写一个用自适应变异粒子群法求解关于四维变量X的函数的最小值的代码 MATLAB...

好的，以下是一个简单的用自适应变异粒子群算法（AMPSO）求解关于四维变量 X 的函数的最小值的 MATLAB 代码：

function [best_value, best_position] = AMPSO(fitness_func, dim, lower_bound, upper_bound, max_iter)

% Set parameters
particle_size = 30;
w_min = 0.4;
w_max = 0.9;
c1 = 2;
c2 = 2;
r1 = rand();
r2 = rand();
max_no_improve = 10;
mutate_rate = 0.1;

% Initialize particles
particle_position = rand(particle_size, dim) .* (upper_bound - lower_bound) + lower_bound;
particle_velocity = rand(particle_size, dim) .* (upper_bound - lower_bound) * 0.05;
particle_fitness = zeros(particle_size, 1);
particle_best_position = particle_position;
particle_best_fitness = Inf(particle_size, 1);
global_best_fitness = Inf;
global_best_position = zeros(1, dim);
no_improve = 0;

% Evaluate fitness
for i = 1:particle_size
    particle_fitness(i) = fitness_func(particle_position(i, :));
    if particle_fitness(i) < particle_best_fitness(i)
        particle_best_fitness(i) = particle_fitness(i);
        particle_best_position(i, :) = particle_position(i, :);
    end
    if particle_fitness(i) < global_best_fitness
        global_best_fitness = particle_fitness(i);
        global_best_position = particle_position(i, :);
    end
end

% Main loop
for iter = 1:max_iter
    % Update velocity and position
    w = w_max - (w_max - w_min) * iter / max_iter;
    for i = 1:particle_size
        r1 = rand();
        r2 = rand();
        if rand() < mutate_rate
            particle_velocity(i, :) = rand(1, dim) .* (upper_bound - lower_bound) * 0.1;
        else
            particle_velocity(i, :) = w * particle_velocity(i, :) + c1 * r1 * (particle_best_position(i, :) - particle_position(i, :)) + c2 * r2 * (global_best_position - particle_position(i, :));
        end
        particle_position(i, :) = particle_position(i, :) + particle_velocity(i, :);
        particle_position(i, :) = max(particle_position(i, :), lower_bound);
        particle_position(i, :) = min(particle_position(i, :), upper_bound);
    end
    
    % Evaluate fitness
    for i = 1:particle_size
        particle_fitness(i) = fitness_func(particle_position(i, :));
        if particle_fitness(i) < particle_best_fitness(i)
            particle_best_fitness(i) = particle_fitness(i);
            particle_best_position(i, :) = particle_position(i, :);
        end
        if particle_fitness(i) < global_best_fitness
            global_best_fitness = particle_fitness(i);
            global_best_position = particle_position(i, :);
            no_improve = 0;
        else
            no_improve = no_improve + 1;
        end
    end
    
    % Check for stagnation
    if no_improve >= max_no_improve
        % Mutate particles
        for i = 1:particle_size
            particle_position(i, :) = rand(1, dim) .* (upper_bound - lower_bound) + lower_bound;
            particle_velocity(i, :) = rand(1, dim) .* (upper_bound - lower_bound) * 0.05;
            particle_fitness(i) = fitness_func(particle_position(i, :));
            if particle_fitness(i) < particle_best_fitness(i)
                particle_best_fitness(i) = particle_fitness(i);
                particle_best_position(i, :) = particle_position(i, :);
            end
            if particle_fitness(i) < global_best_fitness
                global_best_fitness = particle_fitness(i);
                global_best_position = particle_position(i, :);
            end
        end
        no_improve = 0;
    end
    
    % Display progress
    fprintf('Iteration %d: Best fitness value = %f\n', iter, global_best_fitness);
end

% Return best value and position
best_value = global_best_fitness;
best_position = global_best_position;
end

其中，参数 `fitness_func` 是要求解的函数，参数 `dim` 是变量 X 的维度，`lower_bound` 和 `upper_bound` 是 X 各维度的下界和上界，`max_iter` 是迭代次数。函数返回最小值和最小值对应的 X。

请注意，这只是一个简单的实现，可能需要根据具体问题进行调整。