State of charge estimation based on a new dual-polarization-resistance model for electric vehicles

Li-ion batteries have been widely used as the power source of electric vehicles. However, the acquisition of precise state of charge via battery management system remains a problem. A root cause is the complex characteristics of battery polarization, which is affected by the load current. In order to improve the accuracy and reliability of battery state of charge estimation, this paper focuses on the following three aspects: (1) A novel dual-polarization-resistance model is established based on the Thevenin model, in which the polarization resistance can be adaptively adjusted in accordance with the load current, making the battery model more robust. (2) An Extended Kalman Particle Filter is applied in state of charge estimation, and an improved Euler method is proposed for temporal propagation of the state vector, which effectively increases the calculation accuracy. (3) The proposed state of charge estimation algorithm is demonstrated through a set of experiments. By using the dual-polarization-resistance model, the maximum state of charge estimation error based on Extended Kalman Filter is reduced to 2.3%, while using conventional Thevenin model, the maximum error can be as high as 6.2%. Furthermore, by employing Extended Kalman Particle Filter on the dual-polarization-resistance model, the maximum error can further reduce to 1.8%. •A novel dual-polarization-resistance battery model is established.•The polarization resistance can be adjusted in accordance with operating current.•An Extended Kalman Particle Filter is applied for state of charge estimation.•An improved Euler method is proposed for temporal propagation of the state vector.•The proposed model and methods are verified under different load current inputs.