Online state of charge EKF estimation for LiFePO4 battery management systems

State of charge (SOC) is the most important status parameter of energy storage system, which is able to predict the available mileage of electric vehicle. In fact, the accuracy of SOC estimation plays a vital role in the usability and security of the battery. In this paper, we designed an battery characteristic experimental system in which LiFePO4 battery was tested under different conditions. Based on quantities of LiFePO4 battery experiments, an improved second-order battery model was proposed in this paper. To fully consider the practical demands, parameters were acquired by the HPPC composite pulse condition under different factors, such as temperature, charging and discharging rates and SOC. At last, the SOC of the battery was estimated through a state equation established by the Extended Kalman Filter (EKF). Experiments proved that the maximum error of SOC estimation is less than 4.2%. Compared with the original Ah method, the improved method has a better ability to reflect the dynamic performance of batteries suitably, and a better dynamic adaptability. What's more, the SOC estimation algorithm is realized by DSP 5509A in an Battary Manage System (BMS).