遗传算法matlab程序

以下是一个基本的遗传算法的 MATLAB 实现，供您参考：

% 遗传算法 MATLAB 实现

% 初始化遗传算法参数
pop_size = 50; % 群体大小
chrom_length = 20; % 染色体长度
pc = 0.8; % 交叉概率
pm = 0.01; % 变异概率
max_gen = 100; % 最大迭代次数

% 初始化种群
pop = randi([0, 1], pop_size, chrom_length);

% 计算适应度
fitness = sum(pop, 2);

% 迭代
for i = 1:max_gen
    % 选择
    [parents, parent_fitness] = selection(pop, fitness);
    
    % 交叉
    offspring = crossover(parents, pc);
    
    % 变异
    offspring = mutation(offspring, pm);
    
    % 计算适应度
    offspring_fitness = sum(offspring, 2);
    
    % 合并种群和后代
    pop = [pop; offspring];
    fitness = [fitness; offspring_fitness];
    
    % 选择新一代
    [pop, fitness] = elitism(pop, fitness, pop_size);
    
    % 显示当前迭代次数和最佳解
    [~, idx] = max(fitness);
    best_solution = pop(idx, :);
    fprintf('第 %d 次迭代，最佳解为 %s，适应度为 %d\n', i, num2str(best_solution), fitness(idx));
end

% 选择操作，采用锦标赛选择方法
function [parents, parent_fitness] = selection(pop, fitness)
    tournament_size = 2; % 锦标赛大小
    num_parents = size(pop, 1); % 需要选择的父代数量
    parents = zeros(num_parents, size(pop, 2)); % 初始化父代矩阵
    parent_fitness = zeros(num_parents, 1); % 初始化父代适应度向量
    
    for i = 1:num_parents
        % 随机选择 tournament_size 个个体
        idx = randi(size(pop, 1), tournament_size, 1);
        
        % 选择最佳个体作为父代
        [~, best_idx] = max(fitness(idx));
        parents(i, :) = pop(idx(best_idx), :);
        parent_fitness(i) = fitness(idx(best_idx));
    end
end

% 交叉操作，采用单点交叉方法
function offspring = crossover(parents, pc)
    chrom_length = size(parents, 2);
    num_offspring = size(parents, 1);
    offspring = zeros(num_offspring, chrom_length);
    
    for i = 1:2:num_offspring
        if rand < pc
            % 随机选择交叉点
            cross_point = randi(chrom_length - 1);
            
            % 进行交叉
            offspring(i, :) = [parents(i, 1:cross_point), parents(i+1, cross_point+1:end)];
            offspring(i+1, :) = [parents(i+1, 1:cross_point), parents(i, cross_point+1:end)];
        else
            % 不进行交叉
            offspring(i, :) = parents(i, :);
            offspring(i+1, :) = parents(i+1, :);
        end
    end
end

% 变异操作，采用单点变异方法
function offspring = mutation(pop, pm)
    chrom_length = size(pop, 2);
    num_offspring = size(pop, 1);
    offspring = pop;
    
    for i = 1:num_offspring
        for j = 1:chrom_length
            if rand < pm
                % 进行变异
                offspring(i, j) = ~pop(i, j);
            end
        end
    end
end

% 精英策略，选择适应度最高的前 pop_size 个个体作为新一代种群
function [pop, fitness] = elitism(pop, fitness, pop_size)
    [~, idx] = sort(fitness, 'descend');
    pop = pop(idx(1:pop_size), :);
    fitness = fitness(idx(1:pop_size));
end

这是一个基本的遗传算法实现，您可以根据自己的需求进行调整和修改。