用遗传算法编写AGV小车路径规划问题的代码

以下是一个简单的AGV小车路径规划问题的遗传算法实现的代码：

import random

# 遗传算法参数
POP_SIZE = 50  # 种群大小
MUTATION_RATE = 0.1  # 变异概率
CROSSOVER_RATE = 0.8  # 交叉概率
N_GENERATIONS = 20  # 迭代次数

# AGV小车路径规划问题参数
START_POINT = (0, 0)  # 起点坐标
END_POINT = (10, 10)  # 终点坐标
OBSTACLES = [(2, 2), (3, 3), (5, 7)]  # 障碍物坐标列表

# 随机生成一个个体
def generate_individual():
    return [(random.randint(-1, 1), random.randint(-1, 1)) for _ in range(10)]

# 计算个体的适应度，即到达终点的距离
def calculate_fitness(individual):
    x, y = START_POINT
    for step in individual:
        x += step[0]
        y += step[1]
        if (x, y) in OBSTACLES or x < 0 or y < 0:
            return 0  # 遇到障碍物或出界，适应度为0
    return (x - END_POINT[0]) ** 2 + (y - END_POINT[1]) ** 2

# 选择操作
def selection(population):
    fitness = [calculate_fitness(individual) for individual in population]
    total_fitness = sum(fitness)
    roulette_wheel = [sum(fitness[:i+1]) / total_fitness for i in range(len(population))]
    new_population = []
    for i in range(len(population)):
        r = random.random()
        for j in range(len(population)):
            if roulette_wheel[j] >= r:
                new_population.append(population[j])
                break
    return new_population

# 交叉操作
def crossover(parent1, parent2):
    if random.random() < CROSSOVER_RATE:
        child1 = parent1[:5] + parent2[5:]
        child2 = parent2[:5] + parent1[5:]
        return child1, child2
    else:
        return parent1, parent2

# 变异操作
def mutation(individual):
    new_individual = individual.copy()
    for i in range(len(new_individual)):
        if random.random() < MUTATION_RATE:
            new_individual[i] = (random.randint(-1, 1), random.randint(-1, 1))
    return new_individual

# 主函数
def main():
    population = [generate_individual() for _ in range(POP_SIZE)]
    for generation in range(N_GENERATIONS):
        population = selection(population)
        next_population = []
        for i in range(POP_SIZE // 2):
            parent1 = random.choice(population)
            parent2 = random.choice(population)
            child1, child2 = crossover(parent1, parent2)
            child1 = mutation(child1)
            child2 = mutation(child2)
            next_population.extend([child1, child2])
        population = next_population
    best_individual = max(population, key=calculate_fitness)
    print("Best individual:", best_individual)
    print("Fitness:", calculate_fitness(best_individual))

if __name__ == '__main__':
    main()

该代码使用遗传算法来解决AGV小车路径规划问题，其中种群大小为50，变异概率为0.1，交叉概率为0.8，迭代次数为20。在每一次迭代中，使用轮盘赌选择算子选出下一代个体，使用单点交叉算子和随机变异算子对选出的个体进行交叉和变异操作。最终输出最优个体及其适应度。