Add noise budget to determine max noise of sensors

This commit is contained in:
Thomas Dehaeze 2020-09-01 13:48:19 +02:00
parent d24cddcc44
commit 45974748cf

242
org/noise_budgeting.org Normal file
View File

@ -0,0 +1,242 @@
#+TITLE: Noise Budgeting
#+SETUPFILE: ./setup/org-setup-file.org
* Maximum Noise of the Relative Motion Sensors
** Matlab Init :noexport:ignore:
#+begin_src matlab :tangle no :exports none :results silent :noweb yes :var current_dir=(file-name-directory buffer-file-name)
<<matlab-dir>>
#+end_src
#+begin_src matlab :exports none :results silent :noweb yes
<<matlab-init>>
#+end_src
#+BEGIN_SRC matlab
simulinkproject('../');
#+END_SRC
** Initialization
#+begin_src matlab
open('nass_model.slx');
#+end_src
#+begin_src matlab
initializeGround();
initializeGranite();
initializeTy();
initializeRy();
initializeRz();
initializeMicroHexapod();
initializeAxisc();
initializeMirror();
initializeSimscapeConfiguration();
initializeDisturbances('enable', false);
initializeLoggingConfiguration('log', 'none');
initializeController('type', 'hac-dvf');
#+end_src
We set the stiffness of the payload fixation:
#+begin_src matlab
Kp = 1e8; % [N/m]
#+end_src
#+begin_src matlab
initializeNanoHexapod('k', 1e5, 'c', 2e2);
Ms = 50;
initializeSample('mass', Ms, 'freq', sqrt(Kp/Ms)/2/pi*ones(6,1));
#+end_src
#+begin_src matlab
initializeReferences('Rz_type', 'rotating-not-filtered', 'Rz_period', Ms);
#+end_src
** Control System
#+begin_src matlab
Kdvf = 5e3*s/(1+s/2/pi/1e3)*eye(6);
#+end_src
#+begin_src matlab
h = 2.0;
Kl = 2e7 * eye(6) * ...
1/h*(s/(2*pi*100/h) + 1)/(s/(2*pi*100*h) + 1) * ...
1/h*(s/(2*pi*200/h) + 1)/(s/(2*pi*200*h) + 1) * ...
(s/2/pi/10 + 1)/(s/2/pi/10) * ...
1/(1 + s/2/pi/300);
#+end_src
#+begin_src matlab
load('mat/stages.mat', 'nano_hexapod');
K = Kl*nano_hexapod.kinematics.J*diag([1, 1, 1, 1, 1, 0]);
#+end_src
#+begin_src matlab :exports none
%% Name of the Simulink File
mdl = 'nass_model';
%% Micro-Hexapod
clear io; io_i = 1;
io(io_i) = linio([mdl, '/Noises'], 1, 'openinput', [], 'ndL'); io_i = io_i + 1;
io(io_i) = linio([mdl, '/Tracking Error'], 1, 'output', [], 'En'); io_i = io_i + 1;
#+end_src
#+begin_src matlab
%% Run the linearization
G = linearize(mdl, io);
G.InputName = {'ndL1', 'ndL2', 'ndL3', 'ndL4', 'ndL5', 'ndL6'};
G.OutputName = {'Ex', 'Ey', 'Ez', 'Erx', 'Ery', 'Erz'};
#+end_src
#+begin_src matlab :exports none
freqs = logspace(0, 3, 1000);
figure;
hold on;
plot(freqs, abs(squeeze(freqresp(G(1, 1), freqs, 'Hz'))));
plot(freqs, abs(squeeze(freqresp(G(2, 2), freqs, 'Hz'))));
plot(freqs, abs(squeeze(freqresp(G(3, 3), freqs, 'Hz'))));
plot(freqs, abs(squeeze(freqresp(G(4, 4), freqs, 'Hz'))));
plot(freqs, abs(squeeze(freqresp(G(6, 5), freqs, 'Hz'))));
plot(freqs, abs(squeeze(freqresp(G(6, 6), freqs, 'Hz'))));
hold off;
set(gca, 'XScale', 'log'); set(gca, 'YScale', 'log');
ylabel('Amplitude [m/m]'); set(gca, 'XTickLabel',[]);
#+end_src
** Maximum induced vibration's ASD
Required maximum induced ASD of the sample's vibration due to the relative motion sensor noise.
\[ \bm{\Gamma}_x(\omega) = \begin{bmatrix} \Gamma_x(\omega) & \Gamma_y(\omega) & \Gamma_{R_x}(\omega) & \Gamma_{R_y}(\omega) \end{bmatrix} \]
#+begin_src matlab
Gamma_x = [(1e-9)/(1 + s/2/pi/100); % Dx
(1e-9)/(1 + s/2/pi/100); % Dy
(1e-9)/(1 + s/2/pi/100); % Dz
(2e-8)/(1 + s/2/pi/100); % Rx
(2e-8)/(1 + s/2/pi/100)]; % Ry
#+end_src
#+begin_src matlab
freqs = logspace(0, 3, 1000);
#+end_src
Corresponding RMS value in [nm rms, nrad rms]
#+begin_src matlab :exports results :results value table replace :tangle no :post addhdr(*this*)
data2orgtable([1e9*sqrt(trapz(freqs, (abs(squeeze(freqresp(Gamma_x, freqs, 'Hz')))').^2))]', {'Dx [nm]', 'Dy [nm]', 'Dz [nm]', 'Rx [nrad]', 'Ry [nrad]'}, {'Specifications'}, ' %.1f ');
#+end_src
#+RESULTS:
| | Specifications |
|-----------+----------------|
| Dx [nm] | 12.1 |
| Dy [nm] | 12.1 |
| Dz [nm] | 12.1 |
| Rx [nrad] | 241.8 |
| Ry [nrad] | 241.8 |
