手术排程问题：有20个手术间，有40种术式，10个医疗组，已知各医疗组开展相应术式的平均时间，目标函数为：最大化手术室黄金时间利用率并且最小化手术间用时标准差；提供启发式算法python具体代码

这是一个比较复杂的排程问题，需要考虑到多个因素。以下是一个基于遗传算法的启发式算法的 Python 代码实现：

import random
import numpy as np

# 定义问题参数
num_surgeries = 40 # 手术数量
num_rooms = 20 # 手术间数量
num_teams = 10 # 医疗组数量
time_matrix = np.random.rand(num_surgeries, num_teams) # 医疗组开展相应术式的平均时间矩阵

# 定义遗传算法参数
population_size = 100 # 种群大小
mutation_rate = 0.1 # 变异率
crossover_rate = 0.8 # 交叉率
generations = 100 # 迭代次数

# 初始化种群
def init_population():
    population = []
    for i in range(population_size):
        chromosome = np.random.permutation(num_surgeries)
        population.append(chromosome)
    return population

# 评估适应度
def evaluate_fitness(chromosome):
    # 计算手术间用时标准差
    times = np.zeros(num_rooms)
    for i in range(num_surgeries):
        room = np.argmin(times)
        times[room] += time_matrix[chromosome[i], room]
    std_dev = np.std(times)
    # 计算黄金时间利用率
    avg_time = np.mean(time_matrix)
    utilization = np.sum(time_matrix[chromosome]) / (num_rooms * avg_time)
    # 综合考虑
    fitness = utilization / std_dev
    return fitness

# 选择父代
def select_parents(population):
    fitnesses = [evaluate_fitness(chromosome) for chromosome in population]
    total_fitness = sum(fitnesses)
    probabilities = [fitness / total_fitness for fitness in fitnesses]
    parents = np.random.choice(population, size=2, p=probabilities, replace=False)
    return parents

# 单点交叉
def single_point_crossover(parents):
    crossover_point = random.randint(1, num_surgeries - 1)
    child1 = np.concatenate((parents[0][:crossover_point], parents[1][crossover_point:]))
    child2 = np.concatenate((parents[1][:crossover_point], parents[0][crossover_point:]))
    return child1, child2

# 变异
def mutate(chromosome):
    if random.random() < mutation_rate:
        index1, index2 = random.sample(range(num_surgeries), 2)
        chromosome[index1], chromosome[index2] = chromosome[index2], chromosome[index1]
    return chromosome

# 遗传算法主函数
def genetic_algorithm():
    population = init_population()
    for generation in range(generations):
        # 选择父代并产生子代
        new_population = []
        for i in range(population_size // 2):
            parents = select_parents(population)
            if random.random() < crossover_rate:
                child1, child2 = single_point_crossover(parents)
                child1 = mutate(child1)
                child2 = mutate(child2)
                new_population.append(child1)
                new_population.append(child2)
        # 选择下一代种群
        population = new_population
    # 返回最优解
    fitnesses = [evaluate_fitness(chromosome) for chromosome in population]
    best_index = np.argmax(fitnesses)
    return population[best_index]

这段代码中，我们定义了一个 `genetic_algorithm` 函数作为遗传算法的主函数，其中包括了种群初始化、适应度评估、父代选择、交叉、变异等步骤。在每一代种群中，我们选择父代并产生子代，然后通过遗传算法的操作来生成下一代种群。最后，我们返回最优解作为排程结果。

需要注意的是，这里的适应度函数考虑了两个因素：最大化手术室黄金时间利用率并且最小化手术间用时标准差。我们通过综合考虑这两个因素来评估每个染色体的适应度，并选择适应度最高的染色体作为最优解。