A Bayesian Approach for Degradation Analysis with Individual Differences

Abstract

Reliability assessments of long-life and high-reliability products often face the difficulty of lack of failure time data. Meanwhile, there existing individual differences in produces due to the uncertainties during their whole lifetime, such as manufacturing, assembly, work environment and work load. This paper develops a reliability modeling and assessment method considering individual differences based on performance degradation data. Firstly, the model selection method is employed into the analysis of degradation process. And according to the model selection results, the appropriate degradation model for the degradation process can be chosen. Then, the Bayesian method is applied to establish the parameter estimation method and reliability assessment framework based on the appropriate model. At last, a real engineer example is presented to illustrate the effectiveness of the proposed degradation analysis method with individual differences.