Rank-based EWMA procedure for sequentially detecting changes of process location and variability

Abstract

This paper presents a study of a new procedure, which is based on integrating a powerful nonparametric test for the two-sample problem and EWMA control scheme to online sequential monitoring. The proposed procedure, based on individual observation per sample, can be used to monitor the location and the scale parameters of a univariate continuous distribution, simultaneously. An iterative computation procedure is developed for computing the monitoring statistics. A search algorithm for the control limit based on Monte-Carlo simulation and bisection method is derived and a table is provided. The sensitivity analysis on the procedure is studied in detail. Monte-Carlo simulation results show that the proposed procedure is quite robust to nonnormally distributed data, and moreover, it is efficient in detecting various process shifts. A real data example from a chemical reaction process is shown to illustrate the application of our proposed procedure.

Keywords:

• Control charts
• EWMA
• nonparametric
• search algorithm
• statistical process monitoring

Acknowledgements

The authors would like to thank the Editor, the Associate Editor and two anonymous referees for their insightful comments, that improve this paper greatly.

Notes

No potential conflict of interest was reported by the authors.

Additional information

Funding

This paper is supported by the National Natural Science Foundation of China [grant number 11571191], [grant number 11201246], [grant number 11401573], [grant number 11371202], [grant number 11431006], [grant number 11131002] and a grant from City University of Hong Kong (Project No. 9380058). Part of this paper is finished during Li’s visit to Department of Systems Engineering and Engineering Management, at City University of Hong Kong, whose hospitality is appreciated.