Extension of CV-GWMA control chart with fast initial response feature

Abstract

The control chart is widely used statistical process control (SPC) tool that searches for assignable cause of variation and detects any change of process. When the production run is short and process parameters change frequently, it is difficult to monitor the process using traditional control charts. In such a case, the coefficient of variation (CV) is very useful for monitoring the process variability. The CV control chart, however, is not sensitive at small shifts in the magnitude of CV. The CV-EWMA (exponentially weighted moving average), the CV-DEWMA (double exponentially weighted moving average) and the CV-GWMA (generally weighted moving average) control chart which were developed recently is effective in detecting a small shifts of CV. The CVGWMA control chart shows better performance than the EWMA or DEWMA control charts in detecting small shifts of the process. In this paper, we propose an FIR (Fast initial response) CV-GWMA control chart to improve the sensitivity of a CV-GWMA control chart at process start-up or out-of-control process. Moreover, we show the results of the performance study of FIR CV-GWMA control chart by comparing with CV-DEWMA and CV-GWMA control chart. © 2012 American Scientific Publishers. All rights reserved.