一般而言，統計製程管制係以製程能力分析監控製程的品質是否符合工程規格或達到客戶要求的水準，而製程能力指標常用於評估製程的風險。
　　本研究的目的係針對一個多變量卜瓦松製程，在考量品質特性間之相關性及各品質特性對產品影響程度不同情況下，提出一個能正確反映製程風險的新計數型製程能力指標。首先，假設一製程包含q個品質特性，而這q個品質特性之瑕疵數係服從多變量卜瓦松分配。我們藉由瑕疵分數統計量Dw來評估瑕疵數對於製程造成的風險，並利用望小特性下非常態製程能力指標之概念提出新的製程能力指標Cw。
　　接著，我們以數值計算方式評估比較在不同相關係數下當製程平均發生偏移時，各種製程能力指標之表現。結果發現無論在何種相關係數或權重改變之情形下，本研究提出之新製程能力指標Cw皆可正確合理的反映出瑕疵數對於製程所造成的風險。最後，我們以Jiang等(2002)所使用之電信通訊資料及長春化工絕緣紙汙點之嚴重程度為例進行數值實例驗證，說明我們提出之新製程能力指標Cw較能合理反映瑕疵數對多變量卜瓦松製程造成的風險，研究結果可供品管人員未來在評估計數型多重品質特性製程風險時之參考。

Process capability analysis (PCA) is frequently employed by manufacturers to evaluate whether the capability of process can meet the customer’s requirement or not. Practically, process capability indices (PCIs) are often used for assessing the manufacturing risk in most industries.
This paper aims to propose a new attribute process capability index by considering the correlation among different defect types and their degrees of influence on the final product for multivariate Poisson processes. Considering a process has q quality characteristics and the number of defects or q quality characteristics follows a multivariate Poisson distribution, we use the statistic of demerit scheme Dw to measure the risk of number of defects for a multivariate Poisson process. Then, the new process capability index Cw for the statistic Dw can be defined by adopting the concept of non-normal PCIs for the Smaller-the-better (STB) case.
In the numerical analysis, we compare the performance of various PCIs under different coefficients when process mean shifts occurred. The simulation results show that our proposed attribute process capability index can properly reflect the manufacturing risk of a multivariate Poisson process when different weights are assigned to the associated quality characteristics. Finally, two numerical examples with telecommunication data set mentioned in Jiang et al.(2002) and the black spots problem occurred in the insulation paper at Chang Chung Company are used to demonstrate the usefulness of our proposed capability index Cw.

1.Aslam, M., Rao, G. S., Ahmad, L. and Jun, C. H., A control chart for multivariate Poisson distribution using repetitive sampling. Journal of Applied Statistics, 44:1, pp. 123-136 (2016)
2.Borges, W. S. and Ho, L. L., A fraction defective based capability index. Quality and Reliability Engineering International, 17:6, pp. 447-458 (2001)
3.Chan, L. K., Cheng, S. W. and Spiring, F. A., A new measure of process capability: C_pm. Journal of QualityTechnology, 20:3, pp.162-175 (1988)
4.Clement, J. A., Process capability calculation for non-normal distributions. Quality Progress, 22, pp. 95-100 (1989)
5.Chang, P. L., and Lu, K. H., PCI Calculations for Any Shape of Distribution with Percentile. Quality World, technical section, September, pp. 110-114 (1994)
6.Chiu, J. E. and Kuo, T. I., Attribute control chart for multivariate Poisson distribution. Communications in Statistics – Theory and Methods, 37:1, pp. 146-158 (2008)
7.Dodge, H.F., A Method of Rating a Manufactured Product. Bell System Technical Journal, 7:2, pp. 350-368 (1928)
8.Gruska, G. F., Mirkhani, K., and Lamberson, L. R., Non-Normal Data Analysis. Multiface Publishing Co., Michigan (1989)
9.Gunter, B. H., The Use and Abuse of C_pk Part 2. Quality Progress, 22, pp. 108-109 (1989)
10.Holgate, P., Estimation for the bivariate Poisson distribution. Biometrika, 51:1-2, pp. 241-245 (1964)
11.Hoadley, B., The quality measurement plan (QMP). Bell System Technical Journal, 60:2, pp. 215-273 (1981)
12.He, S., He, Z. and Wang, G. A., CUSUM control charts for multivariate Poisson distribution. Communications in Statistics – Theory and Methods, 43:6, pp. 1192-1208 (2014)
13.Juran, J. M., Quality Control Handbook, McGraw-Hill, New York (1974)
14.Jiang, W., Au, S. T., Tsui, K. L. and Xie, M., Process monitoring with univariate and multivariate c-charts. Technical Report, The Logistics Institute, Georgia Tech, and The Logistics Institute-Asia Pacific. National University of Singapore (2002)
15.Kemp, C. D. and Kemp, A. W., Some properties of the ‘Hermite’ distribution. Biometrika, 52:3, pp. 381-394 (1965)
16.Kawamura, K., Direct calculation of maximum likelihood estimator for the bivariate Poisson distribution. Kodai Mathematical Journal, 7:2, pp. 211-221 (1984)
17.Kane, V. E., Process capability indices. Journal of Quality Technology. 18:1, pp. 41-52 (1986)
18.Li, C.I., Pan, J.N. and Huang, M.H., A new demerit control chart for monitoring the quality of multivariate Poisson processes. Journal of Applied Statistics, 46:4, pp. 680-699 (2019)
19.Montgomery, D.C., Introduction to statistical quality control, 6th Ed., John Wiley & Sons, NY, pp. 344-390 (2009)
20.Maiti, S.S., Saha, M., Nanda, A.K., On generalizing process capability indices. Quality Technology & Quantitative Management, 7:3, pp. 279-300 (2010)
21.Pearn, W. L., Lin, G. H., and Chen, K. S., Distributional and inferential properties of process capability indices. Journal of QualityTechnology, 24:4, pp.216-231 (1992)
22.Pearn, W. L., and Chen, K. S., Estimating Process Capability Indices for Non-Normal Pearsonian Populations. Quality and Reliability Engineering International, 11, pp. 86-388 (1995)
23.Pearn, W. L., and Chen, K. S., An Application of Non-normal Process Capability Indices. Quality and Reliability Engineering International, 13, pp. 355-360 (1997)
24.Perakis M. and Xekalaki E., A process capability index that is based on the proportion of conformance. Journal of Statistical Computation and Simulation, 72:9, pp. 707-718 (2002)
25.Vannman, K., A Unified Approach to Capability Indices. Statistical Sinica, 5, pp. 805-820 (1995)
26.蔡琬慈，加權卜瓦松分佈下計數型資料製程能力指標之研究，國立成功大學統計學研究所碩士論文。(2014)