Lithium-ion battery health state estimation based on feature reconstruction and optimized least squares support vector machine

Abstract

The state of health (SOH) of a battery is the main indicator of battery life. In order to improve the SOH estimation accuracy, a model framework for lithium-ion battery health state estimation with feature reconstruction and improved least squares support vector machine is proposed. Firstly, the indirect health features (HF) are obtained by processing multiple health features extracted from the charging and discharging phases through principal component analysis (PCA) to remove the information redundancy among multiple features; then multiple smooth component subsequences of different frequencies are obtained by using variational modal decomposition (VMD) to efficiently capture the overall downtrend and regeneration fluctuations of the data. Then we use the Sparrow Search Algorithm (SSA) to optimize the Least Squares Support Vector Machine (LSSVM) to build an estimation model, and then predict and superimpose the reconstructed fusion features of multiple feature subsequences, and then use the mapping relationship between the reconstructed HI and the SOH for the estimation . The NASA and University of Maryland (CACLE) battery dataset(CACLE) is used to perform validation tests on multiple batteries with different cycle intervals. The results show that the Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) are less than 1% and the method has high estimation accuracy and robustness.