基于遗传算法的混合流水车间批量调度研究

针对混合流水车间批量调度问题(hybrid flow-shop scheduling problem with lot-streaming, HFSPLS), 文章设计改进遗传算法(improved genetic algorithm, IGA)。以最小化最大完工时间为目标函数, 建立混合流水车间批量调度数学模型; 结合先到先服务(first come first serve, FCFS)原则设计仅针对第1阶段的矩阵编码和解码策略; 通过融合精英保留策略和轮盘赌法的选择操作、自适应交叉变异概率增强算法全局搜索能力。最终通过相关案例测试和算法比较验证了该算法的有效性和优越性。

Aiming at the hybrid flow-shop scheduling problem with lot-streaming(HFSP_LS), an improved genetic algorithm(IGA) with adaptive crossover probability and mutation probability is designed. Firstly, the mathematical model of batch scheduling in hybrid flow shop is established by minimizing the maximum completion time as the objective function. Secondly, the matrix encoding and decoding strategies for the first phase are designed according to the first come first serve(FCFS) principle. Then, the global search capability of the algorithm is enhanced by integrating the selective operation of elite retention strategy and roulette method, adaptive crossover probability and multi-fragment crossover operation, adaptive mutation probability and multi-fragment mutation operation. Finally, the effectiveness and superiority of the proposed algorithm are verified by case test and algorithm comparison.