1875 -4.359375 C 3.9375 -4.359375 3.703125 -4.109375 3.703125 -3.890625 C 3.703125 -3.765625 3.75 -3.703125 3.859375 -3.59375 C 4.0625 -3.390625 4.203125 -3.140625 4.203125 -2.78125 C 4.203125 -2.375 3.59375 -0.109375 2.4375 -0.109375 C 1.9375 -0.109375 1.703125 -0.453125 1.703125 -0.96875 C 1.703125 -1.515625 1.96875 -2.234375 2.28125 -3.0625 C 2.34375 -3.234375 2.40625 -3.375 2.40625 -3.5625 C 2.40625 -4 2.078125 -4.359375 1.59375 -4.359375 C 0.65625 -4.359375 0.28125 -2.9375 0.28125 -2.84375 C 0.28125 -2.75 0.40625 -2.75 0.40625 -2.75 C 0.5 -2.75 0.515625 -2.765625 0.5625 -2.921875 C 0.84375 -3.921875 1.28125 -4.140625 1.5625 -4.140625 C 1.640625 -4.140625 1.8125 -4.140625 1.8125 -3.828125 C 1.8125 -3.578125 1.703125 -3.3125 1.640625 -3.140625 C 1.203125 -2 1.078125 -1.546875 1.078125 -1.109375 C 1.078125 -0.046875 1.953125 0.109375 2.40625 0.109375 C 4.0625 0.109375 4.625 -3.15625 4.625 -3.671875 Z M 4.625 -3.671875 "/>
</symbol>
<symbol overflow="visible" id="glyph0-7">
<path style="stroke:none;" d="M 4.296875 -1.40625 C 4.296875 -1.46875 4.25 -1.515625 4.171875 -1.515625 C 4.09375 -1.515625 4.0625 -1.453125 4.03125 -1.375 C 3.8125 -0.734375 3.1875 -0.546875 2.859375 -0.546875 C 2.65625 -0.546875 2.484375 -0.609375 2.265625 -0.6875 C 1.9375 -0.8125 1.78125 -0.84375 1.578125 -0.84375 C 1.578125 -0.84375 1.40625 -0.84375 1.3125 -0.8125 C 1.84375 -1.390625 2.140625 -1.640625 2.484375 -1.953125 C 2.484375 -1.953125 3.09375 -2.484375 3.453125 -2.828125 C 4.390625 -3.75 4.609375 -4.21875 4.609375 -4.265625 C 4.609375 -4.359375 4.5 -4.359375 4.5 -4.359375 C 4.4375 -4.359375 4.40625 -4.34375 4.359375 -4.25 C 4.0625 -3.78125 3.84375 -3.625 3.609375 -3.625 C 3.375 -3.625 3.265625 -3.765625 3.109375 -3.9375 C 2.921875 -4.171875 2.75 -4.359375 2.421875 -4.359375 C 1.6875 -4.359375 1.234375 -3.453125 1.234375 -3.234375 C 1.234375 -3.1875 1.265625 -3.125 1.359375 -3.125 C 1.4375 -3.125 1.46875 -3.1875 1.484375 -3.234375 C 1.671875 -3.6875 2.234375 -3.703125 2.328125 -3.703125 C 2.53125 -3.703125 2.71875 -3.640625 2.9375 -3.5625 C 3.34375 -3.40625 3.453125 -3.40625 3.703125 -3.40625 C 3.34375 -2.984375 2.515625 -2.265625 2.328125 -2.109375 L 1.4375 -1.28125 C 0.765625 -0.625 0.421875 -0.0625 0.421875 0.015625 C 0.421875 0.109375 0.546875 0.109375 0.546875 0.109375 C 0.625 0.109375 0.640625 0.09375 0.703125 -0.015625 C 0.921875 -0.359375 1.21875 -0.625 1.546875 -0.625 C 1.765625 -0.625 1.859375 -0.546875 2.109375 -0.25 C 2.28125 -0.046875 2.453125 0.109375 2.75 0.109375 C 3.71875 0.109375 4.296875 -1.140625 4.296875 -1.40625 Z M 4.296875 -1.40625 "/>
</symbol>
<symbol overflow="visible" id="glyph1-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph1-1">
<path style="stroke:none;" d="M 3.265625 2.375 C 3.265625 2.34375 3.265625 2.328125 3.09375 2.15625 C 1.859375 0.90625 1.546875 -0.953125 1.546875 -2.46875 C 1.546875 -4.1875 1.921875 -5.90625 3.140625 -7.140625 C 3.265625 -7.265625 3.265625 -7.28125 3.265625 -7.3125 C 3.265625 -7.375 3.234375 -7.40625 3.171875 -7.40625 C 3.078125 -7.40625 2.1875 -6.734375 1.59375 -5.484375 C 1.09375 -4.390625 0.984375 -3.296875 0.984375 -2.46875 C 0.984375 -1.703125 1.09375 -0.5 1.625 0.609375 C 2.21875 1.828125 3.078125 2.46875 3.171875 2.46875 C 3.234375 2.46875 3.265625 2.4375 3.265625 2.375 Z M 3.265625 2.375 "/>
</symbol>
<symbol overflow="visible" id="glyph1-2">
<path style="stroke:none;" d="M 2.859375 -2.46875 C 2.859375 -3.234375 2.75 -4.4375 2.203125 -5.546875 C 1.609375 -6.765625 0.765625 -7.40625 0.65625 -7.40625 C 0.609375 -7.40625 0.5625 -7.359375 0.5625 -7.3125 C 0.5625 -7.28125 0.5625 -7.265625 0.75 -7.078125 C 1.71875 -6.109375 2.28125 -4.53125 2.28125 -2.46875 C 2.28125 -0.78125 1.921875 0.953125 0.6875 2.203125 C 0.5625 2.328125 0.5625 2.34375 0.5625 2.375 C 0.5625 2.421875 0.609375 2.46875 0.65625 2.46875 C 0.765625 2.46875 1.65625 1.796875 2.234375 0.546875 C 2.734375 -0.546875 2.859375 -1.640625 2.859375 -2.46875 Z M 2.859375 -2.46875 "/>
</symbol>
<symbol overflow="visible" id="glyph2-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph2-1">
<path style="stroke:none;" d="M 3.28125 0 L 3.28125 -0.25 L 3.015625 -0.25 C 2.3125 -0.25 2.3125 -0.34375 2.3125 -0.5625 L 2.3125 -4.40625 C 2.3125 -4.59375 2.3125 -4.59375 2.109375 -4.59375 C 1.671875 -4.15625 1.03125 -4.15625 0.75 -4.15625 L 0.75 -3.90625 C 0.921875 -3.90625 1.375 -3.90625 1.765625 -4.09375 L 1.765625 -0.5625 C 1.765625 -0.34375 1.765625 -0.25 1.0625 -0.25 L 0.796875 -0.25 L 0.796875 0 L 2.03125 -0.03125 Z M 3.28125 0 "/>
</symbol>
<symbol overflow="visible" id="glyph2-2">
<path style="stroke:none;" d="M 3.5 -1.265625 L 3.265625 -1.265625 C 3.234375 -1.109375 3.171875 -0.703125 3.078125 -0.625 C 3.03125 -0.59375 2.484375 -0.59375 2.390625 -0.59375 L 1.125 -0.59375 C 1.84375 -1.234375 2.09375 -1.421875 2.5 -1.75 C 3.015625 -2.15625 3.5 -2.59375 3.5 -3.25 C 3.5 -4.078125 2.765625 -4.59375 1.875 -4.59375 C 1.015625 -4.59375 0.4375 -4 0.4375 -3.359375 C 0.4375 -3 0.734375 -2.96875 0.796875 -2.96875 C 0.96875 -2.96875 1.171875 -3.09375 1.171875 -3.34375 C 1.171875 -3.453125 1.125 -3.703125 0.765625 -3.703125 C 0.96875 -4.1875 1.453125 -4.34375 1.765625 -4.34375 C 2.46875 -4.34375 2.828125 -3.8125 2.828125 -3.25 C 2.828125 -2.640625 2.390625 -2.171875 2.171875 -1.921875 L 0.5 -0.265625 C 0.4375 -0.203125 0.4375 -0.1875 0.4375 0 L 3.28125 0 Z M 3.5 -1.265625 "/>
</symbol>
<symbol overflow="visible" id="glyph3-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph3-1">
<path style="stroke:none;" d="M 8.046875 -2.96875 C 8.046875 -3.1875 7.859375 -3.1875 7.671875 -3.1875 L 4.5 -3.1875 L 4.5 -6.359375 C 4.5 -6.53125 4.5 -6.734375 4.296875 -6.734375 C 4.09375 -6.734375 4.09375 -6.53125 4.09375 -6.359375 L 4.09375 -3.1875 L 0.921875 -3.1875 C 0.75 -3.1875 0.546875 -3.1875 0.546875 -2.984375 C 0.546875 -2.765625 0.734375 -2.765625 0.921875 -2.765625 L 4.09375 -2.765625 L 4.09375 0.390625 C 4.09375 0.5625 4.09375 0.765625 4.296875 0.765625 C 4.5 0.765625 4.5 0.578125 4.5 0.390625 L 4.5 -2.765625 L 7.671875 -2.765625 C 7.84375 -2.765625 8.046875 -2.765625 8.046875 -2.96875 Z M 8.046875 -2.96875 "/>
</symbol>
<symbol overflow="visible" id="glyph4-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph4-1">
<path style="stroke:none;" d="M 6.859375 -2.46875 C 6.859375 -2.671875 6.671875 -2.671875 6.53125 -2.671875 L 1.140625 -2.671875 C 1 -2.671875 0.8125 -2.671875 0.8125 -2.46875 C 0.8125 -2.265625 1 -2.265625 1.140625 -2.265625 L 6.53125 -2.265625 C 6.671875 -2.265625 6.859375 -2.265625 6.859375 -2.46875 Z M 6.859375 -2.46875 "/>
</symbol>
</g>
<clipPath id="clip1">
<path d="M 0.175781 38 L 47 38 L 47 40 L 0.175781 40 Z M 0.175781 38 "/>
</clipPath>
<clipPath id="clip2">
<path d="M 0.175781 72 L 77 72 L 77 74 L 0.175781 74 Z M 0.175781 72 "/>
</clipPath>
<clipPath id="clip3">
<path d="M 134 78 L 154.957031 78 L 154.957031 103.28125 L 134 103.28125 Z M 134 78 "/>
</clipPath>
<clipPath id="clip4">
<path d="M 134 23 L 154.957031 23 L 154.957031 55 L 134 55 Z M 134 23 "/>
</clipPath>
<clipPath id="clip5">
<path d="M 134 57 L 154.957031 57 L 154.957031 89 L 134 89 Z M 134 57 "/>
</clipPath>
</defs>
<g id="surface1">
<path style="fill-rule:nonzero;fill:rgb(89.99939%,89.99939%,89.99939%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -56.693469 -42.521687 L 56.695512 -42.521687 L 56.695512 42.521033 L -56.693469 42.521033 Z M -56.693469 -42.521687 " transform="matrix(0.991781,0,0,-0.991781,77.567347,60.12077)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-1" x="67.550359" y="11.199191"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-1" x="75.266415" y="11.199191"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-2" x="79.108574" y="11.199191"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-2" x="83.740191" y="11.199191"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(100%,100%,100%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 3.914042 10.169194 L 36.852735 10.169194 L 36.852735 32.847778 L 3.914042 32.847778 Z M 3.914042 10.169194 " transform="matrix(0.991781,0,0,-0.991781,77.567347,60.12077)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-3" x="84.739906" y="41.259079"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-1" x="94.071573" y="42.7408"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-1" x="98.504834" y="41.259079"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-2" x="102.346994" y="41.259079"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-2" x="106.978611" y="41.259079"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(100%,100%,100%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 3.914042 -24.242544 L 36.852735 -24.242544 L 36.852735 -1.567898 L 3.914042 -1.567898 Z M 3.914042 -24.242544 " transform="matrix(0.991781,0,0,-0.991781,77.567347,60.12077)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-3" x="84.739906" y="75.390229"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-2" x="94.071573" y="76.87195"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-1" x="98.504834" y="75.390229"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-2" x="102.346994" y="75.390229"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-2" x="106.978611" y="75.390229"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(100%,100%,100%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -13.593132 21.508486 C -13.593132 25.360458 -16.716459 28.483785 -20.568431 28.483785 C -24.420403 28.483785 -27.54373 25.360458 -27.54373 21.508486 C -27.54373 17.656514 -24.420403 14.537126 -20.568431 14.537126 C -16.716459 14.537126 -13.593132 17.656514 -13.593132 21.508486 Z M -13.593132 21.508486 " transform="matrix(0.991781,0,0,-0.991781,77.567347,60.12077)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph3-1" x="52.86605" y="41.681578"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph4-1" x="47.506466" y="53.445092"/>
</g>
<g clip-path="url(#clip1)" clip-rule="nonzero">
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -77.032512 21.508486 L -32.175549 21.508486 " transform="matrix(0.991781,0,0,-0.991781,77.567347,60.12077)"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.052476 -0.000703674 L 1.609711 1.681088 L 3.086694 -0.000703674 L 1.609711 -1.682495 Z M 6.052476 -0.000703674 " transform="matrix(0.991781,0,0,-0.991781,42.841019,38.788365)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-4" x="4.954104" y="35.001933"/>
</g>
<g clip-path="url(#clip2)" clip-rule="nonzero">
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -77.032512 -12.903252 L -1.217982 -12.903252 " transform="matrix(0.991781,0,0,-0.991781,77.567347,60.12077)"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.054003 0.00172836 L 1.607299 1.68352 L 3.088221 0.00172836 L 1.607299 -1.684002 Z M 6.054003 0.00172836 " transform="matrix(0.991781,0,0,-0.991781,73.54263,72.919683)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-5" x="4.954104" y="69.132091"/>
</g>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -20.568431 -12.903252 L -20.568431 9.901368 " transform="matrix(0.991781,0,0,-0.991781,77.567347,60.12077)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.052063 0.000280675 L 1.609298 1.682072 L 3.086281 0.000280675 L 1.609298 -1.681511 Z M 6.052063 0.000280675 " transform="matrix(0,-0.991781,-0.991781,0,57.168247,53.115602)"/>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 59.144531 72.917969 C 59.144531 71.828125 58.257812 70.945312 57.167969 70.945312 C 56.078125 70.945312 55.191406 71.828125 55.191406 72.917969 C 55.191406 74.011719 56.078125 74.894531 57.167969 74.894531 C 58.257812 74.894531 59.144531 74.011719 59.144531 72.917969 Z M 59.144531 72.917969 "/>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -13.593132 21.508486 L -1.217982 21.508486 " transform="matrix(0.991781,0,0,-0.991781,77.567347,60.12077)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.054003 -0.000703674 L 1.607299 1.681088 L 3.088221 -0.000703674 L 1.607299 -1.682495 Z M 6.054003 -0.000703674 " transform="matrix(0.991781,0,0,-0.991781,73.54263,38.788365)"/>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -44.550698 21.508486 L -44.550698 -34.416003 L 72.402735 -34.416003 " transform="matrix(0.991781,0,0,-0.991781,77.567347,60.12077)"/>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 152.5625 94.253906 L 148.152344 92.582031 L 149.621094 94.253906 L 148.152344 95.921875 Z M 152.5625 94.253906 "/>
<g clip-path="url(#clip3)" clip-rule="nonzero">
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.0543 -0.00183347 L 1.607596 1.683897 L 3.088518 -0.00183347 L 1.607596 -1.683625 Z M 6.0543 -0.00183347 " transform="matrix(0.991781,0,0,-0.991781,146.55796,94.252088)"/>
</g>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 35.359375 38.789062 C 35.359375 37.695312 34.476562 36.8125 33.382812 36.8125 C 32.292969 36.8125 31.40625 37.695312 31.40625 38.789062 C 31.40625 39.878906 32.292969 40.765625 33.382812 40.765625 C 34.476562 40.765625 35.359375 39.878906 35.359375 38.789062 Z M 35.359375 38.789062 "/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-6" x="145.036221" y="90.464308"/>
</g>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 37.349002 21.508486 L 72.402735 21.508486 " transform="matrix(0.991781,0,0,-0.991781,77.567347,60.12077)"/>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 152.5625 38.789062 L 148.152344 37.121094 L 149.621094 38.789062 L 148.152344 40.457031 Z M 152.5625 38.789062 "/>
<g clip-path="url(#clip4)" clip-rule="nonzero">
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.0543 -0.000703674 L 1.607596 1.681088 L 3.088518 -0.000703674 L 1.607596 -1.682495 Z M 6.0543 -0.000703674 " transform="matrix(0.991781,0,0,-0.991781,146.55796,38.788365)"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-7" x="141.152407" y="33.519221"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-1" x="145.747328" y="35.001933"/>
</g>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 37.349002 -12.903252 L 72.402735 -12.903252 " transform="matrix(0.991781,0,0,-0.991781,77.567347,60.12077)"/>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 152.5625 72.917969 L 148.152344 71.25 L 149.621094 72.917969 L 148.152344 74.589844 Z M 152.5625 72.917969 "/>
<g clip-path="url(#clip5)" clip-rule="nonzero">
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.0543 0.00172836 L 1.607596 1.68352 L 3.088518 0.00172836 L 1.607596 -1.684002 Z M 6.0543 0.00172836 " transform="matrix(0.991781,0,0,-0.991781,146.55796,72.919683)"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-7" x="141.152407" y="67.650371"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-2" x="145.747328" y="69.132091"/>
</g>
</g>
</svg>
Side by Side

