粒子群优化算法matlab代码

以下是粒子群优化算法的Matlab代码实现：

% 粒子群优化算法
function [best_x, best_f, conv_curve] = PSO(obj_func, dim, n_particles, max_iter, lb, ub)

% 初始化粒子位置和速度
particles_x = rand(dim, n_particles) .* (ub - lb) + lb;
particles_v = rand(dim, n_particles) .* (ub - lb) * 0.2;

% 初始化最优位置和最优适应度
particles_p = particles_x;
particles_f = obj_func(particles_x);
particles_pbest = particles_p;
particles_fbest = particles_f;

% 初始化全局最优位置和最优适应度
[global_fbest, global_fbest_idx] = min(particles_fbest);
global_pbest = particles_pbest(:,global_fbest_idx);

% 初始化收敛曲线
conv_curve = zeros(max_iter, 1);

% 迭代
for i = 1:max_iter
    % 更新速度和位置
    particles_v = 0.8 * particles_v + rand(dim, n_particles) .* (particles_pbest - particles_x) + rand(dim, n_particles) .* (global_pbest - particles_x);
    particles_x = particles_x + particles_v;
    
    % 处理越界的粒子位置
    particles_x(particles_x < lb) = lb(particles_x < lb);
    particles_x(particles_x > ub) = ub(particles_x > ub);
    
    % 计算粒子的适应度
    particles_f = obj_func(particles_x);
    
    % 更新最优位置和最优适应度
    particles_pbest_update = particles_f < particles_fbest;
    particles_pbest(:, particles_pbest_update) = particles_x(:, particles_pbest_update);
    particles_fbest(particles_pbest_update) = particles_f(particles_pbest_update);
    
    % 更新全局最优位置和最优适应度
    [global_fbest_update, global_fbest_update_idx] = min(particles_fbest);
    if global_fbest_update < global_fbest
        global_pbest = particles_pbest(:,global_fbest_update_idx);
        global_fbest = global_fbest_update;
    end
    
    % 记录收敛曲线
    conv_curve(i) = global_fbest;
end

% 返回最优位置和最优适应度
best_x = global_pbest;
best_f = global_fbest;
end

其中，`obj_func` 是目标函数，`dim` 是问题的维度，`n_particles` 是粒子数，`max_iter` 是最大迭代次数，`lb` 和 `ub` 是问题的变量上下界限制。`best_x` 是最优解，`best_f` 是最优解对应的函数值，`conv_curve` 是收敛曲线。