Estimation of the change point of a uniform process using the EWMA chartand MLE

被引：0

作者：

Shao, Yuehjen E. ^{[1
]}

Hou, Chia-Ding ^{[1
]}

机构：

[1] Department of Statistics and Information Science, Graduate Institute of Applied Statistics, Fu Jen Catholic University, 510 Chung Cheng Rd., Hsinchuang, Taipei County 24205, Taiwan

来源：

ICIC Express Letters | 2009年 / 3卷 / 03期

关键词：

D O I：

暂无

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

The statistical process control (SPC) charts have successfully beenimplemented in industries for more than seventy years. It is a must to searchfor the root causes of a process disturbance when an out-of-control signal istriggered by the SPC chart. The correct and quick determination of the rootcauses is able to significantly improve the process. However, it is a timeconsuming task for this determination if we do not know the change point orstarting time of a disturbance. Consequently, the determination of the changepoint of a process becomes a very important research issue. In contrast to thetypical assumption of normality of a process, this study considers a processwhich is uniformly distributed. This study combines the maximum likelihoodestimation (MIE) and the exponentially weighted moving average (EWMA) chart toestimate the change point of a uniform process. In addition, due to the commonseen in chemical processes, this study considers the case of the sample size ofone. This study derives the exact control limits for a uniform process, andaddresses the fruitful results of the proposed approach by performing a seriesof computer simulations. ICIC International © 2009.

引用

页码：451 / 456