CMStatistics 2022: Start Registration
View Submission - CMStatistics
B1726
Title: Statistical inference for the virtual age imperfect repair model Authors:  Mosa Alsabhi - Durham University (United Kingdom) [presenting]
Abstract: The repairable systems are usually analysed using Nonhomogeneous Poisson Process (NHPP), which represents minimal repairs or as bad as old and the Renewal process (RP) which represents perfect repairs or as good as new. However, some repairs are imperfect, which will be between minimal repairs and perfect repairs. Virtual age models are imperfect repair models used for repairable systems after each repair or maintenance. After each repair, a system can either be rated as bad as old, as good as new, or somewhere between the two. Two methods for estimating virtual age models will be presented based on the Weibull distribution: maximum likelihood estimation (MLE) and the Bayesian estimator method. These two methods are compared using the Monte Carlo simulation by generating different random sample sizes. Then, an example will be provided that will illustrate MLE and Bayesian estimation inference.