麻雀搜索算法python代码

以下是麻雀搜索算法的Python代码：

import math

def sparrow_search(fitness_func, dim, lb, ub, max_iter, num_sparrows, pa, delta_max):
    # 初始化麻雀群体
    sparrow_pos = [lb + (ub - lb) * np.random.random(dim) for i in range(num_sparrows)]
    sparrow_fitness = [fitness_func(sparrow_pos[i]) for i in range(num_sparrows)]
    sparrow_v = np.zeros((num_sparrows, dim))

    # 记录最优位置和适应度
    best_sparrow_pos = sparrow_pos[np.argmin(sparrow_fitness)]
    best_sparrow_fitness = min(sparrow_fitness)

    # 迭代寻优
    for iter in range(max_iter):
        # 更新每个麻雀的速度和位置
        for i in range(num_sparrows):
            # 计算每个麻雀对最优解的影响力
            attract_to_best = pa * np.random.random() * (best_sparrow_pos - sparrow_pos[i])
            # 计算每个麻雀之间的相互作用力
            interact = np.zeros(dim)
            for j in range(num_sparrows):
                if i != j:
                    diff = sparrow_pos[j] - sparrow_pos[i]
                    dist = np.linalg.norm(diff)
                    interact += (sparrow_pos[j] - sparrow_pos[i]) * math.exp(-dist) / dist
            # 计算每个麻雀的随机扰动项
            rand_vec = delta_max * (2 * np.random.random(dim) - 1)
            # 更新速度和位置
            sparrow_v[i] = sparrow_v[i] + attract_to_best + interact + rand_vec
            sparrow_pos[i] = sparrow_pos[i] + sparrow_v[i]
            
            # 限制麻雀位置不超过边界
            sparrow_pos[i] = np.clip(sparrow_pos[i], lb, ub)

            # 更新最优解
            fitness_i = fitness_func(sparrow_pos[i])
            if fitness_i < best_sparrow_fitness:
                best_sparrow_fitness = fitness_i
                best_sparrow_pos = sparrow_pos[i]

    return (best_sparrow_pos, best_sparrow_fitness)

其中，fitness_func 是适应度函数，dim 是问题维数，lb 和 ub 是问题的边界，max_iter 是最大迭代次数，num_sparrows 是麻雀数量，pa 是吸引因子，delta_max 是扰动范围。