Before

Width:  |  Height:  |  Size: 25 KiB

BIN
journal/figs/h_infinity_robust_fusion.pdf → journal/figs/h_infinity_robust_fusion_fb.pdf
View File

Binary file not shown.

BIN
journal/figs/h_infinity_robust_fusion_fb.png Normal file
View File

Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 11 KiB

183
journal/figs/h_infinity_robust_fusion_fb.svg Normal file
View File

@ -0,0 +1,183 @@
<?xml version="1.0" encoding="UTF-8"?>
<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="156.063pt" height="133.48pt" viewBox="0 0 156.063 133.48" version="1.2">
<defs>
<g>
<symbol overflow="visible" id="glyph0-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph0-1">
<path style="stroke:none;" d="M 7.46875 -5.265625 C 7.46875 -6.03125 6.8125 -6.765625 5.515625 -6.765625 L 2.3125 -6.765625 C 2.125 -6.765625 2.015625 -6.765625 2.015625 -6.578125 C 2.015625 -6.453125 2.09375 -6.453125 2.296875 -6.453125 C 2.421875 -6.453125 2.609375 -6.453125 2.71875 -6.4375 C 2.875 -6.421875 2.9375 -6.390625 2.9375 -6.28125 C 2.9375 -6.234375 2.9375 -6.21875 2.90625 -6.09375 L 1.578125 -0.765625 C 1.46875 -0.390625 1.453125 -0.3125 0.671875 -0.3125 C 0.5 -0.3125 0.390625 -0.3125 0.390625 -0.125 C 0.390625 0 0.515625 0 0.546875 0 L 1.796875 -0.03125 L 2.4375 -0.015625 C 2.65625 -0.015625 2.875 0 3.078125 0 C 3.15625 0 3.28125 0 3.28125 -0.203125 C 3.28125 -0.3125 3.1875 -0.3125 3 -0.3125 C 2.640625 -0.3125 2.359375 -0.3125 2.359375 -0.484375 C 2.359375 -0.546875 2.375 -0.59375 2.390625 -0.65625 L 3 -3.125 L 4.6875 -3.125 C 6.09375 -3.125 7.46875 -4.15625 7.46875 -5.265625 Z M 6.546875 -5.515625 C 6.546875 -5.125 6.34375 -4.265625 5.96875 -3.90625 C 5.46875 -3.46875 4.875 -3.390625 4.4375 -3.390625 L 3.046875 -3.390625 L 3.71875 -6.078125 C 3.796875 -6.421875 3.828125 -6.453125 4.25 -6.453125 L 5.203125 -6.453125 C 6.015625 -6.453125 6.546875 -6.1875 6.546875 -5.515625 Z M 6.546875 -5.515625 "/>
</symbol>
<symbol overflow="visible" id="glyph0-2">
<path style="stroke:none;" d="M 3.921875 -1.53125 C 3.921875 -1.890625 3.734375 -2.140625 3.609375 -2.265625 C 3.34375 -2.53125 3.0625 -2.578125 2.625 -2.671875 C 2.265625 -2.75 1.875 -2.828125 1.875 -3.265625 C 1.875 -3.5625 2.109375 -4.15625 2.984375 -4.15625 C 3.234375 -4.15625 3.71875 -4.09375 3.875 -3.703125 C 3.59375 -3.6875 3.390625 -3.484375 3.390625 -3.265625 C 3.390625 -3.125 3.484375 -2.96875 3.703125 -2.96875 C 3.921875 -2.96875 4.15625 -3.140625 4.15625 -3.53125 C 4.15625 -3.96875 3.734375 -4.375 2.984375 -4.375 C 1.671875 -4.375 1.3125 -3.375 1.3125 -2.9375 C 1.3125 -2.15625 2.046875 -2.015625 2.328125 -1.953125 C 2.84375 -1.859375 3.359375 -1.75 3.359375 -1.203125 C 3.359375 -0.9375 3.125 -0.109375 1.9375 -0.109375 C 1.796875 -0.109375 1.046875 -0.109375 0.8125 -0.640625 C 1.1875 -0.578125 1.4375 -0.875 1.4375 -1.15625 C 1.4375 -1.390625 1.28125 -1.5 1.0625 -1.5 C 0.8125 -1.5 0.515625 -1.296875 0.515625 -0.859375 C 0.515625 -0.28125 1.078125 0.109375 1.9375 0.109375 C 3.53125 0.109375 3.921875 -1.09375 3.921875 -1.53125 Z M 3.921875 -1.53125 "/>
</symbol>
<symbol overflow="visible" id="glyph0-3">
<path style="stroke:none;" d="M 7.109375 -0.203125 C 7.109375 -0.3125 7.03125 -0.3125 6.828125 -0.3125 C 6.46875 -0.3125 6.1875 -0.3125 6.1875 -0.484375 C 6.1875 -0.546875 6.21875 -0.59375 6.21875 -0.65625 L 7.5625 -6 C 7.640625 -6.359375 7.671875 -6.453125 8.390625 -6.453125 C 8.640625 -6.453125 8.734375 -6.453125 8.734375 -6.65625 C 8.734375 -6.765625 8.625 -6.765625 8.59375 -6.765625 L 7.328125 -6.734375 L 6.046875 -6.765625 C 5.96875 -6.765625 5.859375 -6.765625 5.859375 -6.5625 C 5.859375 -6.453125 5.9375 -6.453125 6.125 -6.453125 C 6.125 -6.453125 6.34375 -6.453125 6.515625 -6.4375 C 6.6875 -6.421875 6.78125 -6.40625 6.78125 -6.28125 C 6.78125 -6.234375 6.765625 -6.21875 6.734375 -6.09375 L 6.140625 -3.671875 L 3.125 -3.671875 L 3.703125 -6 C 3.796875 -6.359375 3.828125 -6.453125 4.53125 -6.453125 C 4.796875 -6.453125 4.875 -6.453125 4.875 -6.65625 C 4.875 -6.765625 4.765625 -6.765625 4.734375 -6.765625 L 3.46875 -6.734375 L 2.1875 -6.765625 C 2.109375 -6.765625 2 -6.765625 2 -6.5625 C 2 -6.453125 2.09375 -6.453125 2.28125 -6.453125 C 2.28125 -6.453125 2.484375 -6.453125 2.65625 -6.4375 C 2.828125 -6.421875 2.921875 -6.40625 2.921875 -6.28125 C 2.921875 -6.234375 2.90625 -6.21875 2.875 -6.09375 L 1.5625 -0.765625 C 1.453125 -0.390625 1.4375 -0.3125 0.65625 -0.3125 C 0.46875 -0.3125 0.390625 -0.3125 0.390625 -0.109375 C 0.390625 0 0.53125 0 0.53125 0 L 1.78125 -0.03125 L 2.421875 -0.015625 C 2.640625 -0.015625 2.859375 0 3.0625 0 C 3.140625 0 3.265625 0 3.265625 -0.203125 C 3.265625 -0.3125 3.171875 -0.3125 2.984375 -0.3125 C 2.609375 -0.3125 2.34375 -0.3125 2.34375 -0.484375 C 2.34375 -0.546875 2.359375 -0.59375 2.375 -0.65625 L 3.046875 -3.375 L 6.0625 -3.375 L 5.375 -0.640625 C 5.28125 -0.3125 5.09375 -0.3125 4.484375 -0.3125 C 4.328125 -0.3125 4.234375 -0.3125 4.234375 -0.109375 C 4.234375 0 4.375 0 4.375 0 L 5.640625 -0.03125 L 6.265625 -0.015625 C 6.484375 -0.015625 6.703125 0 6.921875 0 C 7 0 7.109375 0 7.109375 -0.203125 Z M 7.109375 -0.203125 "/>
</symbol>
<symbol overflow="visible" id="glyph0-4">
<path style="stroke:none;" d="M 10.375 -6.65625 C 10.375 -6.703125 10.34375 -6.765625 10.265625 -6.765625 C 10.03125 -6.765625 9.75 -6.734375 9.5 -6.734375 C 9.15625 -6.734375 8.796875 -6.765625 8.46875 -6.765625 C 8.40625 -6.765625 8.28125 -6.765625 8.28125 -6.578125 C 8.28125 -6.46875 8.375 -6.453125 8.4375 -6.453125 C 8.671875 -6.453125 9.015625 -6.375 9.015625 -6.078125 C 9.015625 -5.96875 8.96875 -5.890625 8.890625 -5.75 L 6.21875 -1.078125 L 5.84375 -6 C 5.828125 -6.1875 5.8125 -6.453125 6.53125 -6.453125 C 6.703125 -6.453125 6.8125 -6.453125 6.8125 -6.65625 C 6.8125 -6.75 6.703125 -6.765625 6.65625 -6.765625 C 6.265625 -6.765625 5.84375 -6.734375 5.453125 -6.734375 C 5.21875 -6.734375 4.640625 -6.765625 4.40625 -6.765625 C 4.34375 -6.765625 4.21875 -6.765625 4.21875 -6.5625 C 4.21875 -6.453125 4.3125 -6.453125 4.453125 -6.453125 C 4.890625 -6.453125 4.96875 -6.40625 4.984375 -6.21875 L 5.046875 -5.453125 L 2.53125 -1.078125 L 2.15625 -6.109375 C 2.15625 -6.21875 2.15625 -6.453125 2.90625 -6.453125 C 3.015625 -6.453125 3.125 -6.453125 3.125 -6.65625 C 3.125 -6.765625 2.984375 -6.765625 2.984375 -6.765625 C 2.578125 -6.765625 2.171875 -6.734375 1.765625 -6.734375 C 1.421875 -6.734375 1.0625 -6.765625 0.71875 -6.765625 C 0.671875 -6.765625 0.546875 -6.765625 0.546875 -6.578125 C 0.546875 -6.453125 0.640625 -6.453125 0.796875 -6.453125 C 1.28125 -6.453125 1.296875 -6.375 1.3125 -6.09375 L 1.765625 -0.046875 C 1.765625 0.140625 1.78125 0.21875 1.921875 0.21875 C 2.046875 0.21875 2.078125 0.15625 2.15625 0.015625 L 5.078125 -5.046875 L 5.4375 -0.046875 C 5.453125 0.171875 5.484375 0.21875 5.59375 0.21875 C 5.71875 0.21875 5.78125 0.125 5.828125 0.03125 L 9.078125 -5.609375 C 9.328125 -6.046875 9.5625 -6.40625 10.1875 -6.453125 C 10.28125 -6.46875 10.375 -6.46875 10.375 -6.65625 Z M 10.375 -6.65625 "/>
</symbol>
<symbol overflow="visible" id="glyph0-5">
