用遗传算法写一个多参量离散变量全局优化算法代码

以下是一个使用遗传算法进行多参数离散变量全局优化的示例代码：

import random

# 定义目标函数（这里以一个简单的多参数离散变量函数为例）
def objective_function(x):
    return sum(x)

# 定义遗传算法参数
population_size = 100  # 种群大小
num_generations = 100  # 迭代代数
mutation_rate = 0.01  # 变异率
num_variables = 5  # 变量个数
variable_domain = [0, 1, 2, 3, 4]  # 变量取值范围

# 初始化种群
population = []
for _ in range(population_size):
    individual = [random.choice(variable_domain) for _ in range(num_variables)]
    population.append(individual)

# 遗传算法主循环
for generation in range(num_generations):
    # 计算适应度
    fitness_scores = [objective_function(individual) for individual in population]

    # 选择
    selected_population = []
    for _ in range(population_size):
        # 锦标赛选择，选择两个个体中适应度更高的个体
        tournament_size = 5
        tournament_individuals = random.sample(list(enumerate(population)), tournament_size)
        tournament_fitness_scores = [fitness_scores[i] for (i, _) in tournament_individuals]
        selected_individual = tournament_individuals[tournament_fitness_scores.index(max(tournament_fitness_scores))][1]
        selected_population.append(selected_individual)

    # 交叉
    crossover_population = []
    for _ in range(population_size):
        parent1, parent2 = random.sample(selected_population, 2)
        crossover_point = random.randint(1, num_variables - 1)
        child = parent1[:crossover_point] + parent2[crossover_point:]
        crossover_population.append(child)

    # 变异
    mutated_population = []
    for individual in crossover_population:
        if random.random() < mutation_rate:
            mutation_point = random.randint(0, num_variables - 1)
            individual[mutation_point] = random.choice(variable_domain)
        mutated_population.append(individual)

    # 更新种群
    population = mutated_population

# 打印最优解
best_individual = max(population, key=objective_function)
print("最优解：", best_individual)
print("最优值：", objective_function(best_individual))

这个示例代码中，目标函数定义为一个简单的求和函数，种群大小为100，迭代代数为100，变异率为0.01，变量个数为5，变量取值范围为[0, 1, 2, 3, 4]。种群初始化为随机的离散变量取值。然后，遗传算法进行多轮的选择、交叉和变异操作，最终得到最优解和最优值