【智能优化算法】基于遗传算法求解非线性目标函数最小值问题含Matlab源码
2022/2/8 17:17:47
本文主要是介绍【智能优化算法】基于遗传算法求解非线性目标函数最小值问题含Matlab源码,对大家解决编程问题具有一定的参考价值,需要的程序猿们随着小编来一起学习吧!
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。
这篇关于【智能优化算法】基于遗传算法求解非线性目标函数最小值问题含Matlab源码的文章就介绍到这儿,希望我们推荐的文章对大家有所帮助,也希望大家多多支持为之网!
- 2024-11-22[开源]10.3K+ Star!轻量强大的开源运维平台,超赞!
- 2024-11-21Flutter基础教程:新手入门指南
- 2024-11-21Flutter跨平台教程:新手入门详解
- 2024-11-21Flutter跨平台教程:新手入门与实践指南
- 2024-11-21Flutter列表组件教程:初学者指南
- 2024-11-21Flutter列表组件教程:新手入门指南
- 2024-11-21Flutter入门教程:初学者必看指南
- 2024-11-21Flutter入门教程:从零开始的Flutter开发指南
- 2024-11-21Flutter升级教程:新手必读的升级指南
- 2024-11-21Flutter升级教程:轻松掌握Flutter版本更新