Abstract: To research the scheduling problem of shampoo industry,the simulation method was used.Combined with the actual production problems of a shampoo production enterprise,a simulation model based on real-world manufacturing process was built by using Agent-based modeling of AnyLogic.In this model,production tasks and machines in each stage were regarded as Agent.Some scheduling rules were determined to schedule production tasks onto machine.Owing to the various complicated constraints in the actual production problems,it was difficult to directly use some traditional scheduling rules.Based on Shortest Processing Time First (SPT) and Longest Processing Time First (LPT  ) rules,some new dispatching rules were proposed,which could not only cut down the processing time,but also balance the workload of each machine.By analyzing the production data of each machine,some suggestions were put forward.

Key words: hybrid flow shop, simulation, Agent, rule scheduling method, shampoo industry

摘要： 针对液体日用品的生产调度问题,利用仿真的方法对其进行研究。结合某液体洗护用品生产企业的实际生产问题,借助AnyLogic中基于Agent的建模技术,搭建了该企业的生产线仿真模型。将生产任务与各阶段的机器视为Agent实体,根据所采用的调度规则将任务Agent和各阶段机器Agent整合成一个整体,进行相互协调并完成生产。由于实际生产问题带有各种复杂的条件约束,一些传统的调度规则很难直接运用于该问题之中。因此,在最短作业时间优先规则(SPT)和最长作业时间优先规则 (LPT)基础上,结合所研究的具体问题,提出了循环遍历最短作业时间规则/循环遍历最长作业时间规则(CSPT/CLPT)和机器免切换且最短作业时间优先规则/机器免切换且最长作业时间优先规则(FSPT/FLPT),分别应用于第一生产阶段和第二生产阶段。仿真结果表明,新提出的规则能在一定程度上减少最大完工时间并均衡各机器上的工作量。最后,分析了基于新调度方案得到的生产数据,为该液体洗护用品生产企业提出了相应的改进建议。

关键词: 混合流水车间, 仿真, 智能体, 规则调度方法, 液体日用品