| Summary: | 碩士 === 國立高雄第一科技大學 === 運籌管理研究所 === 104 === Re-entrant flow is a common process in IC and PCB industries. The flow sop with the re-entrant flow is much more complex than the general flow shop, resulting in decrease in throughput rate (TH), increase in Work-in-process (WIP) etc. This study mainly constructs a flow shop with the re-entrant flow model and evaluates two two-objective problems with TH and WIP as well as TH and completion time variation, respectively, under both CONWIP control mechanism and Multi-CONWIP control mechanism. First, the multi-objective problem with TH and WIP is measured in terms of control strategies of the two control mechanisms scheduled by system simulation and multi-objective particle swarm optimization algorithm (MOPSO). The performance of two control mechanisms is then discussed. Second, the multi-objective problem with TH and completion time variation is evaluated by system simulation as well as MOPSO, adjusting the control strategies of the two control mechanisms, to examine the performance of these two control mechanisms. The experimental results show that Multi-CONWIP with appropriate control strategies, scheduled by system simulation as well as MOPSO, is able to more effectively control the flow sop with the re-entrant flow than the CONWIP in three performance indexes: TH, WIP, and completion time variation. Subsequently, the results shown in the study can be applied to the real-word the flow sop with the re-entrant flow so as to effectively control TH and completion time variation in the similar level of TH; namely, managers can further reduce uncertainty of the production process by the results shown in the study and enhance their competitiveness
|
|---|
