Adaptive Battery State-of-Charge Estimation using Aging-driven Equivalent Circuit Parameterization and Electrochemical Impedance Spectroscopy

Ensuring the accurate estimation of lithium-ion battery (LIB) state of charge (SOC) over the entire life cycle of the battery is still an open research challenge due to the nonlinear degradation and calendar aging especially for electric vehicle (EV) applications. State-of-the-art SOC estimation techniques such as Coulomb counting and equivalent circuit model (ECM) in fusion with Kalman filtering are ineffective in ensuring accuracy over the entire cycle life of the battery as the model parameters of these do not accommodate the changes of battery characteristics due to aging. This paper proposes an approach for estimating SOC using an adaptive extended Kalman filter (EKF) based on the parameters obtained from electrochemical impedance spectroscopy (EIS). By updating the ECM parameters, the proposed SOC estimation technique can maintain accuracy throughout the battery life, as the ECM parameters are continuously updated based on feedback from the EIS data. The effectiveness of the proposed method is demonstrated through a series of battery cycling (350 cycles) and EIS test data of a Samsung NCA 21700 LIB cell under a controlled laboratory environment.