Abstract: The characteristics of mould heat-treatment shop production scheduling were summarized.Under the dynamic arrival of jobs,the flow-shop batch scheduling problem with two-stage batch machine and constant processing time was researched.Aiming at the minimum makespan,a mixed integer programming mathematic model was built.The problem was proved as NP-hard and the lower bound of the optimal objective function was given.An optimization algorithm for solving large-scale problems was constructed. The upper and lower bounds of batches number were solved,and the mathematical model was designed to solve the optimal solution for given batches number.By comparing each solution obtained from the given batches number,the global optimal solution was obtained.For small-scale problem,through comparing with the solution obtained from the traditional branch and bound method,the algorithm was proved to get the optimal solution and have shorter computation time.For large-scale problems,multiple sets of data were randomly generated in the different production conditions to validate the algorithm and the results show the effectiveness of it.

Key words: batch scheduling, mould, heat-treatment, flow-shop

摘要： 总结模具热处理车间生产调度的特点,在工件动态到达的情况下,研究了两阶段均为批处理机、各批次加工时间为定值的流水车间批调度问题。以最小化最大完成时间为目标建立混合整数规划数学模型,证明问题是NP难并给出最优目标函数值的下界,构建并证明一种求解大规模问题最优解的算法。算法首先求解分批数量上下界,然后构造数学模型并求解在给定分批数量下对应的最优解,最后通过对比各给定分批数量下得到的解并得到全局最优解。对于小规模问题,通过与传统分枝定界法得到的最优解相比较,说明所提算法能得到问题的最优解且计算时间更短;对于大规模问题,在不同生产状况下随机产生多组数据、对所提算法进行验证,结果显示该算法是有效的。

关键词: 批调度, 模具, 热处理, 流水车间