PSO.rar_PSO_pso code_pso swarm_swarm

clear all; nIter = 0; NFCCounter=0; nRuns = 50; VTRCounter=0; for tIter = 1:nRuns % clear all; %Initialization of PSO parameters wmax=0.9; wmin=0.4; itmax=5000; %Maximum iteration number c1=1.49612; c2=1.49612; for iter=1:itmax W(iter)=wmax-((wmax-wmin)/itmax)*iter; end %********************************************************** %Initialization of positions of agents % agents are initialized between -5,+5 randomly N=100; %% Population Size D=30; %% Dimensions %=================================================== a=-5.12; %% Min values of Swarm parameters b=5.12; %% Max values of Swarm Parameters Jr=0.3; %% Probability %=================================================== org_x = a+(b-a)*rand(N,D,1); %% Initial Positions of the PSO Agents x = org_x; mixedX = x; for i=1:N F(i,1,1)=f1(mixedX(i,:,1)); NFCCounter= NFCCounter +1; %FbestNFCs(tIter, NFCCounter) = min(F(:,1,1)); end [F2 indx] = sort(F); clear F; F(:,1,1) = F2(1:N); x = mixedX(indx(1:N),:); %Initialization of velocities of agents %Between -5 , +5, (which can also be started from zero) % m=-1; % n=1; m= 0.1 * (a - b ); n= 0.1 * (b - a ); V=m+(n-m)*rand(N,D,1); %% Inital velocity of the Agents %********************************************************** %Function to be minimized. %x(i,2,1)=x2 %x(i,1,1)=x1 %f=100*(x2-x1^2)^2+(x1-1)^2 % for i=1:N % F(i,1,1)=100*(x(i,2,1)-x(i,1,1)^2)^2+(x(i,1,1)-1)^2; % % NFCCounter= NFCCounter +1 % end %********************************************************** [C,I]=min(abs(F(:,1,1))); B(1,1,1)=C; XX(1,1,1)=I; gbest(1,:,1)=x(I,:,1); % Global Best % gbest(1,2,1)=x(I,2,1); % Global Best %******************************************************** %Matrix composed of gbest vector for p=1:N for r=1:D G(p,r,1)=gbest(1,r,1); end end graphIndex = 1; Fbest(tIter,1)=f1(G(1,:,1)); graphFbest(tIter, graphIndex:NFCCounter) = Fbest(tIter,1); graphIndex = NFCCounter; for i=1:N pbest(i,:,1)=x(i,:,1); end V(:,:,2)=W(1)*V(:,:,1)+c1*rand*(pbest(:,:,1)-x(:,:,1))+c2*rand*(G(:,:,1)-x(:,:,1)); x(:,:,2)=x(:,:,1)+V(:,:,2); Fb(tIter,1)=f1(gbest(1,:,1)); %****************************************************** for j=2:itmax-1 probab = rand; flag = 0; % Calculation of new positions for i=1:N F(i,1,j)=f1(x(i,:,j)); NFCCounter= NFCCounter +1 IterationDisp = tIter end % F(:,:,j) [C,I]=min(abs(F(:,:,j))); %========================================================================== if ((C <= 0.00000001) || (NFCCounter >=1000000)) %0.0000000000000000000000001) || (NFCCounter >=100000)) %10 raise to power -25 %0.0000000000000000000000001 BY Calculator C x(I,:,j) VTRCounter=VTRCounter+1; break; end; %============================================================================ B(1,1,j)=C; gbest(1,:,j)=x(I,:,j); % gbest(1,2,j)=x(I,2,j); Fb(tIter,j)=f1(gbest(1,:,j)); [C,I]=min(Fb(tIter,:)); if Fb(tIter,j)<=C gbest(1,:,j)=gbest(1,:,j); % gbest(1,2,j)=gbest(1,2,j); else gbest(1,:,j)=gbest(1,:,I); % gbest(1,2,j)=gbest(1,2,I); end %Matrix composed of gbest vector for p=1:N for r=1:D G(p,r,j)=gbest(1,r,j); end end Fbest(tIter,j)=f1(G(1,:,j)); graphFbest(tIter, graphIndex:NFCCounter) = Fbest(tIter,j); graphIndex = NFCCounter; for i=1:N [C,I]=min(F(i,1,:)); if F(i,1,j)<=C pbest(i,:,j)=x(i,:,j); else pbest(i,:,j)=x(i,:,I); end end V(:,:,j+1)=W(j)*V(:,:,j)+c1*rand*(pbest(:,:,j)-x(:,:,j))+c2*rand*(G(:,:,j)-x(:,:,j)); % Velocity Clamping % for individual = 1:N % for dim = 1:D % if V(individual,dim,j+1) < m % V(individual,dim,j+1) = m; % end % if V(individual,dim,j+1) > n % V(individual,dim,j+1) = n; % end % % end % end % End Velocity Clamping x(:,:,j+1)=x(:,:,j)+V(:,:,j+1); end nIter(tIter) = j; NFCs(tIter) = NFCCounter; BestFitness(tIter) = C; % plot(x(:,1,j),x(:,2,j),'o'); clear pbest; clear gbest; clear Fb; clear F; NFCCounter = 0; end; MeanNFC = sum(NFCs) / nRuns [M N] = size(graphFbest); for i = 1:N FbestAvg(i) = sum(graphFbest(:,i))/M; end; %hold on, plot(FbestAvg(1,1:MeanNFC), 'color', 'blue'); % SuccessRate=VTRCounter/nRuns MaxNFC = max(NFCs) MinNFC = min(NFCs) STDDEV= std(NFCs) MeanBestFitness =sum(BestFitness)/nRuns MinBestFitness =min(BestFitness) MaxBestFitness =max(BestFitness) MedianFitness =median(BestFitness)