SOC Estimation for Lithium Battery Based on Segmented Model UKF Filter

摘要

The rapidness and accuracy of the state of charge (SOC) estimation is the key technology in the battery management system. To solve the problems in SOC estimation caused by the inherent dynamic and nonlinear property of a lithium battery, first, an improved accurate measurement model is proposed by using segmented model method, in order to mitigate the negative influence of Polarization Effect. And then, the detail procedures and algorithms for lithium battery SOC estimation based on segmented model Unscented Kalman Filter (UKF) are given. Finally, the accuracy and the convergence rate of the segmented model UKF filter are analyzed. Experiments show that the segmented model UKF filter can be used to estimate the SOC quickly and accurately with an estimation error of 3%, which saves half time than the global model. Meanwhile, the adjustment of the initial value of UKF filter does not affect the estimation accuracy.