A Real-Time Embedded Framework for EKF-Based State-of-Charge Estimation of Lithium-Ion Batteries

Abstract

Accuracy of estimation of battery State of Charge (SOC) is one of the major criteria for battery management systems for electric vehicles and power storage applications. Traditional techniques such as Coulomb counting have increasing errors with time, while model-based approach ensures accurate estimation. Estimation of SOC is necessary for the safe operation and extended life of the battery. This paper presents an integrated approach for estimating SOC using Recursive Least Squares (RLS) and the Extended Kalman Filter (EKF) using both simulated and hardware-based implementation methods. Initially, actual values of SOC are gathered using open source datasets (for instance, NASA dataset), and the algorithm based on EKF is developed and implemented in the MATLAB environment. The algorithm is further implemented using the Arduino MEGA to estimate the real-time SOC. The standard values of SOC are computed and then compared with the estimated values using the EKF algorithm. The results indicate that SOC estimation using EKF algorithm yields accurate results under the given operating conditions