Adaptive State-of-Charge Estimation Based on a Split Battery Model for Electric Vehicle Applications

Abstract

The conventional state-of-charge (SoC) estimation methods based on the equivalent circuit model (ECM) integrate all state variables into one augmented state vector. However, the correlations between RC voltages and SoC degrade the stability and accuracy of the estimates. To address this problem, this paper presents an adaptive SoC estimation method based on the split battery model, which divides the conventional augmented battery model into two submodels: the RC voltage submodel and the SoC submodel. Hence, the cross interference between RC voltages and SoC is reduced, which effectively reduces the oscillation in the estimation and improves the estimation accuracy. In addition, the adaptive algorithm is applied on the SoC submodel to improve the system robustness to noise disturbances. A case of a second-order ECM is analyzed and two types of Lithium-ion batteries are employed to verify the universality of the proposed method. Experimental results show that the undesired oscillation is eliminated during the convergence stage and the maximum SoC error is within 1% over a wide SoC range.