|
9-PSO粒子群算法的障碍路径寻优Path Planning-三角形障碍
链接:https://pan.baidu.com/s/1IZ_On5XWVkKIHGrNbQ8H5Q 提取码:33td
具体链接在halcom.cn论坛,联系人QQ:3283892722
该论坛是一个学习交流平台,我会逐一的和大家分享学习。
欢迎大家录制视频,你可在论坛进行打赏分享。
视频专用播放器:http://halcom.cn/forum.php?mod=viewthread&tid=258&extra=page%3D1
主程序如下:
- clc,clear,close all
- warning off
- % 初始化种群
- model = CreateModel();
- % PSO 参数
- c1 = 1.4995;
- c2 = 1.4995;
- Vmin = -1;
- Vmax = 1;
- w = 1; % 权重
- wdamp = 0.98; % 阻尼
- maxiter = 50; % 迭代次数
- sizepop = 20; % 种群数量
- nvar = model.n; % n个未知量
- popmin_x = model.xmin; % x
- popmax_x = model.xmax; % x
- popmin_y = model.ymin; % y
- popmax_y = model.ymax; % y
- init_pop.position = []; % 位置信息
- init_pop.Velocity = []; % 速度信息
- init_pop.fitness = []; % 适应度值信息
- init_pop.sol = []; % 解集信息
- init_pop.Best.position = []; % 个体最优 位置信息
- init_pop.Best.fitness = []; % 个体最优 适应度值信息
- init_pop.Best.sol = []; % 个体最优 解集信息
- % 粒子群
- pop = repmat(init_pop, sizepop, 1);
- clear init_pop
- %% 初始化种群
- fitnesszbest = inf;
- for i=1:sizepop
- pop(i).position = CreateRandomSolution(model); % 初始化个体
- [pop(i).fitness , pop(i).sol]= fun( pop(i).position, model ); % 适应度值
- pop(i).Velocity.x = zeros(1, nvar);
- pop(i).Velocity.y = zeros(1, nvar);
- % 更新个体最优
- pop(i).Best.position = pop(i).position;
- pop(i).Best.fitness = pop(i).fitness;
- pop(i).Best.sol = pop(i).sol;
- if(pop(i).Best.fitness<fitnesszbest)
- fitnesszbest = pop(i).Best.fitness;
- zbest = pop(i).Best;
- end
- end
- %% 迭代寻优
- for i=1:maxiter
- for j=1:sizepop
- % 横坐标的操作
- pop(j).Velocity.x = w*pop(j).Velocity.x + c1*rand(1,nvar).*(pop(j).Best.position.x-pop(j).position.x) ...
- + c2*rand(1,nvar).*(zbest.position.x-pop(j).position.x);
- % 取值范围
- pop(j).Velocity.x = max(pop(j).Velocity.x, Vmin);
- pop(j).Velocity.x = min(pop(j).Velocity.x, Vmax);
- pop(j).position.x = pop(j).position.x + pop(j).Velocity.x;
- pop(j).position.x = max(pop(j).position.x, popmin_x);
- pop(j).position.x = min(pop(j).position.x, popmax_x);
-
- % 纵坐标的操作
- pop(j).Velocity.y = w*pop(j).Velocity.y + c1*rand(1,nvar).*(pop(j).Best.position.y-pop(j).position.y) ...
- + c2*rand(1,nvar).*(zbest.position.y-pop(j).position.y);
- % 取值范围
- pop(j).Velocity.y = max(pop(j).Velocity.y, Vmin);
- pop(j).Velocity.y = min(pop(j).Velocity.y, Vmax);
- pop(j).position.y = pop(j).position.y + pop(j).Velocity.y;
- pop(j).position.y = max(pop(j).position.y, popmin_y);
- pop(j).position.y = min(pop(j).position.y, popmax_y);
-
- [pop(j).fitness , pop(j).sol]= fun( pop(j).position, model ); % 适应度值
-
- if pop(j).fitness<pop(j).Best.fitness
- pop(j).Best.fitness = pop(j).fitness;
- pop(j).Best.position = pop(j).position;
- pop(j).Best.sol = pop(j).sol;
- end
- if(pop(j).Best.fitness<fitnesszbest)
- fitnesszbest = pop(j).Best.fitness;
- zbest = pop(j).Best;
- end
-
- w = w * wdamp;
- end
- fitness_iter(i) = fitnesszbest;
-
- figure(1)
- PlotSolution(zbest.sol,model)
- pause(0.1)
- end
- %% 最后的显示部分
- disp('最优解')
- disp(zbest)
- fprintf('\n')
- figure('color',[1,1,1])
- plot(fitness_iter,'ro-','linewidth',2)
- xlabel('迭代次数'); ylabel('适应度曲线');
- axis tight
- grid on
- figure(1)
- PlotSolution(zbest.sol,model)
- pause(0.1)
复制代码 参考:
【1】结果动态图
【2】子函数参考链接
【3】粒子群算法PSO视频讲解
【4】基于PSO的火车票分配问题:链接:http://pan.baidu.com/s/1nvRYrg9 密码:bfua
【5】基于PSO的机器人路径优化问题:链接:http://pan.baidu.com/s/1jI0uNlk
【6】PSO_source_error_NRMSE_惯性因子gbest,改进的PSO算法结构体化编程
|
|