Estimation and Prediction on State of Health of Electric Vehicle Battery Based on BP Neural Network

Faced with complex battery degradation mechanism, machine learning method is used to establish battery model in this paper. We propose a method to extract features from the factors affecting battery life. In order to obtain the nonlinear relationship between features and State of Health (SOH), the battery SOH estimation model is established by BP neural network. Aiming at the prediction of SOH, the SOH time series characteristics are constructed by using the sliding window as the input of BP neural network to predict the future attenuation of SOH. The experimental results show that the BP neural network model trained by real vehicle data can effectively realize the estimation and prediction of SOH, which provides accurate data support for battery safety management.