遗传算法优化粒子群算法求解shubert函数，ga-pso算法求解测试函数

clc clear close all warning off set(0,'defaultfigurecolor','w') %x y xmax = [5.12 5.12]; xmin = [-5.12 -5.12]; vmax = 0.2*xmax; vmin = -vmax; fun = @shubert; m=2; %程序初始化 % global popsize; %种群规模 gen=20; %设置进化代数 popsize=300; %设置种群规模大小 best_in_history(gen)=inf; %初始化全局历史最优解 best_in_history(:)=inf; %初始化全局历史最优解 best_fitness=inf; fz = zeros(gen,5); %设置种群数量 pc=0.5; %交叉概率选择，0和1之间 pm=0.1; %变异概率选择，0和1之间 pop1 = zeros(popsize,m); pop2 = zeros(popsize,m); pop3 = zeros(popsize,m); pop6 = zeros(gen,m);%存储解码后的每代最优粒子 pop7 = zeros(popsize,m);%存储更新解码后的粒子的位置 for ii1=1:popsize pop1(ii1,:)=funx(xmin,xmax,m); %初始化种群中的粒子位置， pop3(ii1,:)=pop1(ii1,:); %初始状态下个体最优值等于初始位置 pop2(ii1,:)=funv(vmax,m); %初始化种群微粒速度， pop4(ii1,1)=inf; pop5(ii1,1)=inf; end pop0=pop1; xmax = [5.12 5.12]; xmin = [-5.12 -5.12]; c1=2; c2=2; gbest_x=pop1(end,:); % pop1(1:size(num,1),:) = num; %全局最优初始值为种群第一个粒子的位置 for exetime=1:gen ww = 0.7*(gen-exetime)/gen+0.2; for ii4=1:popsize pop2(ii4,:)=(ww*pop2(ii4,:)+c1*rand(1,m).*(pop3(ii4,:)-pop1(ii4,:))+c2*rand(1,m).*(gbest_x-pop1(ii4,:))); %更新速度 for jj = 1:m if pop2(ii4,jj)<vmin(jj) pop2(ii4,jj)=vmin(jj); elseif pop2(ii4,jj)>vmax(jj) pop2(ii4,jj)=vmax(jj); end end end %更新粒子位置 for ii5=1:popsize pop1(ii5,:)=pop1(ii5,:)+pop2(ii5,:); for jj2 = 1:m if pop1(ii5,jj2)>xmax(jj2) pop1(ii5,jj2) = xmax(jj2); elseif pop1(ii5,jj2)<xmin(jj2) pop1(ii5,jj2)=xmin(jj2); end end % if rand>0.85 % k=ceil(m*rand); % pop1(ii5,k) = (xmax( k)-xmin(k)).*rand(1,1)+xmin(k); % end % if pop5(ii5)>sum(pop5)/popsize % pop1(ii5,:) = (xmax(1,m)-xmin(1,m)).*rand(1,m)+xmin(1,m); % end end for jj2 = 1:m if pop1(ii5,jj2)>xmax(jj2) pop1(ii5,jj2) = xmax(jj2); elseif pop1(ii5,jj2)<xmin(jj2) pop1(ii5,jj2)=xmin(jj2); end end if exetime-1>0 plot(1:length(best_in_history(1:exetime-1)),best_in_history(1:exetime-1)); xlabel('迭代次数') ylabel('适应度') title('遗传算法粒子群GA-PSO算法') hold on; pause(0.1) end pop1(end,:) = gbest_x; %计算适应值并赋值 for ii3=1:popsize [my,mx] = fun2(pop1(ii3,:)); % [my,mx] = fun2(gbest_x,num,xmax,xmin); pop5(ii3,1)=my; pop7(ii3,:) = mx; if pop4(ii3,1)>pop5(ii3,1) %若当前适应值优于个体最优值，则进行个体最优信息的更新 pop4(ii3,1)=pop5(ii3,1); %适值更新 pop3(ii3,:)=pop1(ii3,:); %位置坐标更新 end end %计算完适应值后寻找当前全局最优位置并记录其坐标 if best_fitness>min(pop4(:,1)) best_fitness=min(pop4(:,1)) ; %全局最优值 ag = []; ag =find(pop4(:,1)==min(pop4(:,1))); gbest_x(1,:)=(pop1(ag(1),:)); %全局最优粒子的位置 pop6(exetime,:) = pop7(ag(1),:); else fz(exetime,:) = fz(exetime-1,:); if exetime>1 pop6(exetime,:) = pop6(exetime-1,:); end end best_in_history(exetime)=best_fitness; %记录当前全局最优 meany = mean(pop5); %变异 for hh = 1:popsize if pop5(hh)>meany if rand()>pm k=ceil(m*rand()); pop1(hh,k) = (xmax( k)-xmin(k)).*rand(1,1)+xmin(k); end end end %交叉 pany = []; for hh = 1:popsize if pop5(hh)>meany if isempty(pany) pany =[pop1(hh,:) hh]; else panx = [pop1(hh,:) hh]; if rand()>pc k=ceil(m*rand()); pop1(pany(end),k) = panx(k); pop1(panx(end),k) =pany(k); pany = []; end end end end end x = -2:0.1:2; y = -2:0.1:2; x = -10:0.1:10; y = -10:0.1:10; [x,y] = meshgrid(x,y); [m,n] = size(x); z = zeros(m,n); for ii = 1:m for jj = 1:n xx = [x(ii,jj) y(ii,jj)]; z(ii,jj) = shubertfun(xx); end end figure surf(x,y,z) hold on xlabel('x1') ylabel('x2') zlabel('z') axis([-2 2 -2 2 -200 200]) % axis([-10 10 -10 10 -200 200]) % shading interp title('Shubert Function') set(gca,'fontsize',12) colormap jet plot(pop0(:,1),pop0(:,2),'ro','MarkerFaceColor','r') xlabel('X') ylabel('Y') title('初始种群') set(gca,'fontsize',12) view([-130 40]) figure surf(x,y,z) hold on xlabel('x1') ylabel('x2') zlabel('z') axis([-2 2 -2 2 -200 200]) % axis([-10 10 -10 10 -200 200]) % shading interp title('Shubert Function') set(gca,'fontsize',12) colormap jet plot(pop1(:,1),pop1(:,2),'ro','MarkerFaceColor','r') xlabel('X') ylabel('Y') title('收敛后的种群') set(gca,'fontsize',12) view([-130 40])