Computational Modelling and Prediction of Remaining Useful Life of Rolling Bearings in Electric Vehicles (EV)

Abstract

Rolling bearings are essential parts of an electric vehicle's powertrain. They support spinning and stationary components and lower friction between them. However, the constant running and heavy loads in electric cars can cause rolling bearings to degrade, which can lead to unanticipated breakdowns and downtime. The remaining useful life (RUL) of a rolling bearing is the predicted period of time before proactive maintenance or replacement is needed. Predictions of the RUL of rolling bearings in electric vehicles are frequently made using computational modeling techniques that combine operational data. Based on the recent success of the deep neural networks in various artificial intelligence and machine learning domains, the proposed end-to-end deep frameworks for RUL estimation based on Support Vector Machine (SVM), Convolutional Neural Network (CNN), Convolutional and Long-Short-Term Memory (CNN-LSTM) is founded on the examination of actual data gathered from rolling bearings in electric cars. To precisely anticipate the rolling bearings' remaining useful life. The suggested method is verified using experimental data and shows high accuracy in estimating the remaining usable life of rolling bearings in electric cars. The findings of this study may be utilized to create preventative maintenance plans that save downtime and maintenance expenses, boosting the dependability and effectiveness of electric cars.