44 lines
2.4 KiB
Mathematica
44 lines
2.4 KiB
Mathematica
|
function results_ANN = run_ANN(robot, benchmarkSettings, experimentDataStruct, optionsANN, progressBar)
|
||
|
|
|
||
|
|
% Authors: Quentin Leboutet, Julien Roux, Alexandre Janot and Gordon Cheng
|
||
|
|
|
||
|
|
%% Define result data structure:
|
||
|
|
|
||
|
|
if benchmarkSettings.displayProgression == true
|
||
|
|
waitbar(0, progressBar, sprintf('ANN: First Iteration...'));
|
||
|
|
end
|
||
|
|
|
||
|
|
results_ANN.benchmarkSettings = benchmarkSettings; % Data structure containing the benchmark settings
|
||
|
|
results_ANN.options = optionsANN; % Options that are specific to the identification method.
|
||
|
|
results_ANN.Betas = zeros(benchmarkSettings.numberOfInitialEstimates, benchmarkSettings.numberOfExperimentsPerInitialPoint, numel(benchmarkSettings.Beta_obj)); % Identified parameters
|
||
|
|
results_ANN.times = zeros(benchmarkSettings.numberOfInitialEstimates, benchmarkSettings.numberOfExperimentsPerInitialPoint); % Computation times
|
||
|
|
|
||
|
|
%% Run the identification code:
|
||
|
|
|
||
|
|
for initEst = 1:benchmarkSettings.numberOfInitialEstimates % For each set of initial parameters
|
||
|
|
|
||
|
|
Beta_0 = benchmarkSettings.Initial_Beta(:,initEst);
|
||
|
|
|
||
|
|
for expNb = 1:benchmarkSettings.numberOfExperimentsPerInitialPoint % For each trajectory noise
|
||
|
|
tic
|
||
|
|
[Beta_ANN] = ANN_identification(robot, experimentDataStruct, expNb, benchmarkSettings, Beta_0, optionsANN);
|
||
|
|
results_ANN.times(initEst, expNb)=toc;
|
||
|
|
|
||
|
|
if benchmarkSettings.displayProgression == true
|
||
|
|
waitbar(((initEst-1)*benchmarkSettings.numberOfExperimentsPerInitialPoint+expNb)/((benchmarkSettings.numberOfInitialEstimates-1)*benchmarkSettings.numberOfExperimentsPerInitialPoint+benchmarkSettings.numberOfExperimentsPerInitialPoint), progressBar, sprintf('Adaline NN: %d%% done...', floor(100*((initEst-1)*benchmarkSettings.numberOfExperimentsPerInitialPoint+expNb)/((benchmarkSettings.numberOfInitialEstimates-1)*benchmarkSettings.numberOfExperimentsPerInitialPoint+benchmarkSettings.numberOfExperimentsPerInitialPoint))));
|
||
|
|
end
|
||
|
|
|
||
|
|
if optionsANN.verbose == true
|
||
|
|
fprintf('ANN status: initial estimate %d, experiment %d \n', initEst, expNb );
|
||
|
|
fprintf('ANN status: initial parameter error = %d\n', norm(Beta_0-benchmarkSettings.Beta_obj));
|
||
|
|
fprintf('ANN status: estimated parameter error = %d\n', norm(Beta_ANN-benchmarkSettings.Beta_obj));
|
||
|
|
disp('---------------------------------------------')
|
||
|
|
end
|
||
|
|
|
||
|
|
results_ANN.Betas(initEst,expNb,:) = Beta_ANN;
|
||
|
|
end
|
||
|
|
end
|
||
|
|
|
||
|
|
end
|
||
|
|
|