State of Health and Remaining Useful Life Prediction of Li-ion Batteries Based on Improved Support Vector Regression

The state of health (SOH) and remaining useful life (RUL) are both important parameters that reflect the degree of aging of Li-ion batteries. Support vector regression (SVR) has advantages in processing small sample data and time series analysis but is challenging to select kernel parameters. In this paper, we present an improved SVR model based on the Cauchy Variation Ant-lion Optimization (CALO) algorithm to enhance the prediction performance of SOH/RUL. Firstly, we conduct feature selection and select health factors based on correlation analysis. We then use the Ant-lion (ALO) algorithm to optimize kernel parameters and incorporate the self-adaptive T-distribution Cauchy variation into the ALO algorithm to improve its global convergence and local search ability. Finally, the National Aeronautics and Space Administration (NASA) and Center for Advanced Life Cycle Engineering (CALCE) datasets are used to compare and validate the proposed method. The results demonstrate that the proposed method has a faster convergence speed and higher optimization accuracy, which can effectively enhance the accuracy and robustness of the SOH and RUL prediction of Li-ion batteries.