<< computing_experiment computing_experiment AdvancedModelLearningDataTimeGrid >>
computing_experiment >> computing_experiment > AdvancedModelLearningDataEvidence

AdvancedModelLearningDataEvidence

Implements a computational experiment of a user system, specified in the form of an xcos-diagram, with the calculation of the parameters of the output dynamics of the system

Syntax

[SystemOutputChatacteristics..
]=AdvancedModelLearningDataEvidence(SystemOutputCharacteristicsNames,..
SystemOutputCharacteristicsIndexes,..
SystemShemaName,SystemOutputCharacteristicsCompressionFunctions,..
SystemParametrsNames,SystemParametrsValues[,absolutTolErr[,..
relativeTolErr[,simulationTime[,timeStepMax[,Solver[,toleranceTime]]]]]])

Arguments

SystemOutputChatacteristicsNames

Array of names of system output characteristics in xcos-diagram (see block TOWS_c). Array dimension 1xM

SystemOutputCharacteristicsIndexes

An array of indices of the dynamics of the output characteristics in the names of the output characteristics of the system in the xcos diagram (block TOWS_c). Array dimension 1xM

SystemShemaName

Full name of the user system xcos diagram file (extension .zcos, .xcos)

SystemOutputCharacteristicsCompressionFunctions

Cell array (see cell arrays) functions or string array of names of functions for getting the parameters of computed dynamics, specified by SystemOutputCharacteristicsIndexes

SystemParametrsNames

The names of the system parameters in the xcos diagram (see the context property in Xcos diagram structure). Array of strings of dimension 1xM

SystemParametrsValues

System parameter values in the xcos diagram (see the context property in Xcos diagram structure), specified by the corresponding SystemParametrsNames. NxM array of real numbers (each n-th row of N rows in the matrix is the corresponding n-th dynamics)

absolutTolErr

Absolute errors of integration in xcos (see the tol (1) atol property in Xcos diagram structure) of each dynamics. Array of positive real numbers of dimension 1xN

relativeTolErr

Relative errors of integration in xcos (see the tol (2) rtol property in Xcos diagram structure) of each dynamics. Array of positive real numbers of dimension 1xN

simulationTime

Simulation times in xcos (see the tf property in Xcos diagram structure) of each dynamics. Array of positive real numbers of dimension 1xN

timeStepMax

Maximum time integration steps in xcos (see the tol (7) hmax property in Xcos diagram structure) of each dynamics. Array of positive real numbers of dimension 1xN

Solver

Solver views in xcos (see the tol (6) solver property in Xcos diagram structure) of each dynamics. An array of integers ranging from 0 to 100 - solver codes. The dimension of the solver codes is 1xN

toleranceTime

Time errors in xcos are the corresponding minimum steps of integration over time (see the tol (3) ttol property in Xcos diagram structure) for each dynamics. Array of positive real numbers of dimension 1xN

Return values

SystemOutputChatacteristics

Matrix of parameters of system output characteristics. Dimension of matrix PxN

Description

Simulates (computational experiment) a user system, specified in the form of an xcos diagram, for its various parameters (specified in the context).

Returns a matrix of speaker parameters with user-defined output characteristics of the system (in the diagram, these are given as TOWS_c blocks)

All graphic windows open during the simulation are automatically deleted.

Examples

//Forming the path to the xcos-diagram Proba3.zcos
[a, pathlib] = libraryinfo("computing_experimentlib");
clear a;
pathlib = strncpy(pathlib, length(pathlib) - length('macros')) + 'xcos';
pathdiagram = fullfile(pathlib, 'Proba3.zcos');

//Filter functions
function [parametrs]=DynamicFilterFunction1(time, value)
    parametrs=[sum(time.*value) / length(time), sum((time.^2).*value) / (3*length(time))];
endfunction
function [parametr]=DynamicFilterFunction2(time, value)
    parametr=sum(time.*value) / (2*length(time));
endfunction

