A Bayesian life test sampling plan for a weibull lifetime distribution under accelerated type-I censoring

Abstract

A life test sampling plan under accelerated condition is an efficient approach in reliability demonstration, especially for the products with high reliability and long life. It pays more attention to rapid decision-making in determining the acceptance of a batch of products such that the producer and consumer risks are satisfied. Design of a sampling plan depends on the acceptance probability function parameters. The classical design method assumes the function parameters with precise values. In practice, there is uncertainty in those values. Moreover, available prior knowledge, such as history information and expert opinions, can be used in an accelerated life test sampling plan (ALTSP) design to reduce testing resources. In this paper, a Bayesian ALTSP is developed for a Weibull lifetime distribution under type-I censoring to overcome these problems. The Bayesian posterior risk criteria are introduced to construct the acceptance probability function. The uncertainty in the values of the parameters of the function including AF (acceleration factor) is expressed by prior distributions. The MCMC (Markov chain Monte Carlo) method is adopted to obtain plans. The proposed method is demonstrated by an example.