A Novel BCRLS-BP-EKF Method for the State of Charge Estimation of Lithium-ion Batteries

Abstract

State-of-charge of lithium-ion batteries is one significant state parameter for battery management system monitoring. To accurately estimate the state-of-charge in real time, a novel BCRLS-BP-EKF method is proposed innovatively. Based on FFRLS algorithm, bias compensation is added to better capture the real-time operating characteristics of the system. To modify the model error of EKF algorithm, BP neural network is introduced, which has powerful nonlinear mapping and self-learning ability. The estimation error of EKF can be corrected by its learning and training relevant parameters that affect the estimation value of filtering. The data of different complex working conditions are used to verify the feasibility and rationality of the proposed algorithm by building a second-order RC equivalent circuit model. The results show that the root mean square error of the novel BCRLS-BP-EKF method under HPPC and BBDST operating condition can be controlled within 0.11% and 1.41% in state-of-charge estimation, which verifies that the proposed algorithm in this research has high precision and convergence characteristics. The novel BCRLS-BP-EKF method lays a theoretical foundation for accurate state estimation of lithium-ion batteries, which will effectively improve the security and reliability of electric vehicles.