Dynamic generalized fast initial response for EWMA control schemes for monitoring processes with startup and post‐startup problems

Abstract

AbstractThe exponentially weighted moving average (EWMA) control schemes have been proven to be very effective at monitoring random shifts or disturbances in a given process. However, EWMA is somewhat insensitive to the shifts at the process startup. Consequently, fast initial response feature (FIR) or headstart has often been used to increase the sensitivity of EWMA at the process startup. Although FIR feature significantly increases the sensitivity of the EWMA at the startup, its effects diminished after few observations thereby making FIR‐based schemes less sensitive compared to the classical EWMA at the process post‐startup. In this paper, we proposed the dynamic generalized fast initial response for the EWMA control schemes for monitoring processes with startup and post‐startup problems. The proposed scheme is highly sensitive at the startup and has a sensitivity equal to that of the classical EWMA at the process post‐startup. The average run length based performance comparisons of the proposed chart and its counterparts are presented. Real‐life examples are offered to demonstrate the applications of the proposed scheme.