State of charge estimation of power lithium-ion battery based on an adaptive time scale dual extend Kalman filtering

In this paper, we introduce the Unsymmetrical Thevenin model, an improved equivalent circuit model to obtain a more precise SOC estimation. We first propose an Auto-tuning Multiple Forgetting Factors Recursive Least Squares (AMFFRLS) for model parameter identification, then, we proposed an Adaptive Time Scale Dual Extend Kalman Filtering (ATSDEKF) to update the model parameters and Sliding Window Forgetting Factor Approximate Total Recursive Least Squares (SWFFATRLS) to update the maximum available capacity of a lithium-ion battery to obtain more accurate state of charge (SOC) estimation. Numerical experiments demonstrate that the proposed method can get better SOC estimation results compare to the traditional ones. Except for extreme temperatures, such as at 0 degrees C, the root mean square error (RMSE) of the Unsymmetrical Thevenin model is below 1.2%, which is much smaller than the most common Thevenin model with fixed parameters based on Extend Kalman Filtering (EKF).