State of Charge Estimation of Lithium-ion Battery based on Minimum Error Entropy Square Root Cubature Kalman Filter

To account for the impulsive non-Gaussian noises, a minimum error entropy square root cubature Kalman filter (MEE-SRCKF) is proposed to estimate the state of charge (SOC) of lithium-ion batteries. It adopts the robust minimum error entropy (MEE) criterion as the optimization criterion, instead of the MMSE or maximum correntropy criterion (MCC). And then the square root cubature Kalman filter (SRCKF) is used to deal with the nonlinearity. The performance of the proposed algorithm is validated by a pulsed discharge test and an urban dynamometer driving schedule (UDDS) test. Experimental results demonstrate that the proposed approach has significantly improved the SOC estimation accuracy of the lithium-ion battery under non-Gaussian measurement noise compared with the SRCKF algorithm. In addition, given different initial SOC values, MEE-SRCKF algorithm also converges rapidly to its true value.