CS-EWMA Chart for Monitoring Process Dispersion

摘要

Control charts are the most extensively used technique to detect the presence of special cause variations in processes. They can be classified into memory and memoryless control charts. Cumulative sum and exponentially weighted moving average control charts are memory-type control charts as their control structures are developed in such a way that the past information is not ignored as it is done in the case of memoryless control charts, like the Shewhart-type control charts. The present study is based on the proposal of a new memory-type control chart for process dispersion. This chart is named as CS-EWMA chart as its plotting statistic is based on a cumulative sum of the exponentially weighted moving averages. Comparisons with other memory charts used to monitor the process dispersion are done by means of the average run length. An illustration of the proposed technique is done by applying the CS-EWMA chart on a simulated dataset. Copyright © 2012 John Wiley & Sons, Ltd.