監控線性輪廓製程的多變量指數加權移動平均保證管制圖

A Guaranteed MEWMA Control Chart for Monitoring Linear Profiles

藍文萱(Lan, Wen Hsuan)

指導教授 ： 黃榮臣

國立清華大學/理學院/統計學研究所/碩士(2016年)

摘要

在許多工業製造的例子上，產品或製程的品質可以用一個反應變數以及一個或多個解釋變數的函數關係來描述，這種函數關係稱為輪廓製程，而管制圖是統計製程管制(statistical process control，簡稱SPC) 中最為廣泛使用的監控輪廓製程的方法。在本篇論文中，我們探討使用Zou, Tsung and Wang (2007) 的MEWMA (multivariate EWMA) 管制圖來監控一般線性輪廓製程優劣點，並對於實用上的缺點提出補助之道。首先，我們比較此管制圖和其它用來監控線性輪廓製程管制圖的表現，接著我們想要了解需要使用多少組第一階段的管制狀態資料才能使真實的(true) 管制狀態平均連串長度(ARL0) 達到名義的(nominal)ARL0。在一般情形下，通常沒有辦法蒐集到大量的第一階段資料，所以我們藉由拔靴法來調整管制界限，保證在調整後的管制界限之下，有一固定的機率使得真實的ARL0 大於名義的ARL0。最後，我們會用一個實際的例子來說明如何使用調整管制界限與非調整管制界限來監控線性輪廓製程。

關鍵字

線性輪廓製程 ； 多變量指數加權移動平均 ； 保證管制圖

並列摘要

In many industrial manufacturing processes, the quality of a process or product is represented by a relationship between the response variable and one or more explanatory variables. We call this relationship profile process. Control chart is the most widely used to monitor profile process in the statistical process control (SPC). In this article, we are going to investigate strengths and weakness of Zou, Tsung and Wang (2007) MEWMA chart which is used to monitor general linear profiles. Then provide a method to solve the practical shortcomings. First, we compare the performances between this chart and other charts which are used to monitor linear profiles. Then we want to know how many in-control phase I datasets can make true in-control average run length (ARL0) achieve nominal ARL0. In general case, usually can’t collected a lot of phase I datasets. So we adjust control limit by bootstrap, which can guarantee that use the adjusted control
limit have a fixed probability let true ARL0 more than nominal ARL0. At the end, a real example is used to illustrate how to use adjusted and unadjusted control limit for monitoring linear profiles.

並列關鍵字

MEWMA ； Linear Profile

參考文獻

[1] Aly, A. A., Mahmoud, M. A. and Woodall, W. H. (2015). ”A Comparison of the Performance of Phase II Simple Linear Profile Control Charts when Parameters are Estimated”. Communications in Statistics-Simulation and Computation 44, pp. 1432-1440.

[2] Efron, B. (1979). ”Bootstrap Methods: Another Look at the Jackknife”. The Annals of Statistics 7, pp. 1-26.

[3] Gandy, A. and Kvaløy, J. T. (2013). ”Guaranteed Conditional Performance of Control Charts via Bootstrap Methods”. Scandinavian Journal of Statistics 40, pp. 647-668.

[4] Jones, M. A. and Steiner, S. H. (2012). ”Assessing the Effect of Estimation Error on Risk-adjusted CUSUM Chart Performance”. International Journal for Quality in Health Care 24(2), pp. 176-181.

[5] Kang, L. and Albin, S. L. (2000). ”On-Line Monitoring When the Process Yields a Linear Profiles”. Journal of Quality Technology 32, pp. 418-426.