Can anyone help me with the PID tuning using #Particle Swarm Optomization for Load Frequency Control of 3 area system??

PLEASE HELP ME FOR TUNING THE PID CONTROLLER PARAMETERS USING PARTICLE SWARM OPTIMIZATION ALGORITHM FOR A LOAD FREQUENCY CONTROL PROBLEM HAVING THREE AREAS.ACTUALLY I NEED SOME HELP FOR THE MATALB CODE FOR PSO.

[Walter Roberson]

http://www.mathworks.com/matlabcentral/answers/?term=tag%3A%22pso%22

Answer 1

[Pramit Biswas]

start with THIS

Answer 2

[denny]

what's your problem exactly ? I had done some optimize for PID controller run MATLAB and MATLAB/Simulink together. My Email is [e-mail] , I can help you if you have any problem.

[syukron jamil]

can you help me? I want to use PID tunning PSo with coupled tanks system. but I don't understand how to make on simulink

[Walter Roberson]

Note that Denny will not have been notified of your comment

Answer 3

[denny]

You mean the convergence fitness curve? you can get the convergence curve by two ways as follows:

1: use your matrix to record the convergence fitness, you can plot it at any time, I recommend this way to you.

step 1: at you main function , declare a global variable as follow:

              global every_gen_bestvalues

step 2: modefy your ga function optiions, add a plot function by yourself as follow:

options=gaoptimset('Generations',100,'PopulationSize',20,'Display','iter','PlotFcn',@myplot_forga);

step 3: add a matlab function ,name as above, myplot_forga, the myplot_forga is shown as follows:

function state = myplot_forga(options,state,flag)
% set(gcf,'Visible','off')
if state.Generation>=1
    global every_gen_bestvalues
    every_gen_bestvalues=[every_gen_bestvalues; state.Best(end)];
end
end

step 4: you can plot the convergence curve like :

plot(every_gen_bestvalues)

step 5: you can save the variable as a mat file , and then figure it next time.

save('every_gen_bestvalues.mat','plot(every_gen_bestvalues)')

2: You can use the ga toolbox' plot function to plot it, if you close the figure window, you must run the ga optimization function again, so, I do not recommend this way . j ust one step:

options=gaoptimset('Generations',100,'PopulationSize',20,'Display','iter','PlotFcn',@gaplotbestf);