The generally weighted moving average control chart for monitoring the process median

摘要

Though there exist many traditional and specially developed control charts for monitoring and statistically controlling a process, in choosing a suitable control chart certain desirable properties are expected for their use. Such properties include simplicity, fast detection of assignable causes, robustness to underlying assumptions, and economical usefulness. Of course, in monitoring the process mean, the traditional average control charts, such as the Shewhart charts, cumulative control charts and exponentially weighted moving average (EWMA) charts are widely used. However, in order to achieve fast detection of process shifts, the Shewhart median and the EWMA-based median control chart have been developed and are in the literature. In this article, a generally weighted moving average median control chart is proposed that outperforms all other charts of its kind and is also very economical. As an example, simulated data are used for analysis and illustration.