基于改进变邻域搜索算法的成型机分批重调度优化  

Molding Machines Batch Rescheduling Optimization Based on Improved Variable Neighborhood Search

作　　者：徐立云[1] 程赞 宓宏 李爱平[1] XU Liyun;CHENG Zan;MI Hong;LI Aiping(School of Mechanical Engineering,Tongji University,Shanghai 200092,China;Zhejiang Quzhou Lianzhou Refrigerant Co.,Ltd.,Quzhou 324000,China)

机构地区：[1]同济大学机械与能源工程学院,上海200092 [2]浙江衢州联州致冷剂有限公司,浙江衢州324000

出　　处：《同济大学学报（自然科学版）》2020年第10期1460-1469,共10页Journal of Tongji University:Natural Science

基　　金：2017年工信部智能制造新模式项目;国家自然科学基金面上项目(51975417);国家科技重大专项(2011ZX04015-022)。

摘　　要：针对成型机故障和工单交货期提前两类事件,提出一种基于改进变邻域搜索算法的分批重调度方法,基于最小分批原则和非等量分批原则对工单进行批量划分,考虑重调度过程的稳定性与准时性,建立数学模型。设计一种改进的变邻域搜索算法(VNS),通过构建转移邻域和叠加邻域两种邻域结构,提高了搜索的收敛速度和寻优能力。最后以某磁性材料成型车间作为实例进行验证。结果表明,所提重调度方法能够在保证工单准时交付的基础上,提高成型机利用率,为工厂的实际生产决策提供参考。This paper proposed a batch rescheduling method based on improved variable neighborhood search algorithm for two types of events which were molding machine failure and early delivery,and the orders were batch-divided based on the minimum batching principle and non-equal batching principle.Then,a rescheduling mathematical model was established considering the process stability and punctuality.After that,an improved variable neighborhood search algorithm was designed,which had a faster convergence speed and a stronger global optimization ability by constructing transition neighborhood and superposition neighborhood.Finally,a magnetic material molding workshop is taken as an example to verify the results.The results show that the proposed rescheduling method can ensure on-time delivery of order,raise the utilization rate of molding machine,and provide a reference for the production decision of factory.

关 键 词：成型机 分批调度 重调度 变邻域搜索 准时交付 

分 类 号：TH186[机械工程—机械制造及自动化]