ERROR with PSO algorithm for two variable function

Question

Mohammad 2015-7-23

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/230957-error-with-pso-algorithm-for-two-variable-function

编辑： Mohammad 2015-7-23

I write this PSO algorithm. my algorithm worked for single variable Cost Function. However, when I want run this code for two variable Cost Function I see this error

 Error in Iterativepso (line 43)
            particle(i).cost=CostFunction(particle(i).position);

my code :

function [best_cost,best_particle,N_iter]=Iterativepso

clc; clear;

tic

npop=100; maxit=40; w=1; wdamp=0.98;

c1=2; c2=2;

xmin = [0.0001 0.0001]; % Lower band of parameter case30 xmax = [1 1];

nvar=size(xmin,2); dx=xmax-xmin;

vmax=0.1*dx;

empty_particle.position=[]; empty_particle.velocity=[]; empty_particle.cost=[]; empty_particle.pbest=[]; empty_particle.pbestcost=[];

particle=repmat(empty_particle,npop,1);

gbest=zeros(maxit,nvar); gbestcost=zeros(maxit,1);

for it=1:maxit if it==1 gbestcost(1)=inf; for i=1:npop particle(i).velocity=zeros(1,nvar); particle(i).position=xmin+(xmax-xmin).*rand(1,nvar); particle(i).cost=CostFunction(particle(i).position); particle(i).pbest=particle(i).position; particle(i).pbestcost=particle(i).cost;

            if particle(i).pbestcost<gbestcost(it)
                gbest(it,:)=particle(i).pbest;
                gbestcost(it)=particle(i).pbestcost;
            end
        end
    else
        gbest(it,:)=gbest(it-1,:);
        gbestcost(it)=gbestcost(it-1);
        for i=1:npop
            particle(i).velocity=w*particle(i).velocity...
                                +c1*rand*(particle(i).pbest-particle(i).position)...
                                +c2*rand*(gbest(it,:)-particle(i).position);
            particle(i).velocity=min(max(particle(i).velocity,-vmax),vmax);
            particle(i).position=particle(i).position+particle(i).velocity;
            particle(i).position=min(max(particle(i).position,xmin),xmax);
            particle(i).cost=CostFunction(particle(i).position);
            if particle(i).cost<particle(i).pbestcost
                particle(i).pbest=particle(i).position;
                particle(i).pbestcost=particle(i).cost;
                if particle(i).pbestcost<gbestcost(it)
                    gbest(it,:)=particle(i).pbest;
                    gbestcost(it)=particle(i).pbestcost;
                end
            end
        end
    end
    w=w*wdamp;
end
best_particle=gbest(maxit,:);
best_cost=gbestcost(maxit);
N_iter=npop*maxit;
toc