ABSTRACT

The accelerated degradation test (ADT) is an effective method for evaluating the lifetime of high-reliability products. In this paper, a doubly accelerated degradation model based on the inverse Gaussian process is proposed to characterize the ADT data, and then an objective Bayesian approach is presented to analyze the model. Some important noninformative priors including the Jeffreys prior and reference priors under different group orderings are derived. The propriety of the posterior distributions under each prior is validated. A simulation study is carried out to show the superiority of objective Bayesian approach compared with the parametric Bootstrap method. Finally, the approach is applied to analyze a carbon film data.

摘要

加速降解测试（ADT）是评估高可靠性产品寿命的有效方法。本文提出了一种基于逆高斯过程的双加速退化模型来表征ADT数据，然后提出了一种客观的贝叶斯方法来分析该模型。得出了一些重要的非信息性先验，包括不同组排序下的Jeffreys先验和参考先验。验证了每个先验条件下的后验分布的适当性。仿真研究表明，客观贝叶斯方法与参数Bootstrap方法相比具有优越性。最后，该方法被应用于分析碳膜数据。