How is genetic algorithm function's ga optimfunctions used? How do I use that? What is the difference between optimset and optimfunction? Why do I receive error in this code?

% Genetic Algorithm to tune PID parameters

nos_var= 3;

lb = [0 0 0];

ub= [200 200 200];

% GA optionsga_opt = optimoption('ga','Display', 'off', 'Generations', 25, 'PopulationSize',50,'PlotFcns',@gaplotbestf);---- this line has error

obj_fcn= @(k) optimizationofPID(k);

[k,best]= ga((obj_fcn),no_var,[],[],[],[],lb,ub,[],ga_opt);

Output:

Error in GAlgorithmPID (line 9)

ga_opt = optimoption('ga','Display', 'off', 'Generations', 25, 'PopulationSize',50,'PlotFcns',@gaplotbestf);

>> GAlgorithmPID

Unrecognized function or variable 'optimoption'.

Error in GAlgorithmPID (line 9)

ga_opt = optimoption('ga','Display', 'off', 'Generations', 25, 'PopulationSize',50,'PlotFcns',@gaplotbestf);

>>

[Sam Chak]

Two questions:

1. Can you show the math model of the system?

2. How does your cost function in obj_fcn look like? This function usually tells about the desired performance like fast/slow output convergence, and high/low control effort.

The searching range [0 200] for each control gain is also relatively large. Probably most of the iterations lead to instability. Having a model allows you to run some tests and analysis so that the range can be tightened, especially with Ki and Kd. Ideally, the ranges should be set up such that GA searches within the stability regime.

Answer 1

[Alan Weiss]

The correct function name is optimoptions:

ga_opt = optimoptions('ga','Display', 'off', 'Generations', 25, 'PopulationSize',50,'PlotFcns',@gaplotbestf)

ga_opt = 
  ga options:

   Set properties:
                         Display: 'off'
                  MaxGenerations: 25
                         PlotFcn: @gaplotbestf
                  PopulationSize: 50

   Default properties:
             ConstraintTolerance: 1.0000e-03
                     CreationFcn: []
                    CrossoverFcn: []
               CrossoverFraction: 0.8000
                      EliteCount: '0.05*PopulationSize'
                    FitnessLimit: -Inf
               FitnessScalingFcn: @fitscalingrank
               FunctionTolerance: 1.0000e-06
                       HybridFcn: []
         InitialPopulationMatrix: []
          InitialPopulationRange: []
             InitialScoresMatrix: []
             MaxStallGenerations: 50
                    MaxStallTime: Inf
                         MaxTime: Inf
                     MutationFcn: []
    NonlinearConstraintAlgorithm: 'auglag'
                       OutputFcn: []
                  PopulationType: 'doubleVector'
                    SelectionFcn: []
                     UseParallel: 0
                   UseVectorized: 0

Alan Weiss

MATLAB mathematical toolbox documentation