Enhanced generally weighted moving average variance charts for monitoring process variance with individual observations

Abstract

Abstract The generally weighted moving average variance (GWMAV) chart is effective in detecting increases in process variance when only individual observations are available. Recently, the combination of exponentially weighted moving average and cumulative sum (CUSUM) charts for the effective detection of small process shifts has emerged. Inspired by the features, we propose the mixed GWMAV‐CUSUM chart and its reverse order CUSUM‐GWMAV to enhance the detection ability of the GWMAV chart and compare with the existing counterparts. Numerical simulation revealed that the mixed GWMAV‐CUSUM and mixed CUSUM‐GWMAV charts are sensitive to small upward shifts in the process variance and efficient structures compared with their prototypes and their separate charts, that is, GWMAV and CUSUM charts. An industrial dataset was used to illustrate the application of the proposed mixed charts.