摘要
介绍了建立工艺监控模块(PCM)参数数值模型的多元回归方法。多元回归是一种多元统计的处理方法,由该方法得到的一组数值模型中的每一个,都能够与数据源良好拟合。利用单值模拟器,取得了统计性的模拟结果,应用多元回归法得到了回归模型,它在一定范围内能有效地替代耗时的数值模拟。该方法可广泛应用于IC工艺开发、优化、诊断和电子制造自动化等方面。
A multivariate regression method to build statistic models for PCM parameters is presented,which is one of the multivariate statistical methods Each of the models obtained with this method can fit very well with the source data The proposed method is especially suitable for complicated systems,such as VLSI manufacturing system Statistical results are obtained from the simulator that gives a definite value The models obtained can be used efficiently,to a certain extent,to replace the time consuming numerical simulations So,the method is useful in developing,optimizing,and diagnosing VLSI processes,and in manufacture automation
出处
《微电子学》
CAS
CSCD
北大核心
2003年第4期309-312,共4页
Microelectronics
关键词
PCM
工艺监控模块
多元回归模型
集成电路
电子制造
PCM(Process Control Module)
Multivariate regression
Process simulation
Device simulation
