A framework of Bayesian lifetime demonstration based on degradation

In order to validate the lifetime of long-life product, the concept of lifetime demonstration testing (LDT) based on degradation is presented in this paper. For shortening test time as far as possible, the prior information of similar products is considered. Therefore, a framework of Bayesian lifetime demonstration based on degradation is developed by considering the prior information of products. Firstly, we present a Bayesian reliability model based on a generalized degradation process, and take a linear form for example to infer the posterior cumulative distribution function (CDF). Due to the prior information considered, the Bayesian sequential probability ratio test (SPRT) method is used to demonstrate lifetime of the product at test censored time. And the corresponding decision rules are given under the producer´s risk α and consumer´s risk β. Finally, for an illustration of the proposed model and method, an example about the self-healing metallized film pulse capacitors is discussed and some interesting and exciting analytical results are presented.