小学排课优化：使用遗传算法和Python代码生成最佳排课表

下面是使用pandas库和遗传算法来进行排课的Python代码：

import pandas as pd
import numpy as np
import random

# 定义班级、课程和老师的数量
num_classes = 6
num_courses = 7
num_teachers = 11

# 创建一个空的排课表
schedule = pd.DataFrame(index=range(num_classes), columns=range(num_courses))

# 定义课程和老师的列表
courses = ['语文', '数学', '英语', '体育', '物理', '化学', '生物']
teachers = ['A老师', 'B老师', 'C老师', 'D老师', 'E老师', 'F老师', 'G老师', 'H老师', 'I老师', 'J老师', 'K老师']

# 初始化遗传算法的参数
num_generations = 100
population_size = 20
mutation_rate = 0.01

# 定义适应度函数，评估排课的质量
def fitness(schedule):
    # 初始化适应度值为0
    score = 0
    
    # 计算每个班级每天的课程数量
    daily_count = schedule.apply(lambda x: x.value_counts(), axis=1).fillna(0)
    
    # 检查每个班级每天的课程数量是否符合规定
    for i in range(num_classes):
        for j in range(5):
            if daily_count.loc[i, j] > 3:
                score -= 1
    
    # 检查同一课程是否连续上
    for i in range(num_classes):
        for j in range(num_courses - 1):
            if schedule.loc[i, j] == schedule.loc[i, j+1]:
                score -= 1
    
    # 检查同一课程每天是否出现3次
    for i in range(num_classes):
        for j in range(5):
            if schedule.loc[i, j:j+2].value_counts().max() >= 3:
                score -= 1
    
    return score

# 定义交叉函数，用于生成下一代
def crossover(parent1, parent2):
    # 随机选择交叉点
    crossover_point = random.randint(1, num_courses - 2)
    
    # 生成子代
    child = parent1.copy()
    child.iloc[:, crossover_point:] = parent2.iloc[:, crossover_point:]
    
    return child

# 定义变异函数，用于引入随机变异
def mutate(schedule):
    for i in range(num_classes):
        for j in range(num_courses):
            if random.random() < mutation_rate:
                schedule.iloc[i, j] = random.choice(teachers)
    
    return schedule

# 初始化种群
population = []
for _ in range(population_size):
    individual = schedule.copy()
    individual.applymap(lambda x: random.choice(teachers) if pd.isna(x) else x)
    population.append(individual)

# 进化过程
for generation in range(num_generations):
    # 计算每个个体的适应度值
    fitness_scores = [fitness(individual) for individual in population]
    
    # 选择适应度值较高的个体
    selected_indices = np.argsort(fitness_scores)[-population_size:]
    selected_population = [population[i] for i in selected_indices]
    
    # 生成下一代
    next_population = []
    for _ in range(population_size):
        parent1 = random.choice(selected_population)
        parent2 = random.choice(selected_population)
        child = crossover(parent1, parent2)
        child = mutate(child)
        next_population.append(child)
    
    population = next_population

# 输出最优个体
best_individual = max(population, key=fitness)
best_individual.to_excel('schedule.xlsx', index=False)

这段代码使用遗传算法进行排课，通过定义适应度函数来评估排课的质量，然后使用交叉和变异操作生成下一代个体，最终输出最优个体的排课结果到Excel文件中。