A novel method for state of energy estimation of lithium-ion batteries using particle filter and extended Kalman filter

State-of-energy (SOE) estimation of lithium-ion batteries (LIBs) is one of the core functions of battery management systems in electric vehicles. In this study, to improve the accuracy and robustness of SOE estimation, a novel SOE method using a particle filter (PF) and extended Kalman filter (EKF) insensitive to uncertain total available energy loss and ambient temperatures is proposed. First, a battery model is established, and then the model parameters at different temperatures in the whole SOE range are identified. Second, the PF algorithm is chosen to estimate the SOE, whereas the EKF algorithm is used to update the total available energy online. Finally, the effectiveness of the proposed SOE method is verified by experiments under dynamic conditions. The experimental results indicate that the maximum error of the SOE estimation with the proposed PF-EKF algorithm is less than 3% under the dynamic stress test at 0, 25, and 40 degrees C. Moreover, even if there are large initial SOE and total available energy errors, the SOE estimation by the proposed algorithm would be able to quickly converge to it is reference trajectory with high accuracy and robustness.

Automated summary

A novel SOE estimation method using PF and EKF algorithms is proposed in this study, which is able to improve accuracy and robustness by identifying battery model parameters at different temperatures. Experimental results show that the maximum error of the proposed algorithm is less than 3% under dynamic conditions and can quickly converge to its reference trajectory even with large initial errors in SOE and total available energy.