<path style="stroke:none;" d="M 6.84375 -3.6875 C 6.84375 -4.234375 6.578125 -4.375 6.40625 -4.375 C 6.15625 -4.375 5.90625 -4.125 5.90625 -3.90625 C 5.90625 -3.78125 5.96875 -3.71875 6.0625 -3.640625 C 6.171875 -3.53125 6.421875 -3.265625 6.421875 -2.796875 C 6.421875 -2.453125 6.125 -1.484375 5.875 -0.984375 C 5.609375 -0.453125 5.265625 -0.109375 4.78125 -0.109375 C 4.3125 -0.109375 4.046875 -0.40625 4.046875 -0.96875 C 4.046875 -1.25 4.125 -1.5625 4.15625 -1.6875 L 4.578125 -3.359375 C 4.625 -3.578125 4.71875 -3.9375 4.71875 -4 C 4.71875 -4.1875 4.59375 -4.265625 4.4375 -4.265625 C 4.3125 -4.265625 4.140625 -4.1875 4.078125 -4 C 4.046875 -3.921875 3.578125 -2.03125 3.5 -1.765625 C 3.4375 -1.46875 3.421875 -1.296875 3.421875 -1.125 C 3.421875 -1.015625 3.421875 -0.984375 3.421875 -0.9375 C 3.203125 -0.421875 2.90625 -0.109375 2.515625 -0.109375 C 1.71875 -0.109375 1.71875 -0.84375 1.71875 -1.015625 C 1.71875 -1.328125 1.765625 -1.71875 2.234375 -2.9375 C 2.34375 -3.234375 2.40625 -3.375 2.40625 -3.5625 C 2.40625 -4.015625 2.078125 -4.375 1.59375 -4.375 C 0.65625 -4.375 0.28125 -2.9375 0.28125 -2.859375 C 0.28125 -2.75 0.40625 -2.75 0.40625 -2.75 C 0.5 -2.75 0.515625 -2.78125 0.5625 -2.9375 C 0.828125 -3.859375 1.21875 -4.15625 1.5625 -4.15625 C 1.65625 -4.15625 1.8125 -4.15625 1.8125 -3.828125 C 1.8125 -3.59375 1.703125 -3.3125 1.640625 -3.140625 C 1.203125 -1.96875 1.078125 -1.515625 1.078125 -1.140625 C 1.078125 -0.234375 1.75 0.109375 2.484375 0.109375 C 2.65625 0.109375 3.125 0.109375 3.515625 -0.578125 C 3.78125 0.046875 4.453125 0.109375 4.75 0.109375 C 5.5 0.109375 5.9375 -0.515625 6.1875 -1.109375 C 6.53125 -1.875 6.84375 -3.203125 6.84375 -3.6875 Z M 6.84375 -3.6875 "/>
</symbol>
<symbol overflow="visible" id="glyph0-6">
<path style="stroke:none;" d="M 4.3125 -1.421875 C 4.3125 -1.46875 4.265625 -1.515625 4.1875 -1.515625 C 4.09375 -1.515625 4.078125 -1.453125 4.046875 -1.390625 C 3.828125 -0.75 3.1875 -0.5625 2.859375 -0.5625 C 2.671875 -0.5625 2.484375 -0.609375 2.28125 -0.6875 C 1.9375 -0.8125 1.796875 -0.859375 1.578125 -0.859375 C 1.578125 -0.859375 1.40625 -0.859375 1.3125 -0.828125 C 1.859375 -1.390625 2.140625 -1.640625 2.5 -1.953125 C 2.5 -1.953125 3.109375 -2.484375 3.46875 -2.84375 C 4.40625 -3.765625 4.625 -4.234375 4.625 -4.28125 C 4.625 -4.375 4.515625 -4.375 4.515625 -4.375 C 4.453125 -4.375 4.421875 -4.359375 4.375 -4.265625 C 4.078125 -3.796875 3.859375 -3.640625 3.625 -3.640625 C 3.390625 -3.640625 3.265625 -3.78125 3.125 -3.953125 C 2.9375 -4.1875 2.765625 -4.375 2.4375 -4.375 C 1.6875 -4.375 1.234375 -3.453125 1.234375 -3.25 C 1.234375 -3.203125 1.265625 -3.140625 1.359375 -3.140625 C 1.453125 -3.140625 1.46875 -3.1875 1.484375 -3.25 C 1.671875 -3.703125 2.25 -3.71875 2.328125 -3.71875 C 2.53125 -3.71875 2.71875 -3.640625 2.953125 -3.5625 C 3.34375 -3.421875 3.453125 -3.421875 3.71875 -3.421875 C 3.359375 -2.984375 2.53125 -2.28125 2.34375 -2.125 L 1.453125 -1.28125 C 0.765625 -0.625 0.421875 -0.0625 0.421875 0.015625 C 0.421875 0.109375 0.546875 0.109375 0.546875 0.109375 C 0.625 0.109375 0.640625 0.09375 0.703125 -0.015625 C 0.9375 -0.359375 1.234375 -0.640625 1.546875 -0.640625 C 1.765625 -0.640625 1.875 -0.546875 2.125 -0.25 C 2.28125 -0.046875 2.46875 0.109375 2.75 0.109375 C 3.734375 0.109375 4.3125 -1.15625 4.3125 -1.421875 Z M 4.3125 -1.421875 "/>
</symbol>
<symbol overflow="visible" id="glyph1-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph1-1">
<path style="stroke:none;" d="M 3.28125 2.375 C 3.28125 2.34375 3.28125 2.328125 3.109375 2.15625 C 1.875 0.90625 1.5625 -0.953125 1.5625 -2.484375 C 1.5625 -4.203125 1.9375 -5.921875 3.15625 -7.15625 C 3.28125 -7.28125 3.28125 -7.296875 3.28125 -7.328125 C 3.28125 -7.40625 3.234375 -7.4375 3.1875 -7.4375 C 3.078125 -7.4375 2.1875 -6.75 1.609375 -5.5 C 1.09375 -4.40625 0.984375 -3.3125 0.984375 -2.484375 C 0.984375 -1.703125 1.09375 -0.5 1.640625 0.609375 C 2.234375 1.828125 3.078125 2.484375 3.1875 2.484375 C 3.234375 2.484375 3.28125 2.453125 3.28125 2.375 Z M 3.28125 2.375 "/>
</symbol>
<symbol overflow="visible" id="glyph1-2">
<path style="stroke:none;" d="M 2.859375 -2.484375 C 2.859375 -3.25 2.75 -4.453125 2.203125 -5.5625 C 1.609375 -6.78125 0.765625 -7.4375 0.65625 -7.4375 C 0.609375 -7.4375 0.5625 -7.390625 0.5625 -7.328125 C 0.5625 -7.296875 0.5625 -7.28125 0.75 -7.109375 C 1.71875 -6.125 2.28125 -4.546875 2.28125 -2.484375 C 2.28125 -0.78125 1.921875 0.953125 0.6875 2.203125 C 0.5625 2.328125 0.5625 2.34375 0.5625 2.375 C 0.5625 2.4375 0.609375 2.484375 0.65625 2.484375 C 0.765625 2.484375 1.65625 1.796875 2.234375 0.546875 C 2.75 -0.546875 2.859375 -1.640625 2.859375 -2.484375 Z M 2.859375 -2.484375 "/>
</symbol>
<symbol overflow="visible" id="glyph2-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph2-1">
<path style="stroke:none;" d="M 3.5 -1.265625 L 3.265625 -1.265625 C 3.25 -1.109375 3.171875 -0.703125 3.09375 -0.625 C 3.03125 -0.59375 2.5 -0.59375 2.40625 -0.59375 L 1.125 -0.59375 C 1.859375 -1.234375 2.09375 -1.421875 2.515625 -1.75 C 3.03125 -2.171875 3.5 -2.59375 3.5 -3.25 C 3.5 -4.09375 2.765625 -4.609375 1.875 -4.609375 C 1.015625 -4.609375 0.4375 -4 0.4375 -3.359375 C 0.4375 -3.015625 0.734375 -2.96875 0.8125 -2.96875 C 0.96875 -2.96875 1.171875 -3.09375 1.171875 -3.34375 C 1.171875 -3.46875 1.125 -3.71875 0.765625 -3.71875 C 0.984375 -4.203125 1.453125 -4.359375 1.78125 -4.359375 C 2.46875 -4.359375 2.828125 -3.8125 2.828125 -3.25 C 2.828125 -2.65625 2.40625 -2.171875 2.171875 -1.921875 L 0.5 -0.265625 C 0.4375 -0.203125 0.4375 -0.1875 0.4375 0 L 3.296875 0 Z M 3.5 -1.265625 "/>
</symbol>
<symbol overflow="visible" id="glyph2-2">
<path style="stroke:none;" d="M 3.28125 0 L 3.28125 -0.25 L 3.015625 -0.25 C 2.328125 -0.25 2.328125 -0.34375 2.328125 -0.5625 L 2.328125 -4.40625 C 2.328125 -4.59375 2.3125 -4.609375 2.109375 -4.609375 C 1.671875 -4.171875 1.046875 -4.15625 0.75 -4.15625 L 0.75 -3.90625 C 0.921875 -3.90625 1.375 -3.90625 1.765625 -4.109375 L 1.765625 -0.5625 C 1.765625 -0.34375 1.765625 -0.25 1.0625 -0.25 L 0.8125 -0.25 L 0.8125 0 L 2.046875 -0.03125 Z M 3.28125 0 "/>
</symbol>
<symbol overflow="visible" id="glyph3-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph3-1">
<path style="stroke:none;" d="M 8.0625 -2.984375 C 8.0625 -3.203125 7.875 -3.203125 7.6875 -3.203125 L 4.515625 -3.203125 L 4.515625 -6.375 C 4.515625 -6.546875 4.515625 -6.75 4.3125 -6.75 C 4.09375 -6.75 4.09375 -6.5625 4.09375 -6.375 L 4.09375 -3.203125 L 0.921875 -3.203125 C 0.75 -3.203125 0.546875 -3.203125 0.546875 -3 C 0.546875 -2.78125 0.75 -2.78125 0.921875 -2.78125 L 4.09375 -2.78125 L 4.09375 0.390625 C 4.09375 0.5625 4.09375 0.765625 4.296875 0.765625 C 4.515625 0.765625 4.515625 0.578125 4.515625 0.390625 L 4.515625 -2.78125 L 7.6875 -2.78125 C 7.859375 -2.78125 8.0625 -2.78125 8.0625 -2.984375 Z M 8.0625 -2.984375 "/>
</symbol>
<symbol overflow="visible" id="glyph4-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph4-1">
<path style="stroke:none;" d="M 6.875 -2.484375 C 6.875 -2.671875 6.6875 -2.671875 6.546875 -2.671875 L 1.15625 -2.671875 C 1.015625 -2.671875 0.828125 -2.671875 0.828125 -2.484375 C 0.828125 -2.28125 1.015625 -2.28125 1.15625 -2.28125 L 6.546875 -2.28125 C 6.6875 -2.28125 6.875 -2.28125 6.875 -2.484375 Z M 6.875 -2.484375 "/>
</symbol>
</g>
<clipPath id="clip1">
<path d="M 61 108 L 94 108 L 94 132.824219 L 61 132.824219 Z M 61 108 "/>
</clipPath>
<clipPath id="clip2">
<path d="M 81 104 L 114 104 L 114 132.824219 L 81 132.824219 Z M 81 104 "/>
</clipPath>
<clipPath id="clip3">
<path d="M 134 23 L 155.132812 23 L 155.132812 55 L 134 55 Z M 134 23 "/>
</clipPath>
<clipPath id="clip4">
<path d="M 134 57 L 155.132812 57 L 155.132812 89 L 134 89 Z M 134 57 "/>
</clipPath>
</defs>
<g id="surface1">
<path style="fill-rule:nonzero;fill:rgb(89.99939%,89.99939%,89.99939%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -56.693712 -42.52075 L 56.69394 -42.52075 L 56.69394 42.521954 L -56.693712 42.521954 Z M -56.693712 -42.52075 " transform="matrix(0.994032,0,0,-0.994032,77.566293,60.397083)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-1" x="67.526571" y="11.364474"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-1" x="75.260139" y="11.364474"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-2" x="79.111019" y="11.364474"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-2" x="83.753147" y="11.364474"/>
