A model to estimate the state of health of lithium-ion batteries for different charging rates

In recent years, there has been significant research interest in the fast charging of lithium-ion batteries (LIBs). However, the estimation of the State of Health (SOH) for LIBs under fast charging conditions has received relatively little attention. Therefore, a data-driven and improved incremental capacity analysis-based SOH estimation model for LIBs is proposed in this paper, which can estimate the SOH at different charging rates. Firstly, a revised Lorentz voltage capacity (RL-VC) model is constructed using the constant current charging data. Further, the revised Lorentz incremental capacity (RL-IC) curve of the battery is calculated and decomposed according to the RL-VC model. Then, the health features (HFs) are extracted from the original and the decomposed RL-IC curves. The HFs with high correlation are selected through Pearson correlation analysis as inputs to the back-propagation neural network to build the SOH estimation model. Validation of the model is performed on NASA datasets containing one charging rate and experimental datasets with three charging rates. The results indicate that the proposed model can estimate the SOH of LIBs more accurately than the estimation models in the recent literature. The SOH estimation errors are all less than 1 %, and the coefficients of determination are all higher than 0.98 in Hold-out cross validation. It also shows excellent generalization ability under different charging rates. • Developing a robust model for accurate estimation of SOH at various charging rates. • Combining enhanced capacity analysis with neural network for accurate SOH estimation. • Reduces computational load on BMS, mitigating measurement noise and filter effects.