State of charge estimation for Li-ion battery based on LSSVR and CKF

The Lion-ion battery state of charge (SOC) Estimation is an ever significant issue for the intelligent battery management system. Our study aims to estimate the state of charge Lion-ion battery from experimental data sets using the Dynamic least square support vector regression (DLSSVR) approach. In addition, Cubature Kalman Filter (CKF) was used to help find the optimized parameter for DLSSVR. The operation of the algorithm has been validated upon Federal Urban Driving Schedule (FUDS) of 10 C
^°
, 20 C
^°
, 30 C
^°
. The simulation conducted upon LSSVR and (DLSSVR + CKF), the result of which shows that the method of this paper(DLSSVR + CKF) can estimate SOC of li-ion battery with good accuracy and robustness under a dynamic environment.