</g>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(100%,100%,100%);fill-opacity:1;" d="M 61.753906 131.832031 L 93.375 131.832031 L 93.375 109.289062 L 61.753906 109.289062 Z M 61.753906 131.832031 "/>
<g clip-path="url(#clip1)" clip-rule="nonzero">
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -15.907324 -71.863843 L 15.903622 -71.863843 L 15.903622 -49.185526 L -15.907324 -49.185526 Z M -15.907324 -71.863843 " transform="matrix(0.994032,0,0,-0.994032,77.566293,60.397083)"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-3" x="65.056402" y="123.036"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-1" x="73.28798" y="124.521083"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-1" x="77.730308" y="123.036"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-2" x="81.582182" y="123.036"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-2" x="86.22431" y="123.036"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(100%,100%,100%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 3.910168 10.168709 L 36.84894 10.168709 L 36.84894 32.847025 L 3.910168 32.847025 Z M 3.910168 10.168709 " transform="matrix(0.994032,0,0,-0.994032,77.566293,60.397083)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-4" x="84.755131" y="41.492585"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-2" x="94.107977" y="42.978663"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-1" x="98.551299" y="41.492585"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-2" x="102.402179" y="41.492585"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-2" x="107.044308" y="41.492585"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(100%,100%,100%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 3.910168 -24.243701 L 36.84894 -24.243701 L 36.84894 -1.565385 L 3.910168 -1.565385 Z M 3.910168 -24.243701 " transform="matrix(0.994032,0,0,-0.994032,77.566293,60.397083)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-4" x="84.755131" y="75.701197"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-1" x="94.107977" y="77.186281"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-1" x="98.551299" y="75.701197"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-2" x="102.402179" y="75.701197"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-2" x="107.044308" y="75.701197"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(100%,100%,100%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -13.592729 21.509832 C -13.592729 25.360941 -16.716843 28.485055 -20.567952 28.485055 C -24.419061 28.485055 -27.543175 25.360941 -27.543175 21.509832 C -27.543175 17.658723 -24.419061 14.534609 -20.567952 14.534609 C -16.716843 14.534609 -13.592729 17.658723 -13.592729 21.509832 Z M -13.592729 21.509832 " transform="matrix(0.994032,0,0,-0.994032,77.566293,60.397083)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph3-1" x="52.808936" y="41.915049"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph4-1" x="47.437188" y="53.70526"/>
</g>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -77.033855 21.509832 L -32.176295 21.509832 " transform="matrix(0.994032,0,0,-0.994032,77.566293,60.397083)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.052312 0.000641747 L 1.607818 1.682555 L 3.089316 0.000641747 L 1.607818 -1.681271 Z M 6.052312 0.000641747 " transform="matrix(0.994032,0,0,-0.994032,42.761153,39.016263)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-5" x="4.788251" y="35.220244"/>
</g>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -20.567952 -12.906508 L -20.567952 9.901489 " transform="matrix(0.994032,0,0,-0.994032,77.566293,60.397083)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.052766 -0.000197973 L 1.608272 1.681715 L 3.08584 -0.000197973 L 1.608272 -1.682111 Z M 6.052766 -0.000197973 " transform="matrix(0,-0.994032,-0.994032,0,57.120897,53.376017)"/>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 59.101562 73.226562 C 59.101562 72.132812 58.214844 71.246094 57.121094 71.246094 C 56.027344 71.246094 55.140625 72.132812 55.140625 73.226562 C 55.140625 74.320312 56.027344 75.207031 57.121094 75.207031 C 58.214844 75.207031 59.101562 74.320312 59.101562 73.226562 Z M 59.101562 73.226562 "/>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -13.592729 21.509832 L -1.218094 21.509832 " transform="matrix(0.994032,0,0,-0.994032,77.566293,60.397083)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.054473 0.000641747 L 1.609979 1.682555 L 3.087547 0.000641747 L 1.609979 -1.681271 Z M 6.054473 0.000641747 " transform="matrix(0.994032,0,0,-0.994032,73.532442,39.016263)"/>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -44.55093 21.509832 L -44.55093 -34.413773 L 68.530206 -34.413773 L 68.530206 -60.52272 L 21.035814 -60.52272 " transform="matrix(0.994032,0,0,-0.994032,77.566293,60.397083)"/>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 95.28125 120.558594 L 99.699219 122.234375 L 98.230469 120.558594 L 99.699219 118.886719 Z M 95.28125 120.558594 "/>
<g clip-path="url(#clip2)" clip-rule="nonzero">
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.054433 -0.00180019 L 1.609939 1.684042 L 3.087507 -0.00180019 L 1.609939 -1.683713 Z M 6.054433 -0.00180019 " transform="matrix(-0.994032,0,0,0.994032,101.299549,120.560383)"/>
</g>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 35.261719 39.015625 C 35.261719 37.921875 34.375 37.035156 33.28125 37.035156 C 32.1875 37.035156 31.300781 37.921875 31.300781 39.015625 C 31.300781 40.109375 32.1875 40.996094 33.28125 40.996094 C 34.375 40.996094 35.261719 40.109375 35.261719 39.015625 Z M 35.261719 39.015625 "/>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -16.402467 -60.52272 L -68.529978 -60.52272 L -68.529978 -12.906508 L -1.218094 -12.906508 " transform="matrix(0.994032,0,0,-0.994032,77.566293,60.397083)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.054473 -0.00152773 L 1.609979 1.684315 L 3.087547 -0.00152773 L 1.609979 -1.683441 Z M 6.054473 -0.00152773 " transform="matrix(0.994032,0,0,-0.994032,73.532442,73.225044)"/>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 37.348012 21.509832 L 72.400963 21.509832 " transform="matrix(0.994032,0,0,-0.994032,77.566293,60.397083)"/>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 152.730469 39.015625 L 148.3125 37.34375 L 149.78125 39.015625 L 148.3125 40.6875 Z M 152.730469 39.015625 "/>
<g clip-path="url(#clip3)" clip-rule="nonzero">
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.05311 0.000641747 L 1.608616 1.682555 L 3.086184 0.000641747 L 1.608616 -1.681271 Z M 6.05311 0.000641747 " transform="matrix(0.994032,0,0,-0.994032,146.713484,39.016263)"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-6" x="141.295663" y="33.735161"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-2" x="145.901012" y="35.220244"/>
</g>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 37.348012 -12.906508 L 72.400963 -12.906508 " transform="matrix(0.994032,0,0,-0.994032,77.566293,60.397083)"/>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 152.730469 73.226562 L 148.3125 71.550781 L 149.78125 73.226562 L 148.3125 74.898438 Z M 152.730469 73.226562 "/>
<g clip-path="url(#clip4)" clip-rule="nonzero">
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.05311 -0.00152773 L 1.608616 1.684315 L 3.086184 -0.00152773 L 1.608616 -1.683441 Z M 6.05311 -0.00152773 " transform="matrix(0.994032,0,0,-0.994032,146.713484,73.225044)"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-6" x="141.295663" y="67.942779"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-1" x="145.901012" y="69.428856"/>
</g>
</g>
</svg>
Side by Side

