1 简介

2 部分代码

clear all

clc

close all

%% 参数

parameter.nvar = 2;

parameter.xmin = -1;

parameter.xmax = 1;

parameter.m = 50; 

parameter.k = 15;%控制适配值之间差异的常数

parameter.num_part =10;

itermax = 2000;

crossover_probability = 0.5;

mutation_probability = 0.001;

num_part = parameter.num_part;

nvar = parameter.nvar; 

xmin = parameter.xmin;

xmax = parameter.xmax;

m = parameter.m;

   %% 初始化

    generation = repmat([], num_part, 1);

    for i = 1:num_part

        generation(i).x_bi = randi([0,1],1,parameter.nvar*parameter.m);%随机初始

        generation(i).cost = my_obj(generation(i).x_bi,parameter);

    end

   generation_new =generation;   

   dert_mean_cost = 1;

iter = 1;

while(dert_mean_cost>1e-10&&iter<=itermax)

generation = generation_new;

%% 复制选择

[cost_sort index] = sort([generation.cost]');

generation_sort = generation(index);%从小到大 对应

   for i = 1:num_part

      fitness(i) = parameter.k*(num_part-i)/num_part;

   end 

fitness_percent = fitness/sum(fitness);

[generation_selet] = percent_select(generation_sort,fitness,parameter); 

%% 交叉

 [generation_cross] = crossover(generation_selet,crossover_probability,parameter);

%% 变异，                     

[generation_new] = mutation(generation_cross,mutation_probability,parameter);

best(iter).cost = 10;

for i = 1:num_part

    generation_new(i).cost = my_obj(generation_new(i).x_bi,parameter);

    if generation_new(i).cost< best(iter).cost

        best(iter).cost = generation_new(i).cost;

        best(iter).x_bi = generation_new(i).x_bi;

    end

    %-------------------------------

    x_obj =generation_new(i).x_bi;

    b(1) = bi2de(x_obj(1:m));

    b(2) = bi2de(x_obj(m+1:nvar*m));

    x = xmin + b*(xmax-xmin)/(2^m-1);

    tempX(iter,i) =x(1);

    tempY(iter,i) =x(2);

    %-------------------

end

    meancost(iter) = mean([generation_new.cost]);

        disp(['Iteration ' num2str(iter) '| mean cost ' num2str(meancost(iter)) '| best_cost ' num2str(best(iter).cost)]);

     if  iter==1

        dert_mean_cost = 1;

     else

          dert_mean_cost = abs(meancost(iter) - meancost(iter-1));

     end

    iter = iter + 1;

end

  disp('平均最优值')

 meanobj = meancost(iter-1)

  disp('全局最优值')

 bestobj =  best(iter-1).cost

 disp('最优值对应自变量')

nvar = parameter.nvar; 

xmin = parameter.xmin;

xmax = parameter.xmax;

m = parameter.m;

 x_obj = best(iter-1).x_bi;

b(1) = bi2de(x_obj(1:m));

b(2) = bi2de(x_obj(m+1:nvar*m));

x = xmin + b*(xmax-xmin)/(2^m-1)

 %% 画优化曲线

   figure(2)

    plot(1:iter-1,[best.cost]);

    title('最优个体适配值曲线');

   figure(3)

    plot(1:iter-1,meancost(1:iter-1));

     title('平均适配值曲线');

3 仿真结果

4 参考文献

[1]刘鲭洁,陈桂明,杨旗. "基于Matlab工具的遗传算法求解有约束最优化问题." 兵工自动化 27.11(2008):2.

博主简介：擅长智能优化算法、神经网络预测、信号处理、元胞自动机、图像处理、路径规划、无人机等多种领域的Matlab仿真，相关matlab代码问题可私信交流。

部分理论引用网络文献，若有侵权联系博主删除。