遗传算法经典实例matlab_蚁群算法matlab实例

%目标函数
function y = fobj(x)
if x <= 0
    y = sin(x);
elseif x <= 2
        y = x^2/2;
else 
    y = 3-x/2;
end
end

%主程序
%清楚窗口内容和变量
clear
clc   

pc = 0.8 ;%交叉率
pm = 0.05 ;%变异率

Iter_N =100; %迭代最大代数
popsize = 10; %种群规模，个体的集合

LB = -10;
UB = 10;  %左右边界
DELTA = 1e-2; 
x=LB:0.05:UB;

for i = 1:length(x)
    y(i)=fobj(x(i));  %目标函数
end
plot(x,y)
hold on;

1. 初始化种群
function pop = ini_pop(LB,UB,DELTA,popsize)
N = ceil(log2((UB-LB)/DELTA+1)); %
for i = 1:popsize
    pop(i,:)=randi(2,[1,N])-1; %
end

%2. fitness 计算适应性
function y = fitness(LB,UB,x)
popsize = size(x,1);
for i = 1:popsize
xtemp = x(i,:);
N = length(xtemp);
xdec = Bin2Dec(xtemp(1:N));
xx = LB+(UB-LB)/(2^N-1)*xdec;
y(i) = fobj(xx);
end

%主程序
pop=ini_pop(LB,UB,DELTA,popsize); %生成初始种群
fit_pop = fitness(LB,UB,pop) %计算适应度
[best_fit,ind]=max(fit_pop); %找到最大适应度个体
best_pop = pop(ind,:); %记录最佳个体
for k = 1:Iter_N
newpop = selection(pop,fit_pop); %轮盘赌选择新种群
newpop = crossover(newpop,pc); %交叉算子
newpop = mutation(newpop,pm); %变异算子
fit_pop = fitness(LB,UB,newpop);%计算适应度
[bf,ind] = max(fit_pop);
if bf<best_fit
R=randi(popsize);
newpop(R,:)=best_pop;
fit_pop(R)=best_fit;
else
best_pop=newpop(ind,:);
end
pop = newpop;
fstar(k)=best_fit;
N=length(best_pop);
xdec=Bin2Dec(best_pop(1:N));
xx=LB+(UB-LB)/(2^N-1)*xdec;
best_fit;
plot(xx,best_fit,'r*');
pause(0.1);
end