After

Width:  |  Height:  |  Size: 25 KiB

BIN
journal/figs/h_infinity_robust_fusion_plant.pdf Normal file
View File

Binary file not shown.

BIN
journal/figs/h_infinity_robust_fusion_plant.png Normal file
View File

Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 9.8 KiB

170
journal/figs/h_infinity_robust_fusion_plant.svg Normal file
View File

@ -0,0 +1,170 @@
<?xml version="1.0" encoding="UTF-8"?>
<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="156.063pt" height="133.48pt" viewBox="0 0 156.063 133.48" version="1.2">
<defs>
<g>
<symbol overflow="visible" id="glyph0-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph0-1">
<path style="stroke:none;" d="M 7.46875 -5.265625 C 7.46875 -6.03125 6.8125 -6.765625 5.515625 -6.765625 L 2.3125 -6.765625 C 2.125 -6.765625 2.015625 -6.765625 2.015625 -6.578125 C 2.015625 -6.453125 2.09375 -6.453125 2.296875 -6.453125 C 2.421875 -6.453125 2.609375 -6.453125 2.71875 -6.4375 C 2.875 -6.421875 2.9375 -6.390625 2.9375 -6.28125 C 2.9375 -6.234375 2.9375 -6.21875 2.90625 -6.09375 L 1.578125 -0.765625 C 1.46875 -0.390625 1.453125 -0.3125 0.671875 -0.3125 C 0.5 -0.3125 0.390625 -0.3125 0.390625 -0.125 C 0.390625 0 0.515625 0 0.546875 0 L 1.796875 -0.03125 L 2.4375 -0.015625 C 2.65625 -0.015625 2.875 0 3.078125 0 C 3.15625 0 3.28125 0 3.28125 -0.203125 C 3.28125 -0.3125 3.1875 -0.3125 3 -0.3125 C 2.640625 -0.3125 2.359375 -0.3125 2.359375 -0.484375 C 2.359375 -0.546875 2.375 -0.59375 2.390625 -0.65625 L 3 -3.125 L 4.6875 -3.125 C 6.09375 -3.125 7.46875 -4.15625 7.46875 -5.265625 Z M 6.546875 -5.515625 C 6.546875 -5.125 6.34375 -4.265625 5.96875 -3.90625 C 5.46875 -3.46875 4.875 -3.390625 4.4375 -3.390625 L 3.046875 -3.390625 L 3.71875 -6.078125 C 3.796875 -6.421875 3.828125 -6.453125 4.25 -6.453125 L 5.203125 -6.453125 C 6.015625 -6.453125 6.546875 -6.1875 6.546875 -5.515625 Z M 6.546875 -5.515625 "/>
</symbol>
<symbol overflow="visible" id="glyph0-2">
<path style="stroke:none;" d="M 3.921875 -1.53125 C 3.921875 -1.890625 3.734375 -2.140625 3.609375 -2.265625 C 3.34375 -2.53125 3.0625 -2.578125 2.625 -2.671875 C 2.265625 -2.75 1.875 -2.828125 1.875 -3.265625 C 1.875 -3.5625 2.109375 -4.15625 2.984375 -4.15625 C 3.234375 -4.15625 3.71875 -4.09375 3.875 -3.703125 C 3.59375 -3.6875 3.390625 -3.484375 3.390625 -3.265625 C 3.390625 -3.125 3.484375 -2.96875 3.703125 -2.96875 C 3.921875 -2.96875 4.15625 -3.140625 4.15625 -3.53125 C 4.15625 -3.96875 3.734375 -4.375 2.984375 -4.375 C 1.671875 -4.375 1.3125 -3.375 1.3125 -2.9375 C 1.3125 -2.15625 2.046875 -2.015625 2.328125 -1.953125 C 2.84375 -1.859375 3.359375 -1.75 3.359375 -1.203125 C 3.359375 -0.9375 3.125 -0.109375 1.9375 -0.109375 C 1.796875 -0.109375 1.046875 -0.109375 0.8125 -0.640625 C 1.1875 -0.578125 1.4375 -0.875 1.4375 -1.15625 C 1.4375 -1.390625 1.28125 -1.5 1.0625 -1.5 C 0.8125 -1.5 0.515625 -1.296875 0.515625 -0.859375 C 0.515625 -0.28125 1.078125 0.109375 1.9375 0.109375 C 3.53125 0.109375 3.921875 -1.09375 3.921875 -1.53125 Z M 3.921875 -1.53125 "/>
</symbol>
<symbol overflow="visible" id="glyph0-3">
<path style="stroke:none;" d="M 10.375 -6.65625 C 10.375 -6.703125 10.34375 -6.765625 10.265625 -6.765625 C 10.03125 -6.765625 9.75 -6.734375 9.5 -6.734375 C 9.15625 -6.734375 8.796875 -6.765625 8.46875 -6.765625 C 8.40625 -6.765625 8.28125 -6.765625 8.28125 -6.578125 C 8.28125 -6.46875 8.375 -6.453125 8.4375 -6.453125 C 8.671875 -6.453125 9.015625 -6.375 9.015625 -6.078125 C 9.015625 -5.96875 8.96875 -5.890625 8.890625 -5.75 L 6.21875 -1.078125 L 5.84375 -6 C 5.828125 -6.1875 5.8125 -6.453125 6.53125 -6.453125 C 6.703125 -6.453125 6.8125 -6.453125 6.8125 -6.65625 C 6.8125 -6.75 6.703125 -6.765625 6.65625 -6.765625 C 6.265625 -6.765625 5.84375 -6.734375 5.453125 -6.734375 C 5.21875 -6.734375 4.640625 -6.765625 4.40625 -6.765625 C 4.34375 -6.765625 4.21875 -6.765625 4.21875 -6.5625 C 4.21875 -6.453125 4.3125 -6.453125 4.453125 -6.453125 C 4.890625 -6.453125 4.96875 -6.40625 4.984375 -6.21875 L 5.046875 -5.453125 L 2.53125 -1.078125 L 2.15625 -6.109375 C 2.15625 -6.21875 2.15625 -6.453125 2.90625 -6.453125 C 3.015625 -6.453125 3.125 -6.453125 3.125 -6.65625 C 3.125 -6.765625 2.984375 -6.765625 2.984375 -6.765625 C 2.578125 -6.765625 2.171875 -6.734375 1.765625 -6.734375 C 1.421875 -6.734375 1.0625 -6.765625 0.71875 -6.765625 C 0.671875 -6.765625 0.546875 -6.765625 0.546875 -6.578125 C 0.546875 -6.453125 0.640625 -6.453125 0.796875 -6.453125 C 1.28125 -6.453125 1.296875 -6.375 1.3125 -6.09375 L 1.765625 -0.046875 C 1.765625 0.140625 1.78125 0.21875 1.921875 0.21875 C 2.046875 0.21875 2.078125 0.15625 2.15625 0.015625 L 5.078125 -5.046875 L 5.4375 -0.046875 C 5.453125 0.171875 5.484375 0.21875 5.59375 0.21875 C 5.71875 0.21875 5.78125 0.125 5.828125 0.03125 L 9.078125 -5.609375 C 9.328125 -6.046875 9.5625 -6.40625 10.1875 -6.453125 C 10.28125 -6.46875 10.375 -6.46875 10.375 -6.65625 Z M 10.375 -6.65625 "/>
</symbol>
<symbol overflow="visible" id="glyph0-4">
<path style="stroke:none;" d="M 6.84375 -3.6875 C 6.84375 -4.234375 6.578125 -4.375 6.40625 -4.375 C 6.15625 -4.375 5.90625 -4.125 5.90625 -3.90625 C 5.90625 -3.78125 5.96875 -3.71875 6.0625 -3.640625 C 6.171875 -3.53125 6.421875 -3.265625 6.421875 -2.796875 C 6.421875 -2.453125 6.125 -1.484375 5.875 -0.984375 C 5.609375 -0.453125 5.265625 -0.109375 4.78125 -0.109375 C 4.3125 -0.109375 4.046875 -0.40625 4.046875 -0.96875 C 4.046875 -1.25 4.125 -1.5625 4.15625 -1.6875 L 4.578125 -3.359375 C 4.625 -3.578125 4.71875 -3.9375 4.71875 -4 C 4.71875 -4.1875 4.59375 -4.265625 4.4375 -4.265625 C 4.3125 -4.265625 4.140625 -4.1875 4.078125 -4 C 4.046875 -3.921875 3.578125 -2.03125 3.5 -1.765625 C 3.4375 -1.46875 3.421875 -1.296875 3.421875 -1.125 C 3.421875 -1.015625 3.421875 -0.984375 3.421875 -0.9375 C 3.203125 -0.421875 2.90625 -0.109375 2.515625 -0.109375 C 1.71875 -0.109375 1.71875 -0.84375 1.71875 -1.015625 C 1.71875 -1.328125 1.765625 -1.71875 2.234375 -2.9375 C 2.34375 -3.234375 2.40625 -3.375 2.40625 -3.5625 C 2.40625 -4.015625 2.078125 -4.375 1.59375 -4.375 C 0.65625 -4.375 0.28125 -2.9375 0.28125 -2.859375 C 0.28125 -2.75 0.40625 -2.75 0.40625 -2.75 C 0.5 -2.75 0.515625 -2.78125 0.5625 -2.9375 C 0.828125 -3.859375 1.21875 -4.15625 1.5625 -4.15625 C 1.65625 -4.15625 1.8125 -4.15625 1.8125 -3.828125 C 1.8125 -3.59375 1.703125 -3.3125 1.640625 -3.140625 C 1.203125 -1.96875 1.078125 -1.515625 1.078125 -1.140625 C 1.078125 -0.234375 1.75 0.109375 2.484375 0.109375 C 2.65625 0.109375 3.125 0.109375 3.515625 -0.578125 C 3.78125 0.046875 4.453125 0.109375 4.75 0.109375 C 5.5 0.109375 5.9375 -0.515625 6.1875 -1.109375 C 6.53125 -1.875 6.84375 -3.203125 6.84375 -3.6875 Z M 6.84375 -3.6875 "/>
</symbol>
<symbol overflow="visible" id="glyph0-5">
<path style="stroke:none;" d="M 5.375 -1.421875 C 5.375 -1.515625 5.296875 -1.515625 5.265625 -1.515625 C 5.15625 -1.515625 5.15625 -1.46875 5.125 -1.34375 C 4.984375 -0.78125 4.796875 -0.109375 4.375 -0.109375 C 4.171875 -0.109375 4.078125 -0.234375 4.078125 -0.5625 C 4.078125 -0.78125 4.1875 -1.25 4.265625 -1.59375 L 4.546875 -2.671875 C 4.578125 -2.8125 4.671875 -3.1875 4.71875 -3.34375 C 4.765625 -3.5625 4.859375 -3.9375 4.859375 -4 C 4.859375 -4.1875 4.71875 -4.265625 4.578125 -4.265625 C 4.53125 -4.265625 4.265625 -4.265625 4.1875 -3.921875 L 3.453125 -0.9375 C 3.4375 -0.90625 3.046875 -0.109375 2.3125 -0.109375 C 1.796875 -0.109375 1.703125 -0.5625 1.703125 -0.921875 C 1.703125 -1.46875 1.984375 -2.265625 2.234375 -2.9375 C 2.359375 -3.234375 2.40625 -3.375 2.40625 -3.5625 C 2.40625 -4.015625 2.09375 -4.375 1.59375 -4.375 C 0.65625 -4.375 0.28125 -2.9375 0.28125 -2.859375 C 0.28125 -2.75 0.40625 -2.75 0.40625 -2.75 C 0.5 -2.75 0.515625 -2.78125 0.5625 -2.9375 C 0.8125 -3.796875 1.1875 -4.15625 1.5625 -4.15625 C 1.65625 -4.15625 1.8125 -4.15625 1.8125 -3.828125 C 1.8125 -3.59375 1.703125 -3.3125 1.640625 -3.15625 C 1.28125 -2.171875 1.0625 -1.5625 1.0625 -1.078125 C 1.0625 -0.140625 1.75 0.109375 2.28125 0.109375 C 2.9375 0.109375 3.296875 -0.34375 3.46875 -0.5625 C 3.578125 -0.15625 3.921875 0.109375 4.34375 0.109375 C 4.703125 0.109375 4.921875 -0.125 5.078125 -0.4375 C 5.25 -0.796875 5.375 -1.421875 5.375 -1.421875 Z M 5.375 -1.421875 "/>
