An adaptive prediction approach for rolling bearing remaining useful life based on multistage model with three-source variability

Abstract

The prediction of remaining useful life (RUL) of rolling bearing is core content of equipment prognosis and health management. For rolling bearings, the degradation trend can be divided into multiple stages, and each stage has uncertain changes. Therefore, a new approach of bearing RUL prediction based on stochastic process model is proposed in this paper. Firstly, a new stochastic degradation model is established, which integrates the characteristics of multistage and multi-variability of degradation trend. Then, the statistical process control (SPC) is applied to stage division for the first time, which divides degradation stages and adaptively switches degradation models. At the same time, in the absence of prior information, update model parameters online by using parameters estimation method based on expectation maximization (EM) algorithm and predict RUL distribution in different degradation stages. Finally, the effectiveness of this approach is verified by empirical study of simulation example and XJTU-SY bearing data. The results show that this approach can divide different stages of rolling bearing and provide RUL prediction of corresponding stages.