基于遗传算法航班优化代码

遗传算法可以用于航班优化问题，以下是一个基于遗传算法的航班优化代码示例：

import random

# 航班信息（起飞时间、到达时间、出发地、目的地）
flights = [('BA001', '08:00', '10:00', 'LHR', 'JFK'),
           ('BA002', '12:00', '14:00', 'JFK', 'LHR'),
           ('BA003', '09:00', '12:00', 'LHR', 'CDG'),
           ('BA004', '11:00', '14:00', 'CDG', 'LHR'),
           ('BA005', '10:00', '12:00', 'LHR', 'FRA'),
           ('BA006', '13:00', '15:00', 'FRA', 'LHR'),
           ('BA007', '10:00', '13:00', 'LHR', 'MAD'),
           ('BA008', '14:00', '17:00', 'MAD', 'LHR'),
           ('BA009', '11:00', '13:00', 'LHR', 'AMS'),
           ('BA010', '14:00', '16:00', 'AMS', 'LHR')]

# 目标函数，评估航班安排的优劣
def fitness(solution):
    total_time = 0
    for i in range(len(solution)-1):
        flight1 = flights[solution[i]]
        flight2 = flights[solution[i+1]]
        if flight1[3] != flight2[2]:
            total_time += 2  # 转机时间为2小时
        total_time += (int(flight2[2][:2]) - int(flight1[1][:2]))  # 飞行时间
    return -total_time  # 目标是最小化总时间

# 遗传算法
def genetic_algorithm(population_size, fitness_func, gene_size, generation_num):
    # 初始化种群
    population = [[i for i in range(gene_size)] for _ in range(population_size)]
    for i in range(population_size):
        random.shuffle(population[i])
    # 迭代
    for generation in range(generation_num):
        # 评估种群
        scores = [fitness_func(solution) for solution in population]
        # 选择
        elites = []
        for i in range(population_size // 5):  # 选择前20%的精英个体
            max_index = scores.index(max(scores))
            elites.append(population[max_index])
            del population[max_index]
            del scores[max_index]
        # 交叉
        for i in range(population_size - len(elites)):
            parent1 = random.choice(population)
            parent2 = random.choice(population)
            child = [0] * gene_size
            start = random.randint(0, gene_size-1)
            end = random.randint(start, gene_size-1)
            child[start:end+1] = parent1[start:end+1]
            for j in range(gene_size):
                if parent2[j] not in child:
                    for k in range(gene_size):
                        if child[k] == 0:
                            child[k] = parent2[j]
                            break
            population.append(child)
        # 精英保留策略
        population = elites + population[:population_size-len(elites)]
    # 返回最优解
    scores = [fitness_func(solution) for solution in population]
    max_index = scores.index(max(scores))
    return population[max_index]

# 调用遗传算法寻找最优解
solution = genetic_algorithm(population_size=100, fitness_func=fitness, gene_size=len(flights), generation_num=1000)
print(solution)

在上面的代码中，我们定义了一个航班信息列表 `flights`，每个元素包含航班号、起飞时间、到达时间、出发地和目的地。然后定义了一个目标函数 `fitness`，用于评估航班安排的优劣，目标是最小化总时间。接着定义了一个遗传算法函数 `genetic_algorithm`，其中包括种群初始化、迭代、选择、交叉和精英保留策略等操作。最后调用遗传算法函数寻找最优解，并将结果打印出来。

需要注意的是，这只是一个简单的示例代码，实际应用中还需要考虑诸如航班延误、航班取消等情况，以及更加复杂的约束条件和目标函数。