</symbol>
<symbol overflow="visible" id="glyph0-6">
<path style="stroke:none;" d="M 4.640625 -3.6875 C 4.640625 -4.21875 4.375 -4.375 4.203125 -4.375 C 3.953125 -4.375 3.71875 -4.125 3.71875 -3.90625 C 3.71875 -3.78125 3.765625 -3.71875 3.875 -3.609375 C 4.078125 -3.40625 4.203125 -3.15625 4.203125 -2.796875 C 4.203125 -2.375 3.609375 -0.109375 2.453125 -0.109375 C 1.9375 -0.109375 1.71875 -0.453125 1.71875 -0.96875 C 1.71875 -1.53125 1.984375 -2.25 2.28125 -3.078125 C 2.359375 -3.234375 2.40625 -3.375 2.40625 -3.5625 C 2.40625 -4.015625 2.09375 -4.375 1.59375 -4.375 C 0.65625 -4.375 0.28125 -2.9375 0.28125 -2.859375 C 0.28125 -2.75 0.40625 -2.75 0.40625 -2.75 C 0.5 -2.75 0.515625 -2.78125 0.5625 -2.9375 C 0.859375 -3.9375 1.28125 -4.15625 1.5625 -4.15625 C 1.640625 -4.15625 1.8125 -4.15625 1.8125 -3.84375 C 1.8125 -3.59375 1.71875 -3.328125 1.640625 -3.15625 C 1.203125 -2 1.078125 -1.546875 1.078125 -1.125 C 1.078125 -0.046875 1.953125 0.109375 2.40625 0.109375 C 4.078125 0.109375 4.640625 -3.171875 4.640625 -3.6875 Z M 4.640625 -3.6875 "/>
</symbol>
<symbol overflow="visible" id="glyph0-7">
<path style="stroke:none;" d="M 4.3125 -1.421875 C 4.3125 -1.46875 4.265625 -1.515625 4.1875 -1.515625 C 4.09375 -1.515625 4.078125 -1.453125 4.046875 -1.390625 C 3.828125 -0.75 3.1875 -0.5625 2.859375 -0.5625 C 2.671875 -0.5625 2.484375 -0.609375 2.28125 -0.6875 C 1.9375 -0.8125 1.796875 -0.859375 1.578125 -0.859375 C 1.578125 -0.859375 1.40625 -0.859375 1.3125 -0.828125 C 1.859375 -1.390625 2.140625 -1.640625 2.5 -1.953125 C 2.5 -1.953125 3.109375 -2.484375 3.46875 -2.84375 C 4.40625 -3.765625 4.625 -4.234375 4.625 -4.28125 C 4.625 -4.375 4.515625 -4.375 4.515625 -4.375 C 4.453125 -4.375 4.421875 -4.359375 4.375 -4.265625 C 4.078125 -3.796875 3.859375 -3.640625 3.625 -3.640625 C 3.390625 -3.640625 3.265625 -3.78125 3.125 -3.953125 C 2.9375 -4.1875 2.765625 -4.375 2.4375 -4.375 C 1.6875 -4.375 1.234375 -3.453125 1.234375 -3.25 C 1.234375 -3.203125 1.265625 -3.140625 1.359375 -3.140625 C 1.453125 -3.140625 1.46875 -3.1875 1.484375 -3.25 C 1.671875 -3.703125 2.25 -3.71875 2.328125 -3.71875 C 2.53125 -3.71875 2.71875 -3.640625 2.953125 -3.5625 C 3.34375 -3.421875 3.453125 -3.421875 3.71875 -3.421875 C 3.359375 -2.984375 2.53125 -2.28125 2.34375 -2.125 L 1.453125 -1.28125 C 0.765625 -0.625 0.421875 -0.0625 0.421875 0.015625 C 0.421875 0.109375 0.546875 0.109375 0.546875 0.109375 C 0.625 0.109375 0.640625 0.09375 0.703125 -0.015625 C 0.9375 -0.359375 1.234375 -0.640625 1.546875 -0.640625 C 1.765625 -0.640625 1.875 -0.546875 2.125 -0.25 C 2.28125 -0.046875 2.46875 0.109375 2.75 0.109375 C 3.734375 0.109375 4.3125 -1.15625 4.3125 -1.421875 Z M 4.3125 -1.421875 "/>
</symbol>
<symbol overflow="visible" id="glyph1-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph1-1">
<path style="stroke:none;" d="M 3.28125 2.375 C 3.28125 2.34375 3.28125 2.328125 3.109375 2.15625 C 1.875 0.90625 1.5625 -0.953125 1.5625 -2.484375 C 1.5625 -4.203125 1.9375 -5.921875 3.15625 -7.15625 C 3.28125 -7.28125 3.28125 -7.296875 3.28125 -7.328125 C 3.28125 -7.40625 3.234375 -7.4375 3.1875 -7.4375 C 3.078125 -7.4375 2.1875 -6.75 1.609375 -5.5 C 1.09375 -4.40625 0.984375 -3.3125 0.984375 -2.484375 C 0.984375 -1.703125 1.09375 -0.5 1.640625 0.609375 C 2.234375 1.828125 3.078125 2.484375 3.1875 2.484375 C 3.234375 2.484375 3.28125 2.453125 3.28125 2.375 Z M 3.28125 2.375 "/>
</symbol>
<symbol overflow="visible" id="glyph1-2">
<path style="stroke:none;" d="M 2.859375 -2.484375 C 2.859375 -3.25 2.75 -4.453125 2.203125 -5.5625 C 1.609375 -6.78125 0.765625 -7.4375 0.65625 -7.4375 C 0.609375 -7.4375 0.5625 -7.390625 0.5625 -7.328125 C 0.5625 -7.296875 0.5625 -7.28125 0.75 -7.109375 C 1.71875 -6.125 2.28125 -4.546875 2.28125 -2.484375 C 2.28125 -0.78125 1.921875 0.953125 0.6875 2.203125 C 0.5625 2.328125 0.5625 2.34375 0.5625 2.375 C 0.5625 2.4375 0.609375 2.484375 0.65625 2.484375 C 0.765625 2.484375 1.65625 1.796875 2.234375 0.546875 C 2.75 -0.546875 2.859375 -1.640625 2.859375 -2.484375 Z M 2.859375 -2.484375 "/>
</symbol>
<symbol overflow="visible" id="glyph2-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph2-1">
<path style="stroke:none;" d="M 3.28125 0 L 3.28125 -0.25 L 3.015625 -0.25 C 2.328125 -0.25 2.328125 -0.34375 2.328125 -0.5625 L 2.328125 -4.40625 C 2.328125 -4.59375 2.3125 -4.609375 2.109375 -4.609375 C 1.671875 -4.171875 1.046875 -4.15625 0.75 -4.15625 L 0.75 -3.90625 C 0.921875 -3.90625 1.375 -3.90625 1.765625 -4.109375 L 1.765625 -0.5625 C 1.765625 -0.34375 1.765625 -0.25 1.0625 -0.25 L 0.8125 -0.25 L 0.8125 0 L 2.046875 -0.03125 Z M 3.28125 0 "/>
</symbol>
<symbol overflow="visible" id="glyph2-2">
<path style="stroke:none;" d="M 3.5 -1.265625 L 3.265625 -1.265625 C 3.25 -1.109375 3.171875 -0.703125 3.09375 -0.625 C 3.03125 -0.59375 2.5 -0.59375 2.40625 -0.59375 L 1.125 -0.59375 C 1.859375 -1.234375 2.09375 -1.421875 2.515625 -1.75 C 3.03125 -2.171875 3.5 -2.59375 3.5 -3.25 C 3.5 -4.09375 2.765625 -4.609375 1.875 -4.609375 C 1.015625 -4.609375 0.4375 -4 0.4375 -3.359375 C 0.4375 -3.015625 0.734375 -2.96875 0.8125 -2.96875 C 0.96875 -2.96875 1.171875 -3.09375 1.171875 -3.34375 C 1.171875 -3.46875 1.125 -3.71875 0.765625 -3.71875 C 0.984375 -4.203125 1.453125 -4.359375 1.78125 -4.359375 C 2.46875 -4.359375 2.828125 -3.8125 2.828125 -3.25 C 2.828125 -2.65625 2.40625 -2.171875 2.171875 -1.921875 L 0.5 -0.265625 C 0.4375 -0.203125 0.4375 -0.1875 0.4375 0 L 3.296875 0 Z M 3.5 -1.265625 "/>
</symbol>
<symbol overflow="visible" id="glyph3-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph3-1">
<path style="stroke:none;" d="M 8.0625 -2.984375 C 8.0625 -3.203125 7.875 -3.203125 7.6875 -3.203125 L 4.515625 -3.203125 L 4.515625 -6.375 C 4.515625 -6.546875 4.515625 -6.75 4.3125 -6.75 C 4.09375 -6.75 4.09375 -6.5625 4.09375 -6.375 L 4.09375 -3.203125 L 0.921875 -3.203125 C 0.75 -3.203125 0.546875 -3.203125 0.546875 -3 C 0.546875 -2.78125 0.75 -2.78125 0.921875 -2.78125 L 4.09375 -2.78125 L 4.09375 0.390625 C 4.09375 0.5625 4.09375 0.765625 4.296875 0.765625 C 4.515625 0.765625 4.515625 0.578125 4.515625 0.390625 L 4.515625 -2.78125 L 7.6875 -2.78125 C 7.859375 -2.78125 8.0625 -2.78125 8.0625 -2.984375 Z M 8.0625 -2.984375 "/>
</symbol>
<symbol overflow="visible" id="glyph4-0">
<path style="stroke:none;" d=""/>
</symbol>
<symbol overflow="visible" id="glyph4-1">
<path style="stroke:none;" d="M 6.875 -2.484375 C 6.875 -2.671875 6.6875 -2.671875 6.546875 -2.671875 L 1.15625 -2.671875 C 1.015625 -2.671875 0.828125 -2.671875 0.828125 -2.484375 C 0.828125 -2.28125 1.015625 -2.28125 1.15625 -2.28125 L 6.546875 -2.28125 C 6.6875 -2.28125 6.875 -2.28125 6.875 -2.484375 Z M 6.875 -2.484375 "/>
</symbol>
</g>
<clipPath id="clip1">
<path d="M 134 78 L 155.132812 78 L 155.132812 111 L 134 111 Z M 134 78 "/>
</clipPath>
<clipPath id="clip2">
<path d="M 134 23 L 155.132812 23 L 155.132812 55 L 134 55 Z M 134 23 "/>
</clipPath>
<clipPath id="clip3">
<path d="M 134 57 L 155.132812 57 L 155.132812 89 L 134 89 Z M 134 57 "/>
</clipPath>
</defs>
<g id="surface1">
