State of health estimation for fast-charging lithium-ion battery based on incremental capacity analysis

Abstract

Recently, lithium-ion batteries with fast-charging capability up to 3C rate are progressively equipped in modern consumer products. State of health (SOH) estimation specifically developed for fast-charging batteries is still an open-question, and the majority of the published methods is for batteries under lower charging rate below 1C. This study proposes a novel SOH method in particular for fast-charging batteries based on the incremental capacity (IC) analysis and Gaussian process regression (GPR). Firstly, the battery aging under fast-charging conditions is discussed according to the IC analysis, and a different characteristic of the IC curve is found for fast-charging batteries beyond the one observed under a lower charging rate. Then a novel feature extracted from the IC curves is introduced for SOH estimation of fast-charging batteries, and this novel feature is further identified and verified both from physical mechanism analysis and from quantitative comparison with classical features on battery aging correlation. Finally, a GPR model is established and trained with the dataset of extracted features for SOH estimation of fast-charging batteries. The proposed method is demonstrated on two fast-charging batteries Datasets, in which one of them simulates the real-life application, and the proposed method can achieve more than 90% reduction on mean absolute percentage error.