A Joint Model Identification and Battery SOC Estimation Method Based on ETASEKF

Accurate and robust state-of-charge (SOC) estimation is critical for ensuring the reliability and efficiency of lithium-ion battery systems. To address model uncertainties, this article proposes an enhanced Thevenin equivalent circuit model (ECM), where model uncertainties are explicitly prescribed as a variable terminal voltage across a variable resistor. By modeling this time-varying voltage using a random walk model and integrating it into the system states, an enhanced Thevenin augmented state extended Kalman Filter (ETASEKF) is developed to enable joint online model identification and SOC estimation. The proposed method effectively compensates model deviations in real-time without requiring additional offline parameter identification, significantly enhancing estimation robustness. Experimental results validate that ETASEKF consistently outperforms conventional fractional-order extended Kalman filter (FOEKF) methods in terms of accuracy, robustness, and filter consistency under various dynamic and uncertain operating conditions, demonstrating its practical applicability in real-world battery management systems (BMSs).