An overlapping decomposition method for performance analysis of the two-loop closed production system in semiconductor assembly factory

摘要

Multi-loop closed production systems widely exist in different industries such as semiconductor assembly and test factories. Measuring the system performance is very difficult because of the complex interactions between the loops. It is fairly significant for the design, improvement and control of the production systems. In this article, we propose an approximation method based on the overlapping decomposition philosophy and introduce the iterative procedures to estimate a two-loop closed production system's performance. The accuracy of the estimates is verified by the experiments and the results show that the approximation method provides good results. We also illustrate the application of the method in a semiconductor factory. First, a pre-treatment aggregation procedure was performed to make the real data correspond with the algorithm. Second, the method was used to measure the performance of the system and the comparison between the analytical results and the practical results are illustrated. Finally, we introduce the case study of improvement application in the semiconductor assembly factory. The case study shows that the proposed method is applicable to practical system performance evaluation, system design and improvement.