Towards a Novel Indirect Battery Health Indicator Based on Electrochemical Impedance Spectroscopy for RUL Estimation in Electric Vehicles

With the significant growth of the electric vehicles' market, the role of battery prognostic has currently assumed a central point. In this context, selecting an effective health indicator is essential for accurately assessing the ageing condition and reliability of battery cells. Although the direct measurement of the capacity is a widely spread health indicator, it can be unfeasible in automotive applications due to the necessity of continuous discharge current measurements. Thus, it is necessary to use other indirect measurements capable of capturing the degradations occurred in the cell. Considering the necessity of fast and accurate methods to estimate the state-ofhealth (SOH) of the battery, this work proposes a novel and simple approach to determine an indirect health indicator by using the information collected by Electrochemical Impedance Spectroscopy (EIS) right after the charge. More specifically, two health indicators are discussed: one obtained solely from EIS features and one that integrates additional features from charge and discharge processes. The monitoring performances are validated on a largely variable dataset, to verify the robustness of the predictors on cells cycled under different operating conditions. Eventually, the paper provides a comparison with the classical SOH definition, showing promising results for the realization of on-board SOH monitoring systems.