LINEAR RELATIONSHIPS TO ASSIST IN PARAMETER SELECTION OF A CUSUM CONTROL CHART

Abstract

The cumulative sum (CUSUM) chart is widely employed in quality control. These charts are designed to exhibit acceptable average run lengths (ARL) both when the process is in and out of control. For such a chart, for specified ARLs, this paper introduces a technique employing a simple linear approximation for parameter selection. Graphs are produced to provide the coefficients for the fit. The simplicity of this approach should encourage its adoption as an additional tool in the statistical process control kit.

Keywords

• ARL
• average run length
• cumulative sum chart
• CUSUM
• SPC
• statistical process control