State-of-charge estimation using a self-adaptive noise extended Kalman filter for lithium batteries

• Published in: 2014 IEEE PES Asia-Pacific Power and Energy Engineering Conference (APPEEC) pp. 1 - 5
• Main Authors: Daiming Yang, Jianzheng Liu, Yi Wang, Man Chen, Baihua Zhang, Yongqi Li
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.12.2014
• Subjects: Batteries; Battery energy storage system (BESS); Estimation; extended Kalman filter (EKF); Integrated circuit modeling; Kalman filters; lithium battery; Noise; state-of-charge (SOC) estimation; System-on-chip; Voltage measurement
• ISSN: 2157-4839
• DOI: 10.1109/APPEEC.2014.7066097

Extended Kalman filter (EKF) is a typical method for state-of-charge (SOC) estimation of lithium batteries. An equivalent circuit model for the lithium battery is established, and model parameters are identified by experimental methods. Based on the battery model and model parameters, an EKF is proposed for SOC estimation. Analysis is conducted on the effect of the model coefficients of EKF to the stability and rapidity of SOC estimation. By calculating the sensitivity of SOC to EMF under different SOC stages and according to the needs of various operating conditions, the EKF process noise and measurement noise are adaptively adjusted to regulate the proportion of the measurement update to estimation result. The regulation contributes to tracking the SOC and improving the smoothness of the estimation result. It is validated by experiments on lithium iron phosphate (LiFeP04) battery that the performance of the EKF on SOC estimation in different situations is optimized by the noise adaptive algorithm.