Modified EWMA Control Chart for Skewed Distributions and Contaminated Processes
In this paper, we proposed the variable control chart for monitoring process mean by modified the traditional Exponentially Weighted Moving Average (EWMA) control chart for skewed and contaminated processes. The proposed EWMA control charts are modified both of mean and standard deviation in the control limits of the traditional EWMA control chart based on three different heuristic methods, namely, the Modified Exponentially Weighted Moving Average (MEWMA) control chart, the Modified Weighted Standard Deviation Exponentially Weighted Moving Average (MWSD-EWMA) control chart, and the Modified Weighted Variance Exponentially Weighted Moving Average (MWV-EWMA) control chart. A comparison of the performance measures is the average run length (ARL) were compared with the traditional EWMA control chart in various situations via simulation. The results show that for Gamma distribution, the MWSD-EWMA control chart is performs well for small sample size (), low levels of , and are low in all levels of . The MEWMA control chart is the better than other charts for moderate levels of on all levels of and relatively small differences with the MWV-EWMA control chart. When the process distribution is Weibull, the MWSD-EWMA control chart is performs well for small sample size and are low in all values of and . If the sample sizes are moderate and large , the MWSD-EWMA control chart is better than other charts for very low level of in all values of and . For lognormal distribution, we found that the results as well as the Weibull distribution in all levels of and . Finally, all charts are similar performances for all values of and when high levels of in all skewed distribution processes.
Champ CW, Woodall WH. Exact results for Shewhart control charts with supplementary runs rules. Technometrics. 1987; 29: 393-399.
Chan LK, Cui HJ. Skewness correction and R charts for skewed distributions. Nav Res Logist. 2003; 50: 1-19.
Chang YS, Bai DS. Control charts for positively popultions with weighted standard deviation. Qual Reliab Eng Int. 2001; 17: 397-406.
Choobinech F, Ballard JL. Control-limits of QC charts for skewed distributions using weighted-variance. IEEE Trans Reliab. 1987; 36: 473-477.
Crowder SV. A simple method for studying run-length distributions of exponentially weighted moving average charts. Technometrics. 1987; 29: 401-417.
Crowder SV. Design of exponentially weighted moving average schemes. J Qual Tech. 1989; 21: 155-162.
Derya K, Canan H. Control charts for skewed distributions: Weibull, gamma, and lognormal. Metodoloski. 2012; 9: 95-106.
Derya K. Robust control chart for monitoring the skewed and contaminated process. Hacettepe J Math Stat. 2016; 46: 1-19.
Khoo MBC, Atta AMA. An EWMA control chart for monitoring the mean of skewed populations using weighted variance. Proceeding of the 2008 IEEE International Conference on Industrial Engineering and Engineering Management; 2008 Dec 8-11; Singapore. IEEE; 2008. p. 218-223.
Khoo MBC, Atta AMA, Chen CH. Proposed and S control charts for skewed distributions. Proceeding of the 2009 IEEE International Conference on Industrial Engineering and Engineering Management; 2009 Dec 8-11; Singapore. IEEE; 2009. p. 389-393.
Lucas JM, Saccucci MS. Exponentially weighted moving average control schemes: properties and enhancements. Technometrics. 1990; 32(1): 1-12.
Montgomery DC. Introduction to statistical quality control. 5th ed. New York: John Wiley & Sons; 2005.
Muhammad A, Osama-H A, Chi-Hyuck J. A control chart for gamma distribution using multiple dependent state sampling. Ind Eng Manag Syst. 2017; 16: 109-117.
Page ES. Control charts with warning lines. Biometrics. 1955; 42: 243-257.
Roberts SW. Properties of control chart zone tests. Bell Syst Tech J. 1958; 37: 83-114.
Roberts SW. Control chart tests based on geometric moving averages. Technometrics. 1959; 42: 97-102.