<path style="fill-rule:nonzero;fill:rgb(89.99939%,89.99939%,89.99939%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -56.693712 -42.52075 L 56.69394 -42.52075 L 56.69394 42.521954 L -56.693712 42.521954 Z M -56.693712 -42.52075 " transform="matrix(0.994032,0,0,-0.994032,77.566293,60.397083)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-1" x="67.526571" y="11.364474"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-1" x="75.260139" y="11.364474"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-2" x="79.111019" y="11.364474"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-2" x="83.753147" y="11.364474"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(100%,100%,100%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 3.910168 10.168709 L 36.84894 10.168709 L 36.84894 32.847025 L 3.910168 32.847025 Z M 3.910168 10.168709 " transform="matrix(0.994032,0,0,-0.994032,77.566293,60.397083)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-3" x="84.755131" y="41.492585"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-1" x="94.107977" y="42.978663"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-1" x="98.551299" y="41.492585"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-2" x="102.402179" y="41.492585"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-2" x="107.044308" y="41.492585"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(100%,100%,100%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 3.910168 -24.243701 L 36.84894 -24.243701 L 36.84894 -1.565385 L 3.910168 -1.565385 Z M 3.910168 -24.243701 " transform="matrix(0.994032,0,0,-0.994032,77.566293,60.397083)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-3" x="84.755131" y="75.701197"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-2" x="94.107977" y="77.186281"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-1" x="98.551299" y="75.701197"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-2" x="102.402179" y="75.701197"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph1-2" x="107.044308" y="75.701197"/>
</g>
<path style="fill-rule:nonzero;fill:rgb(100%,100%,100%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -13.592729 21.509832 C -13.592729 25.360941 -16.716843 28.485055 -20.567952 28.485055 C -24.419061 28.485055 -27.543175 25.360941 -27.543175 21.509832 C -27.543175 17.658723 -24.419061 14.534609 -20.567952 14.534609 C -16.716843 14.534609 -13.592729 17.658723 -13.592729 21.509832 Z M -13.592729 21.509832 " transform="matrix(0.994032,0,0,-0.994032,77.566293,60.397083)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph3-1" x="52.808936" y="41.915049"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph4-1" x="47.437188" y="53.70526"/>
</g>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -77.033855 21.509832 L -32.176295 21.509832 " transform="matrix(0.994032,0,0,-0.994032,77.566293,60.397083)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.052312 0.000641747 L 1.607818 1.682555 L 3.089316 0.000641747 L 1.607818 -1.681271 Z M 6.052312 0.000641747 " transform="matrix(0.994032,0,0,-0.994032,42.761153,39.016263)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-4" x="4.788251" y="35.220244"/>
</g>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -77.033855 -12.906508 L -1.218094 -12.906508 " transform="matrix(0.994032,0,0,-0.994032,77.566293,60.397083)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.054473 -0.00152773 L 1.609979 1.684315 L 3.087547 -0.00152773 L 1.609979 -1.683441 Z M 6.054473 -0.00152773 " transform="matrix(0.994032,0,0,-0.994032,73.532442,73.225044)"/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-5" x="4.788251" y="69.428856"/>
</g>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -20.567952 -12.906508 L -20.567952 9.901489 " transform="matrix(0.994032,0,0,-0.994032,77.566293,60.397083)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.052766 -0.000197973 L 1.608272 1.681715 L 3.08584 -0.000197973 L 1.608272 -1.682111 Z M 6.052766 -0.000197973 " transform="matrix(0,-0.994032,-0.994032,0,57.120897,53.376017)"/>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 59.101562 73.226562 C 59.101562 72.132812 58.214844 71.246094 57.121094 71.246094 C 56.027344 71.246094 55.140625 72.132812 55.140625 73.226562 C 55.140625 74.320312 56.027344 75.207031 57.121094 75.207031 C 58.214844 75.207031 59.101562 74.320312 59.101562 73.226562 Z M 59.101562 73.226562 "/>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -13.592729 21.509832 L -1.218094 21.509832 " transform="matrix(0.994032,0,0,-0.994032,77.566293,60.397083)"/>
<path style="fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.054473 0.000641747 L 1.609979 1.682555 L 3.087547 0.000641747 L 1.609979 -1.681271 Z M 6.054473 0.000641747 " transform="matrix(0.994032,0,0,-0.994032,73.532442,39.016263)"/>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M -44.55093 21.509832 L -44.55093 -34.413773 L 72.400963 -34.413773 " transform="matrix(0.994032,0,0,-0.994032,77.566293,60.397083)"/>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 152.730469 94.605469 L 148.3125 92.933594 L 149.78125 94.605469 L 148.3125 96.277344 Z M 152.730469 94.605469 "/>
<g clip-path="url(#clip1)" clip-rule="nonzero">
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.05311 0.000397378 L 1.608616 1.68231 L 3.086184 0.000397378 L 1.608616 -1.681516 Z M 6.05311 0.000397378 " transform="matrix(0.994032,0,0,-0.994032,146.713484,94.605864)"/>
</g>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 35.261719 39.015625 C 35.261719 37.921875 34.375 37.035156 33.28125 37.035156 C 32.1875 37.035156 31.300781 37.921875 31.300781 39.015625 C 31.300781 40.109375 32.1875 40.996094 33.28125 40.996094 C 34.375 40.996094 35.261719 40.109375 35.261719 39.015625 Z M 35.261719 39.015625 "/>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-6" x="145.188292" y="90.809487"/>
</g>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 37.348012 21.509832 L 72.400963 21.509832 " transform="matrix(0.994032,0,0,-0.994032,77.566293,60.397083)"/>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 152.730469 39.015625 L 148.3125 37.34375 L 149.78125 39.015625 L 148.3125 40.6875 Z M 152.730469 39.015625 "/>
<g clip-path="url(#clip2)" clip-rule="nonzero">
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.05311 0.000641747 L 1.608616 1.682555 L 3.086184 0.000641747 L 1.608616 -1.681271 Z M 6.05311 0.000641747 " transform="matrix(0.994032,0,0,-0.994032,146.713484,39.016263)"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-7" x="141.295663" y="33.735161"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-1" x="145.901012" y="35.220244"/>
</g>
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 37.348012 -12.906508 L 72.400963 -12.906508 " transform="matrix(0.994032,0,0,-0.994032,77.566293,60.397083)"/>
<path style=" stroke:none;fill-rule:nonzero;fill:rgb(0%,0%,0%);fill-opacity:1;" d="M 152.730469 73.226562 L 148.3125 71.550781 L 149.78125 73.226562 L 148.3125 74.898438 Z M 152.730469 73.226562 "/>
<g clip-path="url(#clip3)" clip-rule="nonzero">
<path style="fill:none;stroke-width:0.99628;stroke-linecap:butt;stroke-linejoin:miter;stroke:rgb(0%,0%,0%);stroke-opacity:1;stroke-miterlimit:10;" d="M 6.05311 -0.00152773 L 1.608616 1.684315 L 3.086184 -0.00152773 L 1.608616 -1.683441 Z M 6.05311 -0.00152773 " transform="matrix(0.994032,0,0,-0.994032,146.713484,73.225044)"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph0-7" x="141.295663" y="67.942779"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-2" x="145.901012" y="69.428856"/>
</g>
</g>
</svg>
Side by Side

