Residual useful life prediction of the vehicle isolator based on Bayesian inference

Abstract

The shock and vibration of the isolation system of the vehicle are random, so the life model of the isolator cannot make an accurate prediction through the empirical formula. Also, the empirical model is not good at updating the relationship between the existing data and the observed information, so the prediction results do not change according to the data update and cannot reflect the current data information well. Therefore, a life prediction method for the vehicle isolator is proposed in this paper based on the principal component analysis and Bayesian inference. The uncertainty of model parameters is introduced in this method, and the residual useful life of the isolator is predicted considering historical data. The proposed method is compared with the existing methods, and the results show that the proposed method has good prediction accuracy.