** Computation of the maximum relative motion sensor noise
Let's note $G$ the transfer function from the 6 sensor noise $n$ to the 5dof pose error $x$.
We have:
\[ x_i = \sum_{j=1}^6 G_{ij}(s) n_j, \quad i = 1 \dots 5 \]
In terms of ASD:
\[ \Gamma_{x_i}(\omega) = \sqrt{\sum_{j=1}^6 |G_{ij}(j\omega)|^2 \cdot {\Gamma_{n_j}}^2(\omega)}, \quad i = 1 \dots 5 \]
Let's suppose that the ASD of all the sensor noise are equal:
\[ \Gamma_{n_j} = \Gamma_{n}, \quad j = 1 \dots 6 \]
We then have an upper bound of the sensor noise for each of the considered motion errors:
\[ \Gamma_{n_i, \text{max}}(\omega) = \frac{\Gamma_{x_i}(\omega)}{\sqrt{\sum_{j=1}^6 |G_{ij}(j\omega)|^2}}, \quad i = 1 \dots 5 \]
#+begin_src matlab
Gamma_ndL = zeros(5, length(freqs));
for in = 1:5
Gamma_ndL(in, :) = abs(squeeze(freqresp(Gamma_x(in), freqs, 'Hz')))./sqrt(sum(abs(squeeze(freqresp(G(in, :), freqs, 'Hz'))).^2))';
end
#+end_src
#+begin_src matlab :exports none
figure;
hold on;
for in = 1:5
plot(freqs, Gamma_ndL(in, :), 'k-');
end
hold off;
set(gca, 'XScale', 'log'); set(gca, 'YScale', 'log');
xlabel('Frequency [Hz]'); ylabel('ASD [$\frac{m}{\sqrt{Hz}}$]');
#+end_src
#+begin_src matlab :tangle no :exports results :results file replace
exportFig('figs/noise_budget_ndL_max_asd.pdf', 'width', 'wide', 'height', 'normal');
#+end_src
#+name: fig:noise_budget_ndL_max_asd
#+caption: Maximum estimated ASD of the relative motion sensor noise
#+RESULTS:
[[file:figs/noise_budget_ndL_max_asd.png]]
If the noise ASD of the relative motion sensor is bellow the maximum specified ASD for all the considered motion:
\[ \Gamma_n < \Gamma_{n_i, \text{max}}, \quad i = 1 \dots 5 \]
Then, the motion error due to sensor noise should be bellow the one specified.
#+begin_src matlab
Gamma_ndL_max = min(Gamma_ndL(1:5, :));
#+end_src
Let's take a sensor with a white noise up to 1kHz that is bellow the specified one:
#+begin_src matlab
Gamma_ndL_ex = abs(squeeze(freqresp(min(Gamma_ndL_max)/(1 + s/2/pi/1e3), freqs, 'Hz')));
#+end_src
#+begin_src matlab :exports none
figure;
hold on;
plot(freqs, Gamma_ndL_max, 'k-', 'DisplayName', 'Specification');
plot(freqs, Gamma_ndL_ex, 'DisplayName', 'Sensor Example');
hold off;
set(gca, 'XScale', 'log'); set(gca, 'YScale', 'log');
xlabel('Frequency [Hz]'); ylabel('ASD [m/sqrt(Hz)]');
#+end_src
#+begin_src matlab :tangle no :exports results :results file replace
exportFig('figs/relative_motion_sensor_noise_ASD_example.pdf', 'width', 'wide', 'height', 'normal');
#+end_src
#+name: fig:relative_motion_sensor_noise_ASD_example
#+caption: Requirement maximum ASD of the sensor noise + example of a sensor validating the requirements
#+RESULTS:
[[file:figs/relative_motion_sensor_noise_ASD_example.png]]
The corresponding RMS value of the sensor noise taken as an example is [nm RMS]:
#+begin_src matlab :results replace value
1e9*sqrt(trapz(freqs, Gamma_ndL_max.^2))
#+end_src
#+RESULTS:
: 519.29
** Verification of the induced motion error
Verify that by taking the sensor noise, we have to wanted displacement error
From the sensor noise PSD $\Gamma_n(\omega)$, we can estimate the obtained displacement PSD $\Gamma_x(\omega)$:
\[ \Gamma_{x,i}(\omega) = \sqrt{ \sum_{j=1}^{6} |G_{ij}|^2(j\omega) \cdot \Gamma_{n,j}^2(\omega) }, \quad i = 1 \dots 5 \]
#+begin_src matlab
Gamma_xest = zeros(5, length(freqs));
for in = 1:5
Gamma_xest(in, :) = sqrt(sum(abs(squeeze(freqresp(G(in, :), freqs, 'Hz'))).^2.*Gamma_ndL_max.^2));
end
#+end_src
#+begin_src matlab :exports results :results value table replace :tangle no :post addhdr(*this*)
data2orgtable([1e9*sqrt(trapz(freqs, (Gamma_xest.^2)')); 1e9*sqrt(trapz(freqs, (abs(squeeze(freqresp(Gamma_x, freqs, 'Hz')))').^2))]', {'Dx [nm]', 'Dy [nm]', 'Dz [nm]', 'Rx [nrad]', 'Ry [nrad]'}, {'Results', 'Specifications'}, ' %.1f ');
#+end_src
#+RESULTS:
| | Results | Specifications |
|-----------+---------+----------------|
| Dx [nm] | 8.9 | 12.1 |
| Dy [nm] | 9.3 | 12.1 |
| Dz [nm] | 10.2 | 12.1 |
| Rx [nrad] | 110.2 | 241.8 |
| Ry [nrad] | 107.8 | 241.8 |