Battery State of Charge (SOC) Estimation Based on Extended Kalman Filter for Electric Vehicles

The State of Charge (SOC) is a crucial part of Battery Management Systems. Since SOC cannot be measured directly, it is difficult to determine the charge state of a battery cell. Conventional SOC estimation methods often fail to maintain accuracy under dynamic operating conditions, and these limitations become more pronounced in lithium iron phosphate (LFP) cells due to their flat relationship between SOC and open circuit voltage (OCV), and temperature dependent behavior. To address these challenges, this study introduces an SOC estimation approach that employs a combination of forgetting-factor recursive least squares (FFRLS) and the Extended Kalman Filter (EKF). A dual polarization equivalent circuit representation of the LFP battery is established using measurements obtained through a hybrid pulse power characterization (HPPC) test. The FFRLS algorithm is employed to identify time-varying model parameters online, enabling real-time adaptation to changes in battery behavior. Subsequently, the EKF algorithm is utilized to estimate the SOC iteratively based on the identified model. Experimental results demonstrate that the proposed FFRLS-EKF method achieves high estimation accuracy and strong robustness under varying load conditions, providing an effective solution for advanced battery management systems.