Health prognostics framework of lithium-ion batteries for real-world electric vehicles based on charging data

Lithium-ion batteries’ state of health (SOH) is critical to safely managing electric vehicles (EVs). This paper proposes a battery health prognostic framework based on charging data for real-world EVs. First, the capacity of the battery is calculated using the ampere-hour integration method based on historical charging data. The capacity trend is fitted using the locally weighted linear regression (LWLR) algorithm and the effect of charging depth on the capacity calculation results is eliminated. Then, the features strongly correlated with battery capacity are extracted using Pearson correlation analysis (PCA). Finally, Gaussian process regression (GPR) is employed to establish the capacity estimation model of lithium-ion batteries. The effectiveness and accuracy of the proposed method are validated on EVs from different regions. The results indicate that the mean absolute error (MAE) and the root mean square error (RMSE) of the estimated capacity on the test set are both less than 0.5.