用matlab禁忌搜索算法解决0-1背包问题的代码

以下是用Matlab实现的禁忌搜索算法解决0-1背包问题的代码，希望对你有所帮助：

function [best_solution, best_fitness] = tabu_search(capacity, weights, profits, max_iterations, tabu_size)

% 初始化参数
n_items = length(weights);
current_solution = zeros(1, n_items); % 当前解
best_solution = zeros(1, n_items); % 最优解
best_fitness = 0; % 最优解的适应度值
current_fitness = 0; % 当前解的适应度值

% 初始化tabu列表
tabu_list = zeros(n_items, tabu_size);

% 迭代搜索
for i = 1:max_iterations
    % 生成邻域解
    neighborhood = generate_neighborhood(current_solution, tabu_list);
    
    % 计算邻域解的适应度值
    neighborhood_fitness = zeros(1, size(neighborhood, 1));
    for j = 1:size(neighborhood, 1)
        neighborhood_fitness(j) = calculate_fitness(capacity, weights, profits, neighborhood(j, :));
    end
    
    % 选择最优的邻域解
    [neighborhood_fitness_sorted, sort_index] = sort(neighborhood_fitness, 'descend');
    for j = 1:length(sort_index)
        if ~tabu(neighborhood(sort_index(j), :), tabu_list)
            current_solution = neighborhood(sort_index(j), :);
            current_fitness = neighborhood_fitness_sorted(j);
            break;
        end
    end
    
    % 更新tabu列表
    tabu_list = [tabu_list(:, 2:end), current_solution'];
    
    % 更新最优解
    if current_fitness > best_fitness
        best_solution = current_solution;
        best_fitness = current_fitness;
    end
end

end

% 生成邻域解
function neighborhood = generate_neighborhood(solution, tabu_list)
n_items = length(solution);
neighborhood = repmat(solution, n_items, 1);
for i = 1:n_items
    if solution(i) == 0
        neighborhood(i, i) = 1;
    else
        neighborhood(i, i) = 0;
    end
end
neighborhood = setdiff(neighborhood, tabu_list', 'rows');
end

% 计算适应度值
function fitness = calculate_fitness(capacity, weights, profits, solution)
if sum(solution.*weights) <= capacity
    fitness = sum(solution.*profits);
else
    fitness = 0;
end
end

% 判断是否在tabu列表中
function is_tabu = tabu(solution, tabu_list)
is_tabu = false;
for i = 1:size(tabu_list, 2)
    if isequal(solution, tabu_list(:, i)')
        is_tabu = true;
        break;
    end
end
end

这里的禁忌搜索算法采用了邻域解生成、适应度计算、tabu列表维护等基本步骤，其中邻域解生成使用了简单的交换算子，并使用tabu列表来避免陷入局部最优解。