A cubature Kalman filter for online state-of-charge estimation of lithium-ion battery using a gas-liquid dynamic model

Abstract

A novel state of charge (SOC) estimation method is proposed based on a gas-liquid dynamic model using a Cubature Kalman filter (CKF) with state constraints to improve the accuracy of estimations. The constraints derived from the principle of aerodynamic are introduced, which does not need the temperature correction coefficient and can achieve rapid SOC convergence due to its unique iterative form. Non-linear Kalman filter used to eliminate the violent jitter of the original algorithm when switching working conditions are also incorporated. Then, the proposed method is tested using a cylindrical-format ternary lithium-ion battery with a nominal capacity of 2.6 Ah and the estimated SOC were compared with the experimental results. As a result, comparative studies of two non-linear filters revealed that the CKF has the most outstanding performance based on concerning estimation accuracy, recovery time from an initial offset, and computational time. That is 36.4 % more accurate and the convergence time is 97.5 % shorter than the Extended Kalman filter (EKF) when the initial offset is 60 %, respectively.