Identifying the time of step change in process parameter for Maxwell distribution

Abstract

Due to the quick identification of the root causes for an out-of-control process, the estimation of exact time of a process change would be helpful thing for the process improvement. In contrast to the typical normal assumption, this study realizes that a process may follow a skewed Maxwell distribution. In this paper, using maximum likelihood estimation, a considerably efficient change point model is proposed for V control chart which underlying distribution is Maxwell distribution. The required chart statistics are calculated with its distributional properties. The proposed method is used when V control chart signals a change in a process parameter. The proposed method, when used with V control chart would be helpful for process engineers both in controlling and identifying a permanent step change in a process scale parameter. An illustrative example is given to understand the use of the proposed method.