On enhanced sensitivity of nonparametric EWMA control charts for process monitoring

Authors

1 Department of Mathematics and Statistics, King Fahd University of Petroleum and Minerals, Dhahran, 31261, Saudi Arabia.

2 Department of Mathematics, COMSATS Institute of Information Technology, Wah Cantt 47040, Pakistan.

Abstract

Control charts are popular tounderols for process monitoring and are mainly classi ed into memory and memory-less structures. Exponentially Weighted Moving Average control charts (EWMA) are widely used for the detection of small shifts in process parameters. The traditional design structures of these types of charts rely on the stringent assumption of normality. In practice, quality characteristics generally do not follow normality; hence, alternative charting structures are needed. Nonparametric control chart is one such alternative that do not depend on the mentioned distributional assumption(s). This study deals with nonparametric EWMA charts for an ecient detection of smaller shifts in location. We have investigated asymptotic, time-varying, and Fast Initial Response (FIR) based time varying control limits for nonparametric EWMA charts. We have considered two variants of FIR with time varying limits for nonparametric EWMA structure. A variety of run length properties have been investigated for performance evaluations. It is observed that FIR-based nonparametric EWMA charts with time-varying limits are quite sensitive to detecting smaller shifts, and they o er attractive run length properties. A real-life example is also provided to illustrate the application of the time-varying and FIR-based nonparametric EWMA control charts.

Keywords

Scientia Iranica
Volume 24, Issue 1 - Serial Number 1
Transactions on Industrial Engineering (E)
February 2017
Pages 424-438

Files

Share

How to cite

Statistics