Li-ion battery SOC estimation using EKF based on a model proposed by extreme learning machine

• Main Authors: Du, Jiani, Liu, Zhitao, Chen, Can, Wang, Youyi
• Other Authors: School of Electrical and Electronic Engineering
• Format: Conference or Workshop Item
• Language: English
• Published: 2013
• Online Access: https://hdl.handle.net/10356/98879; http://hdl.handle.net/10220/12835
• Institution: Nanyang Technological University

In this paper, a method for modeling and estimation of Li-ion battery state of charge (SOC) using extreme learning machine (ELM) and extended Kalman filter (EKF) is proposed. The Li-ion battery model from ELM, which is established by training the data from the battery block in MATLAB/Simulation, could describe the dynamics of Li-ion battery very well. And it has higher accuracy and needs less calculation than using the traditional neural networks. Moreover, the battery model and discrete SOC definition equation constitute state-space equations, and EKF is used to estimate the SOC of Li-ion battery. Comparing the actual SOC with the estimated SOC by simulation, it reveals that the method proposed in this paper has good performance on Li-ion battery SOC estimation.