entropy.m

遗传算法求解TSP问题,MATLAB模拟实验,对采用遗传算法解优化问题很有帮助

% 计算种群平均信息熵
function aver = entropy(oldpop,outfp)

popsize = length(oldpop);
lchrom = length(oldpop(1).chrom);
N = zeros(lchrom,lchrom);
H = zeros(1,lchrom);

for j = 1:lchrom
    for i = 1:popsize
        for k = 1:lchrom
            if oldpop(i).chrom(j) == k
                N(j,k) = N(j,k) + 1;  % 统计第j位上出现字符k的个体数
            end
        end
    end
    P(j,:) = N(j,:)/popsize;  % j位上各字符出现概率
    for k = 1:lchrom
        if P(j,k) ~= 0
            H(j) = H(j) - P(j,k)*log2(P(j,k));  % H(j)表第j位上的信息量
        end
    end
end
aver = sum(H)/lchrom;
fprintf(outfp,'   种群的平均信息量为：   %8.2f\n',aver);