基于非参数统计量的多元控制图研究  

Research on Multivariate Control Chart Based on Nonparametric Statistics

作　　者：郭佳晟 刘以建[1] GUO Jia-sheng;LIU Yi-jian(School of Logistics Engineering,Shanghai Maritime University,Shanghai 201306,China)

机构地区：[1]上海海事大学物流工程学院,上海201306

出　　处：《组合机床与自动化加工技术》2021年第9期122-124,129,共4页Modular Machine Tool & Automatic Manufacturing Technique

摘　　要：多元控制图是现代工业生产中常用的质量控制方法。传统控制图需要假定数据的分布情况,常为多元正态分布,但在实际生产过程中过程数据的分布情况难以作出简单假定。针对这一情况,需要采用非参数(Nonparametric)控制图,即不依赖数据分布就可以对分布的参数作出判断。提出结合Wilcoxon秩和检验和Ansari-Bradley检验的非参数控制图(WAB控制图),同时监控分布的位置和尺度参数,放大监控效果。通过蒙特卡洛(Monte Carlo)模拟法分析了WAB控制图在过程失控时的性能表现,并与T2控制图进行比较。结果表明WAB控制图具有良好的稳定性和灵敏性。Multivariate control chart is a commonly used quality control method in modern industrial production.Traditional control charts need to assume the distribution of data,which is often a multivariate normal distribution,but it is difficult to make simple assumptions about the distribution of process data in the actual production process.In response to this situation,it is necessary to use a nonparametric control chart,which means that the parameters of the distribution can be judged without relying on the data distribution.This paper proposes a non-parametric control chart combining Wilcoxon rank-sum test and Ansari-Bradley test(WAB control chart),and monitors the position and scale parameters of the distribution at the same time to enlarge the monitoring effect.The performance of the WAB control chart when the process is out of control is analyzed by Monte Carlo simulation method,and compared with the T2control chart.The results show that the WAB control chart has good stability and sensitivity.

关 键 词：多元控制图 非参数检验 HotellingT2控制图 非参数方法 

分 类 号：TH16[机械工程—机械制造及自动化] TG506[金属学及工艺—金属切削加工及机床]