SOC EKF Estimation based on a Second-order LiFePO4 Battery Model

Abstract

An accurate battery State of Charge (SOC) estimation has great significance in improving battery life and vehicle performance. An improved second-order battery model is proposed in this paper through quantities of LiFePO4 battery experiments. The parameters of the model were acquired by the HPPC composite pulse condition under different temperature, charging and discharging rates, SOC. Based on the model, battery SOC is estimated by Extended Kalman Filter (EKF). Comparison of three different pulse conditions shows that the average error of SOC estimation of this algorithm is about 4.2%. The improved model is able to reflect the dynamic performance of batteries suitably, and the SOC estimation algorithm is provided with higher accuracy and better dynamic adaptability.

DOI: http://dx.doi.org/10.11591/telkomnika.v11i8.2629