晶圆制造物料运输系统性能分析建模方法

Performance analysis modeling approach for wafer fabrication material handling system

为在自动化物料搬运系统设计阶段快速有效地分析该系统的性能,提出一种扩展Markov模型的建模方法,该方法减少了自动化物料搬运系统性能分析模型的规模,避免了模型的状态空间规模爆炸。采用某晶圆制造企业的实际生产数据进行测试,并与仿真分析方法比较。结果表明,扩展Markov模型与仿真模型在自动化物料搬运系统搬运量、运输小车利用率和空载小车到达平均时间间隔等指标方面的相对误差值为96%,处于[-8%,10%]的区间范围;与仿真模型相比,扩展Markov模型的性能分析效率提高了600倍以上。由此验证了所提扩展Markov模型建模方法的有效性。

To rapidly and effectively analyze the performances of Automated Material Handling System （AMHS） dur- ing the design phase, an Extended Markov Chain Model （EMCM） based modeling approach was proposed. The scale of AMHS＇s performance analytical model was reduced and the state space explosion phenomenon of the model was avoided. With production data from a 300ram semiconductor wafer fabrication line, the proposed modeling ap- proach was compared with simulation analytical approach. The result showed that the 96% relative error value of performance indexes between EMCM and simulation model such as AMHS＇s movement, vehicle＇s utilization ratio and interval of vehicle＇s mean arrival time within the range of [--8%, 10%]. The analytical efficiency of EMCM was 600 times higher than the simulation model. It demonstrated that the proposed EMCM based modeling approach was effective.