Battery life estimation based on cloud data for electric vehicles

Abstract To evaluate the safety and life-cycle of electric vehicles (EVs), automobile companies usually retain the driving data of EVs on the cloud for monitoring and management. The recording period of the cloud data is generally as long as 10–30 s at present, so the dynamic driving condition of EVs is hard to be revealed with the cloud data. But the charging data are stable, which makes it possible to estimate the battery life based on the charging cloud data. Battery life estimation includes capacity estimation and internal resistance estimation. In this paper, the capacity is directly estimated by the ampere hour integral method. The estimation results are modified based on the temperature data and optimized by the Kalman filter (KF). We further propose the Fuzzy logic (FL) to control the observation noise which effectively improves the accuracy of the estimation results. Then, the battery life is predicted by the Arrhenius empirical model. The sudden changes of voltage and current in the charging data are used for estimating the internal resistance. The internal resistance prediction is achieved using a similar process to the capacity prediction. The sampling test shows that the errors of the battery life estimation method are less than 4%.