After

Width:  |  Height:  |  Size: 25 KiB

BIN
journal/figs/hinf_filters_results.pdf
View File

Binary file not shown.

BIN
journal/figs/hinf_filters_results.png
View File

Binary file not shown.
Side by Side Swipe Overlay

Before

Width:  |  Height:  |  Size: 113 KiB

After

Width:  |  Height:  |  Size: 83 KiB

BIN
journal/figs/hinf_filters_results_mixed_sensitivity.pdf Normal file
View File

Binary file not shown.

BIN
journal/figs/hinf_filters_results_mixed_sensitivity.png Normal file
View File

Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 106 KiB

BIN
journal/figs/uncertainty_set_super_sensor.pdf
View File

Binary file not shown.

BIN
journal/figs/uncertainty_set_super_sensor.png
View File

Binary file not shown.
Side by Side Swipe Overlay

Before

Width:  |  Height:  |  Size: 26 KiB

After

Width:  |  Height:  |  Size: 27 KiB

14
journal/figs/uncertainty_set_super_sensor.svg
View File

@ -53,6 +53,15 @@
<symbol overflow="visible" id="glyph3-2">
<path style="stroke:none;" d="M 3.484375 -1.25 L 3.25 -1.25 C 3.234375 -1.109375 3.15625 -0.703125 3.078125 -0.625 C 3.015625 -0.59375 2.484375 -0.59375 2.390625 -0.59375 L 1.125 -0.59375 C 1.84375 -1.234375 2.078125 -1.421875 2.5 -1.75 C 3.015625 -2.15625 3.484375 -2.578125 3.484375 -3.234375 C 3.484375 -4.078125 2.75 -4.578125 1.875 -4.578125 C 1.015625 -4.578125 0.4375 -3.984375 0.4375 -3.34375 C 0.4375 -3 0.734375 -2.96875 0.796875 -2.96875 C 0.96875 -2.96875 1.171875 -3.078125 1.171875 -3.328125 C 1.171875 -3.453125 1.125 -3.6875 0.765625 -3.6875 C 0.96875 -4.1875 1.4375 -4.34375 1.765625 -4.34375 C 2.453125 -4.34375 2.8125 -3.796875 2.8125 -3.234375 C 2.8125 -2.640625 2.390625 -2.15625 2.171875 -1.90625 L 0.5 -0.265625 C 0.4375 -0.203125 0.4375 -0.1875 0.4375 0 L 3.28125 0 Z M 3.484375 -1.25 "/>
</symbol>
<symbol overflow="visible" id="glyph3-3">
<path style="stroke:none;" d="M 6.1875 0 L 6.1875 -0.25 C 5.75 -0.25 5.671875 -0.25 5.671875 -0.546875 L 5.671875 -2.09375 C 5.671875 -2.6875 5.375 -3.046875 4.640625 -3.046875 C 4.09375 -3.046875 3.71875 -2.75 3.53125 -2.40625 C 3.390625 -2.890625 3.03125 -3.046875 2.515625 -3.046875 C 1.953125 -3.046875 1.59375 -2.734375 1.390625 -2.375 L 1.390625 -3.046875 L 0.375 -2.96875 L 0.375 -2.71875 C 0.828125 -2.71875 0.890625 -2.671875 0.890625 -2.328125 L 0.890625 -0.546875 C 0.890625 -0.25 0.828125 -0.25 0.375 -0.25 L 0.375 0 C 0.375 0 0.875 -0.03125 1.15625 -0.03125 C 1.421875 -0.03125 1.890625 0 1.953125 0 L 1.953125 -0.25 C 1.5 -0.25 1.4375 -0.25 1.4375 -0.546875 L 1.4375 -1.78125 C 1.4375 -2.515625 2.015625 -2.859375 2.46875 -2.859375 C 2.953125 -2.859375 3.015625 -2.46875 3.015625 -2.125 L 3.015625 -0.546875 C 3.015625 -0.25 2.9375 -0.25 2.5 -0.25 L 2.5 0 C 2.5 0 2.984375 -0.03125 3.28125 -0.03125 C 3.53125 -0.03125 4.015625 0 4.078125 0 L 4.078125 -0.25 C 3.625 -0.25 3.5625 -0.25 3.5625 -0.546875 L 3.5625 -1.78125 C 3.5625 -2.515625 4.125 -2.859375 4.578125 -2.859375 C 5.0625 -2.859375 5.125 -2.46875 5.125 -2.125 L 5.125 -0.546875 C 5.125 -0.25 5.0625 -0.25 4.609375 -0.25 L 4.609375 0 C 4.609375 0 5.109375 -0.03125 5.40625 -0.03125 C 5.65625 -0.03125 6.125 0 6.1875 0 Z M 6.1875 0 "/>
</symbol>
<symbol overflow="visible" id="glyph3-4">
<path style="stroke:none;" d="M 3.75 -0.625 L 3.75 -1 L 3.515625 -1 L 3.515625 -0.625 C 3.515625 -0.5625 3.515625 -0.234375 3.265625 -0.234375 C 3.015625 -0.234375 3.015625 -0.546875 3.015625 -0.640625 L 3.015625 -1.84375 C 3.015625 -2.21875 3.015625 -2.484375 2.6875 -2.734375 C 2.40625 -2.96875 2.078125 -3.078125 1.671875 -3.078125 C 1.015625 -3.078125 0.546875 -2.828125 0.546875 -2.40625 C 0.546875 -2.1875 0.703125 -2.078125 0.890625 -2.078125 C 1.078125 -2.078125 1.203125 -2.203125 1.203125 -2.390625 C 1.203125 -2.515625 1.15625 -2.65625 0.96875 -2.71875 C 1.21875 -2.890625 1.609375 -2.890625 1.65625 -2.890625 C 2.046875 -2.890625 2.46875 -2.625 2.46875 -2.046875 L 2.46875 -1.84375 C 2.078125 -1.828125 1.625 -1.796875 1.125 -1.609375 C 0.5 -1.390625 0.3125 -1.015625 0.3125 -0.703125 C 0.3125 -0.109375 1.03125 0.0625 1.515625 0.0625 C 2.078125 0.0625 2.390625 -0.234375 2.546875 -0.515625 C 2.5625 -0.234375 2.75 0.03125 3.078125 0.03125 C 3.078125 0.03125 3.75 0.03125 3.75 -0.625 Z M 2.46875 -0.96875 C 2.46875 -0.3125 1.890625 -0.125 1.578125 -0.125 C 1.21875 -0.125 0.890625 -0.359375 0.890625 -0.703125 C 0.890625 -1.078125 1.21875 -1.609375 2.46875 -1.65625 Z M 2.46875 -0.96875 "/>
</symbol>
<symbol overflow="visible" id="glyph3-5">
<path style="stroke:none;" d="M 3.96875 0 L 3.96875 -0.25 C 3.5 -0.25 3.453125 -0.296875 3.28125 -0.46875 L 2.34375 -1.546875 L 2.28125 -1.609375 C 2.28125 -1.640625 2.734375 -2.125 2.796875 -2.1875 C 3.046875 -2.453125 3.28125 -2.71875 3.828125 -2.734375 L 3.828125 -2.984375 C 3.640625 -2.96875 3.4375 -2.953125 3.25 -2.953125 C 3.046875 -2.953125 2.78125 -2.953125 2.5625 -2.984375 L 2.5625 -2.734375 C 2.6875 -2.703125 2.703125 -2.625 2.703125 -2.5625 C 2.703125 -2.5625 2.703125 -2.453125 2.609375 -2.328125 L 2.125 -1.78125 L 1.546875 -2.4375 C 1.484375 -2.515625 1.46875 -2.5625 1.46875 -2.578125 C 1.46875 -2.6875 1.5625 -2.734375 1.65625 -2.734375 L 1.65625 -2.984375 C 1.40625 -2.953125 1.140625 -2.953125 0.890625 -2.953125 C 0.671875 -2.953125 0.421875 -2.953125 0.21875 -2.984375 L 0.21875 -2.734375 C 0.53125 -2.734375 0.71875 -2.734375 0.890625 -2.515625 L 1.75 -1.53125 L 1.8125 -1.453125 C 1.8125 -1.4375 1.296875 -0.875 1.21875 -0.796875 C 0.96875 -0.515625 0.734375 -0.265625 0.171875 -0.25 L 0.171875 0 C 0.390625 -0.015625 0.53125 -0.03125 0.75 -0.03125 C 0.953125 -0.03125 1.234375 -0.015625 1.4375 0 L 1.4375 -0.25 C 1.359375 -0.265625 1.3125 -0.3125 1.3125 -0.40625 C 1.3125 -0.53125 1.375 -0.625 1.484375 -0.75 L 1.96875 -1.28125 L 2.578125 -0.59375 C 2.703125 -0.453125 2.703125 -0.4375 2.703125 -0.390625 C 2.703125 -0.25 2.546875 -0.25 2.515625 -0.25 L 2.515625 0 C 2.578125 0 3.078125 -0.03125 3.296875 -0.03125 C 3.515625 -0.03125 3.75 -0.015625 3.96875 0 Z M 3.96875 0 "/>
</symbol>
</g>
<clipPath id="clip1">
<path d="M 68 0 L 198 0 L 198 103.097656 L 68 103.097656 Z M 68 0 "/>
@ -187,5 +196,10 @@
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph2-3" x="59.603297" y="47.328129"/>
</g>
<g style="fill:rgb(0%,0%,0%);fill-opacity:1;">
<use xlink:href="#glyph3-3" x="65.485251" y="48.808527"/>
<use xlink:href="#glyph3-4" x="71.970573" y="48.808527"/>
<use xlink:href="#glyph3-5" x="75.905301" y="48.808527"/>
</g>
</g>
</svg>

Side by Side Swipe Overlay

Before

Width:  |  Height:  |  Size: 27 KiB

After

Width:  |  Height:  |  Size: 31 KiB

BIN
journal/figs/weight_formula.pdf
View File

Binary file not shown.

BIN
journal/figs/weight_formula.png
View File

Binary file not shown.
Side by Side Swipe Overlay

Before

Width:  |  Height:  |  Size: 77 KiB

After

Width:  |  Height:  |  Size: 76 KiB

BIN
journal/figs/weights_W1_W2.pdf
View File

Binary file not shown.

BIN
journal/figs/weights_W1_W2.png
View File

Binary file not shown.
Side by Side Swipe Overlay

Before

Width:  |  Height:  |  Size: 78 KiB

After

Width:  |  Height:  |  Size: 58 KiB

42
journal/ref.bib
View File

@ -75,11 +75,10 @@
}
@article{zimmermann92_high_bandw_orien_measur_contr,
author = {M. Zimmermann and W. Sulzer},
author = {Zimmermann, M and Sulzer, W},
title = {High Bandwidth Orientation Measurement and Control Based on
Complementary Filtering},
journal = {Robot Control 1991},
pages = {525-530},
journal = {Robot Control},
year = 1992,
doi = {10.1016/B978-0-08-041276-4.50093-5},
publisher = {Elsevier},
@ -254,7 +253,6 @@
booktitle = {Proceedings of IEEE International Conference on Intelligent
Engineering Systems},
year = 1997,
pages = {nil},
doi = {10.1109/ines.1997.632450},
month = {-},
}
@ -319,7 +317,6 @@
booktitle = {Proceedings of the 44th IEEE Conference on Decision and
Control},
year = 2005,
pages = {nil},
doi = {10.1109/cdc.2005.1582367},
month = {-},
}
@ -354,3 +351,38 @@
large scale gravitational wave detectors},
year = 2016,
}
@article{yeh05_model_contr_hydraul_actuat_two,
author = {T-J Yeh and C-Y Su and W-J Wang},
title = {Modelling and Control of a Hydraulically Actuated
Two-Degree-Of-Freedom Inertial Platform},
journal = {Proceedings of the Institution of Mechanical Engineers,
Part I: Journal of Systems and Control Engineering},
volume = 219,
number = 6,
pages = {405-417},
year = 2005,
doi = {10.1243/095965105x33527},
}
@inproceedings{dehaeze19_compl_filter_shapin_using_synth,
author = {Dehaeze, Thomas and Vermat, Mohit and Collette, Christophe},
title = {Complementary Filters Shaping Using $\mathcal{H}_\infty$
Synthesis},
booktitle = {7th International Conference on Control, Mechatronics and
Automation (ICCMA)},
year = 2019,
pages = {459-464},
doi = {10.1109/ICCMA46720.2019.8988642},
language = {english},
}
@misc{dehaeze21_new_method_desig_compl_filter_code,
author = {Thomas Dehaeze},
doi = {10.5281/zenodo.3894342},
howpublished = {Source Code on Zonodo},
month = 09,
title = {A New Method of Designing Complementary Filters for Sensor
Fusion using $\mathcal{H}_\infty$ Synthesis},
year = 2021,
}