Remaining useful life prediction method of rolling bearing based on SKF-KF-Bayes

Accurate prediction for residual life of rolling bearing plays an important role in safe and reliable operation of mechanical equipment. Here, aiming at problems in life prediction of rolling bearing, such as, not being able to accurately distinguish degradation stage of rolling bearing and how to effectively use historical degradation data and real-time monitoring data, a method for rolling bearing performance degradation modeling and remaining useful life (RUL) prediction based on combination of switching Kalman filters(SKF) recognition, Kalman filters (KF) single step prediction and Bayes update was proposed. Firstly, combined with performance monitoring data of rolling bearing vibration signals, SKF method was used to identify the change point of bearing performance degradation. Secondly, the random effect exponential degradation model was used to describe the process of bearing performance degradation, and the maximum likelihood estimation of the model's unknown parameters was given based on the performance data of the same kind bearings. Then, KF single-step prediction was used to modify the monitoring data at the present moment, and random parameters in the model were updated in real time based on Bayes method to derive the bearing residual life distribution model, and calculate the residual life of rolling bearing. Finally, the applicability and effectiveness of the proposed method were verified through analyzing test data of rolling bearing. © 2021, Editorial Office of Journal of Vibration and Shock. All right reserved.