模糊資料之品管圖架構

On construction of control charts for fuzzy data

李俊賢(Chun-Hsien Li)

指導教授 ： 楊敏生

中原大學/理學院/應用數學研究所/碩士(2004年)

https://doi.org/10.6840/cycu200400232

摘要

摘要
在這篇論文中，我們探討語意型資料的管制圖架構方法，利用Raz and Wang他們所提出的方法來架構品質管制圖，再與一般多變量p-chart 所用的Marcucci方法做靈敏度的比較。接著，我們並對Raz and Wang所提出的表示值及轉換的方法做完整的探討，針對L-R 型模糊資料的隸屬函數與幾何平均數的表示值，是否有更佳的品質管制圖的表現，另外也設法在不解模糊化的狀態下架構出來的品質管制圖是否也可以分辨出哪些有失控狀況的點。最後，雖然我們在靈敏度上Marcucci方法表現的較好，但是如果依是否會損失樣本中所給的訊息方面來說，Raz and Wang的方法處理語意型資料的確是有它的優點在。

關鍵字

品管圖

並列摘要

Abstract
In this paper, we discussed how to construct a control chart in fuzzy data. Using the approach which Raz and Wang proposed to construct control charts and compare Raz and Wang approach with the marcucci approach, which is more sensitive. Then, we discuss completely the representative value and two transform approaches that Raz and Wang purposed, and to see the representative value of geometric mean and the membership function of L-R type fuzzy data have the better outcome of the control charts or not. And try to construct control charts without convert the fuzzy set into scalars and to see that can identify the control charts where is out of control or not. Finally, although the marcucci approach is more sensitive, if as we consider the lose of information of sample, then the Raz and Wang approach is better to handle the fuzzy data.

並列關鍵字

control charts

參考文獻

[1] L.A. Zadeh, Fuzzy sets, Inform. Control, 1965 , Vol. 8, pp. 338-353.

[2] T. Raz and J.H. Wang, On the construction of control charts using linguistic variables, Int. J. Prod. Res., 1990 , Vol. 28, pp.477-487.

[3] M. Marcucci, Monitoring multinomial process. Journal of Quality Technology, 1985 , Vol. 17, pp. 86-91

[4] H. Taleb and M. Limam, On fuzzy and Probabilistic control charts, Int. J. Prod. Res. , 2002 , Vol. 40, pp. 2849-2863.

[5] A. J. Duncan, Quality Control and Industrial Statistics, 5th ed.,1986 , Homewood, IL, Richard D. Irwin.