《基于改进教与学算法的配电网无功优化》matlab代码复现

%% 清理内存空间 clc clear close all %% 算法参数的设定 index=input('请选择测试函数：1-Sphere，2-Griewank，3-Rosenbrock，4-Ackley，5-Rastrign'); x_num=50; % 班级人数 dim=2; % 问题的维度/决策变量的个数 it_max=100; % 最大迭代次数 [x_min,x_max,v_min,v_max]=set_pop(index,dim); % 设置位置，速度的上下限 N_ES=3; % 精英人数 %% 步骤1：班级初始化 x=x_min+rand(x_num,dim).*(x_max-x_min); % 初始化班级 x_teacher=x(1,:); % 初始化老师位置 fitness=zeros(1,x_num); % 所有个体的适应度 fitness_new=zeros(1,x_num); % 更新后个体适应度 fitness_teacher=inf; % 初始化老师的适应度 % 初始的适应度 for k=1:x_num % 计算适应度值 switch index case 1 fitness(k)=Sphere(x(k,:)); case 2 fitness(k)=Griewank(x(k,:)); case 3 fitness(k)=Rosenbrock(x(k,:)); case 4 fitness(k)=Ackley(x(k,:)); case 5 fitness(k)=Rastrign(x(k,:)); otherwise warning('输入有误，请重新输入') end if fitness(k)<fitness_teacher fitness_teacher=fitness(k); x_teacher=x(k,:); end end %% 迭代求最优解 for it=1:it_max %% 步骤2：选取并保留精英解 [~,fitness_rank]=sort(fitness); ES_index=fitness_rank(1:N_ES); %% 步骤3：老师的教学 x_mean=mean(x); TF=randi([1,2]); x_differance=rand*(x_teacher-x_mean); x_new=x+x_differance; for k=1:x_num % 计算适应度值 switch index case 1 fitness(k)=Sphere(x(k,:)); fitness_new(k)=Sphere(x_new(k,:)); case 2 fitness(k)=Griewank(x(k,:)); fitness_new(k)=Griewank(x_new(k,:)); case 3 fitness(k)=Rosenbrock(x(k,:)); fitness_new(k)=Rosenbrock(x_new(k,:)); case 4 fitness(k)=Ackley(x(k,:)); fitness_new(k)=Ackley(x_new(k,:)); case 5 fitness(k)=Rastrign(x(k,:)); fitness_new(k)=Rastrign(x_new(k,:)); otherwise warning('输入有误，请重新输入') end if fitness(k)>fitness_new(k) x(k,:)=x_new(k,:); end end %% 步骤4：学生之间互相学习 for k=1:x_num studied_set=setdiff(1:x_num,k); studied_object=studied_set(randi([1,x_num-1])); if fitness(studied_object)<fitness(k) x_new(k,:)=x(k,:)+rand*(x(studied_object,:)-x(k,:)); else x_new(k,:)=x(k,:)-rand*(x(studied_object,:)-x(k,:)); end % 计算适应度值 switch index case 1 fitness(k)=Sphere(x(k,:)); fitness_new(k)=Sphere(x_new(k,:)); case 2 fitness(k)=Griewank(x(k,:)); fitness_new(k)=Griewank(x_new(k,:)); case 3 fitness(k)=Rosenbrock(x(k,:)); fitness_new(k)=Rosenbrock(x_new(k,:)); case 4 fitness(k)=Ackley(x(k,:)); fitness_new(k)=Ackley(x_new(k,:)); case 5 fitness(k)=Rastrign(x(k,:)); fitness_new(k)=Rastrign(x_new(k,:)); otherwise warning('输入有误，请重新输入') end if fitness(k)>fitness_new(k) x(k,:)=x_new(k,:); end end %% 步骤5：精英解替换最差学员。 [~,fitness_rank]=sort(-fitness); worst_index=fitness_rank(1:N_ES); x(worst_index,:)=x(ES_index,:); %% 步骤6：重复个体裁剪 [x_unique,ia,ic]=unique(x,'stable','rows'); for kk=setdiff(1:x_num,ia) kkk=randi([1,dim]); x(kk,kkk)=x_min(kkk)+rand*(x_max(kkk)-x_min(kkk)); end %% 更新越限的粒子 for k=1:dim x(x(:,k)>x_max(k),k)=x_max(k); x(x(:,k)<x_min(k),k)=x_min(k); end %% 更新适应度值 for k=1:x_num switch index case 1 fitness(k)=Sphere(x(k,:)); case 2 fitness(k)=Griewank(x(k,:)); case 3 fitness(k)=Rosenbrock(x(k,:)); case 4 fitness(k)=Ackley(x(k,:)); case 5 fitness(k)=Rastrign(x(k,:)); otherwise warning('输入有误，请重新输入') end if fitness(k)<fitness_teacher fitness_teacher=fitness(k); x_teacher=x(k,:); end end figure(1) scatter(x(:,1),x(:,2),100) title(['第',num2str(it),'次迭代粒子分布情况']) % axis([x_min(1) x_max(1) x_min(2) x_max(2)]) % drawnow % F=getframe(gcf); % I=frame2im(F); % [I,map]=rgb2ind(I,256); % if it == 1 % imwrite(I,map,'fun5.gif','gif', 'Loopcount',inf,'DelayTime',0.2); % else % imwrite(I,map,'fun5.gif','gif','WriteMode','append','DelayTime',0.2); % end % hold off pause(0.01) end