A data-driven approach for estimating state-of-health of lithium-ion batteries considering internal resistance

Abstract

State-of-health (SOH) estimation of lithium-ion batteries is an important issue in electric vehicle energy management. The complication of the internal electrochemical reaction mechanism and the uncertainty of the external operating conditions pose a significant challenge to SOH estimation. This paper develops a data-driven approach to estimate the SOH of lithium-ion batteries with consideration of the battery's internal resistance, which is used as a bridge to effectively integrate the equivalent circuit model (ECM) and the data-driven method. We try to identify the internal resistance under constant current charging conditions by simplifying the ECM. The poles and offsets are extracted from the differential thermal voltammetry, differential thermal capacity, and incremental capacity curves as thermoelectric coupling features. Then the internal resistance and thermoelectric coupling features are combined as model inputs. An explanation boosting machine (EBM) is used to construct the SOH estimator according to the good fitting performance and interpretability. The model parameters of EBM are optimized by using an ant colony algorithm to improve its robustness. Finally, comparative experiments between features and the model are carried out on the Oxford dataset. The results demonstrate that the mean absolute error of the proposed method is less than 1%.