//Array of filter functions
DynamicFiltersFunctions = cell(1, 4);
DynamicFiltersFunctions{1} = DynamicFilterFunction1;
DynamicFiltersFunctions{2} = DynamicFilterFunction2;
DynamicFiltersFunctions{3} = DynamicFilterFunction2;
DynamicFiltersFunctions{4} = DynamicFilterFunction1;

//Running the basic function
SystemParametrsNames = ['b', 'a', 'd'];//Parameter names
SystemParametrsValues = [ 1.5, 0.81, 3.81;
                         2.25, 0.51, 6.81;
                         1.11, 0.93, 3.51];//Parameter values
[SystemOutputCharacteristics...//Output characteristics
 ]=AdvancedModelLearningDataEvidence(['A3', 'A3', 'A2', 'A1'],...//System output characteristic names
                                     [3, 2, 1, 2],...//Dynamics indices (in dynamics names)
                                     fullfile(pathdiagram),...//System diagram name
                                     DynamicFiltersFunctions,...//Speaker compression function
                                     SystemParametrsNames,...//Parameter names that define dynamics
                                     SystemParametrsValues,...//The values of the parameters that determine the dynamics
                                     [0.5E-6,0.51E-6,0.57E-6],..//Absolute errors
                                     [0.69E-6,0.57E-6,0.75E-6],..//Relative errors
                                     [10.0,11.1,15.3],..//Simulation times
                                     [0.93E-4,0.87E-4,0.81E-4],..//Maximum time integration steps
                                     [0,3,4],..//Solvers type
                                     [2.1E-10,1.5E-10,1.23E-10]..//Time errors
                                     );

//Result
disp(SystemOutputCharacteristics);

Examples

//Forming the path to the xcos-diagram Proba3.zcos
[a, pathlib] = libraryinfo("computing_experimentlib");
clear a;
pathlib = strncpy(pathlib, length(pathlib) - length('macros')) + 'xcos';
pathdiagram = fullfile(pathlib, 'Proba3.zcos');

//Filter functions
function [parametrs]=DynamicFilterFunction1(time, value)
    parametrs=[sum(time.*value) / length(time), sum((time.^2).*value) / (3*length(time))];
endfunction
function [parametr]=DynamicFilterFunction2(time, value)
    parametr=sum(time.*value) / (2*length(time));
endfunction

//Running the basic function
SystemParametrsNames = ['b', 'a', 'd'];//Parameter names
SystemParametrsValues = [ 1.5, 0.81, 3.81;
                         2.25, 0.51, 6.81;
                         1.11, 0.93, 3.51];//Parameter values
[SystemOutputCharacteristics...//Output characteristics
 ]=AdvancedModelLearningDataEvidence(['A3', 'A3', 'A2', 'A1'],..//System output characteristic names
                                     [3, 2, 1, 2],...//Dynamics indices (in dynamics names)
                                     fullfile(pathdiagram),..//System diagram name
                                     ['DynamicFilterFunction1',..
                                      'DynamicFilterFunction2',..
                                      'DynamicFilterFunction2',..
                                      'DynamicFilterFunction1'],...//Speaker compression function
                                     SystemParametrsNames,..//Parameter names that define dynamics
                                     SystemParametrsValues,..//The values of the parameters that determine the dynamics
                                     [0.5E-6,0.51E-6,0.57E-6],..//Absolute errors
                                     [0.69E-6,0.57E-6,0.75E-6],..//Relative errors
                                     [10.0,11.1,15.3],..//Simulation times
                                     [0.93E-4,0.87E-4,0.81E-4],..//Maximum time integration steps
                                     [0,3,4],..//Solvers type
                                     [2.1E-10,1.5E-10,1.23E-10]..//Time errors
                                     );

//Result
disp(SystemOutputCharacteristics);

See also


Report an issue
<< computing_experiment computing_experiment AdvancedModelLearningDataTimeGrid >>