二元偏斜常態測量之兩階段管制圖經濟設計

Economic Design of Two Stage Control Charts with Bivariate Skew Normal Measurements.

呂俊佑(CHUN-YOU LU)

指導教授 ： 蘇南誠

國立臺北大學/商學院/統計學系/碩士(2016年)

摘要

在很多的實例中, 需要花費較高的成本來監控主要的品質特性,我們稱之為主要監測變數, 但是監控其替代變數則較具有經濟效益。為了處理使用不對稱性的主要監測變數及替代變數所建構之兩階段管制圖, 我們考慮服從二元偏斜常態分配的製程變數。當主要監測變數和替代變數間的關聯可被明確闡述時, 監測主要變數之管制圖可替代為交替監控主要監測變數或是其替代變數的管制圖。本論文將在兩階段管制圖設計和使用馬可夫鍊逼近法的經濟設計的架構下, 探討主要變數和替代變數服從二元偏斜常態分佈的情況。

關鍵字

主要監測變數 ； 替代變數 ； 偏斜常態分配 ； 兩階段管制圖設計 ； 馬可夫鍊逼近法 ； 經濟設計

並列摘要

In many instances, the cost is high to monitor primary quality characteristics
called performance variable, but it could be economical to monitor its sur-
rogate. To cover asymmetric processes for two-stage charting methods using both performance and surrogate variables, bivariate skew normal distribution
is considered as the underlying distribution of process variables. When the
correlation relationship between the performance variable and its surrogate
is specified, a charting procedure to monitor either the performance variable
or its surrogate in an alternating fashion rather than monitoring the per-
formance variable alone is proposed. The proposed two-stage control charts
are constructed under an economic design using Markov chain approach. Its
advantages over the existing methods will be presented.

並列關鍵字

Performance variable ； Surrogate variable ； Skew-normal distribution ； Two-stage charting ； Economic design ； Markov chain approach

參考文獻

Azzalini, A., and Capitanio, A. (1999). Statistical applications of the

Su, N. C., Chiang, J. Y., Chen, S. C., Tsai, T. R., and Shyr, Y. (2014).

Azzalini, A., and Dalla Valle, A. (1996). The multivariate skew-normal

multivariate skew normal distribution. Journal of the Royal Statistical

Costa, A. F., and Rahim, M. A. (2004). Joint X and R charts with two-