EWMA charts for censored data using bayes and mle estimates for rayleigh distribution

This article deals with the monitoring of censored data using Exponentially Weighted Moving Average (EWMA) control chart for Rayleigh lifetimes under type I censoring. The focus is stability of the process mean level for which we have considered the specified parameter(s) as well as the unspecified parameter(s) cases [where Bayes and Maximum Likelihood Estimates (MLE) have been considered]. For this purpose, Conditional Expected Values (CEVs) are used and control structure is worked out for type I censored environment. Average Run Length (ARL) study is carried out to evaluate the performance of the said chart for the known parameter, Bayes and MLE estimation cases. An example is also included for illustration purposes. It is observed that for the case of specified parameter(s) CEV EWMA has an ideal ARL performance for all types of censoring rates in Rayleigh distributed process. For the case of unspecified parameter(s), Bayes based CEV EWMA has superior ARL performance as compared to that of MLE based for the low censoring rates, while for higher censoring rates the scenario is reversed.