Hybrid exponentially weighted moving average control chart using Bayesian approach

Abstract

A control chart is an important tool to monitor an industrial process and the use of prior knowledge by Bayesian theory can be helpful in-control charting. In this paper, we have used the Bayesian approach with two different loss function (LF) symmetric and asymmetric loss function namely Linex LF and squared error LF under informative (conjugate) prior and non informative prior (uniform and Jeffery prior) to develop the hybrid exponentially weighted moving average control chart. The average run length and standard deviation of run length have been used as the performance measure for the proposed Bayesian HEWMA control chart using posterior and predictive posterior distribution. An extensive simulation study is conducted to evaluate the proposed control chart and a real data example is presented to demonstrate the implementations of the proposal.