matlab

机械臂运动路径设计问题中代码遗传算法没有gafk_operator

2019-07-17 14:31发布

站内问答 / matlab
2635 1 1134
模型一程序
clear;clc;
AB=140;BC=255;CD=255;DE=65;
x_obj=input('x_Obj=');
y_obj=input('y_Obj=');
z_obj=input('z_Obj=');
obj=[x_obj,y_obj,z_obj];
%=========Theta取值范围====
max_range=[180 125+90 138 270 133.5+90];
min_range=[-180 -125+90 -138 -270 -120+90];
%=========Theta取值范围end====
M=400;T=5000;pc=0.8;pm=0.1;%M为种群规模，T为迭代次数，PC为交叉率，PM为变异率
for i=1:M
    for j=1:5
        pop(i,j)=round(((min_range(j)+(max_range(j)-min_range(j))*rand))*10)*0.1;%每个种群的取值
    end
    E(i,1)=-(BC*cos(pop(i,2)*(pi/180))+CD*cos(pop(i,3)*(pi/180)))*sin(pop(i,1)*(pi/180))-DE*cos(pop(i,5)*(pi/180))*sin(pop(i,4)*(pi/180));
    E(i,2)=(BC*cos(pop(i,2)*(pi/180))+CD*cos(pop(i,3)*(pi/180)))*cos(pop(i,1)*(pi/180))+DE*sin(pop(i,5)*(pi/180));
    E(i,3)=BC*sin(pop(i,2)*(pi/180))+AB+CD*sin(pop(i,3)*(pi/180))-DE*cos(pop(i,5)*(pi/180))*cos(pop(i,4)*(pi/180));
end
%=======误差作为目标函数=====
for i=1:M
    Objfun(i,1)=0.99*sqrt((obj(1)-E(i,1)).^2+(obj(2)-E(i,2)).^2+(obj(3)-E(i,3)).^2)+0.01*round(0.5*round(max(abs(pop(i,:)))));%round(0.5*round(max(abs(BestofPop(:)))))
end
%=======误差作为目标函数end=====
for t=1:T
    FitnV=ranking(Objfun);
    parent_pop=select('rws',pop,FitnV);
    child_pop=gafk_operator(parent_pop,t,T,pc,pm,min_range,max_range);
    com_pop=round([parent_pop;child_pop]*10)*0.1;
    [N,temp]=size(com_pop);
    for i=1:N
        E(i,1)=-(BC*cos(com_pop(i,2)*(pi/180))+CD*cos(com_pop(i,3)*(pi/180)))*sin(com_pop(i,1)*(pi/180))-DE*cos(com_pop(i,5)*(pi/180))*sin(com_pop(i,4)*(pi/180));
        E(i,2)=(BC*cos(com_pop(i,2)*(pi/180))+CD*cos(com_pop(i,3)*(pi/180)))*cos(com_pop(i,1)*(pi/180))+DE*sin(com_pop(i,5)*(pi/180));
        E(i,3)=BC*sin(com_pop(i,2)*(pi/180))+AB+CD*sin(com_pop(i,3)*(pi/180))-DE*cos(com_pop(i,5)*(pi/180))*cos(com_pop(i,4)*(pi/180));
        Objfun_mind(i,1)=0.99*sqrt((obj(1)-E(i,1)).^2+(obj(2)-E(i,2)).^2+(obj(3)-E(i,3)).^2)+0.01*round(0.5*round(max(abs(com_pop(i,:)))));
    end
    [sorted_obj,index_obj]=sort(Objfun_mind);
    for k=1:M
        pop(k,:)=com_pop(index_obj(k),:);
        Objfun(k,:)=Objfun_mind(index_obj(k),:);
    end
    [best_obj(t,1),index]=min(Objfun);
    best_pop(t,:)=pop(index,:);
    best_E(t,:)=E(index,:);
end
[JofObj,daishu]=min(best_obj(:,1));
BestofPop=best_pop(daishu,:);
BestofE=best_E(daishu,:);
error=sqrt((obj(1)-BestofE(1)).^2+(obj(2)-BestofE(2)).^2+(obj(3)-BestofE(3)).^2);%误差
step=round(0.5*round(max(abs(BestofPop(:)))));%步子数
plot(best_obj(:,1),'-')
www.xszlw.com  求大神帮忙哈，急用！！！！
友情提示: 此问题已得到解决，问题已经关闭,关闭后问题禁止继续编辑，回答。
该问题目前已经被作者或者管理员关闭, 无法添加新回复
1条回答
wangxh868
1楼-- · 2019-07-17 19:37

一周热门 更多>