<![CDATA[Accurate state of charge (SOC) estimation is critical for the reliable operation of battery management systems (BMS) in electric vehicles (EVs) and energy storage applications. This paper presents a method for online identification of Thevenin model (TM) parameters and SOC estimation using the extended Kalman filter (EKF). The objective is to improve estimation accuracy by precisely characterizing the SOC-dependent variations of model parameters, including open-circuit voltage (VOCV), internal resistance R1, polarization resistance R2, and capacitance C2. These parameters are identified using least squares regression based on experimental discharge data from a 1.83 Ah lithium-ion (Li-ion) battery. The resulting model is validated under pulsed discharge conditions, achieving a mean absolute error (MAE) of 0.0059 V and root mean square error (RMSE) of 0.0074 V, indicating high modeling accuracy. Subsequently, an EKF algorithm is implemented using the identified model to estimate SOC in real time. Experimental results show excellent performance with an SOC estimation MAE of 0.059% and RMSE of 0.0798%, demonstrating high precision, fast convergence, and stability. The method effectively combines empirical parameter identification with a recursive filtering technique, offering a practical and embeddable solution for BMS applications. The study concludes that accurate parameter modeling significantly enhances EKF-based SOC estimation, providing a robust foundation for real-time battery monitoring and control. ]]>
Copyrights © 2026
