Remaining Useful Life Prediction Using a Novel Two-Stage Wiener Process With Stage Correlation

Remaining useful life (RUL) prediction of products is a critical component of prognostics and health management. Recently, the RUL prediction based on a two-stage degradation process has received increasing attention. However, existing works generally assume that the two stages are mutually independent, which is not reasonable in many applications. To address this problem, we propose a novel two-stage Wiener process model with stage correlation and a Bayesian approach for RUL prediction. Different from previous studies, we incorporate the stage correlation into the prior distributions of model parameters to improve the accuracy of predictions. We also derive the RUL distribution through the total probability formula to comprehensively consider the possibilities that the product fails at different stages. Once real-time monitoring data are available, we employ the Gibbs sampling algorithm to update the posterior distributions of model parameters as well as the RUL distribution. The superiority of the proposed method is demonstrated through a simulation study and an application to the bearing degradation data.

Abstract licence: CC BY-NC-ND

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.