A novel SOC estimation method for Li-ion batteries based on improved Kalman filter with parameter online identification

State-of-Charge(SOC) is a key factor for evaluating the accumulation statues of energy battery in Electric-Vehicles (EVs). For Kalman Filter (KF) can usually be used to estimate the SOC of EVs while the effectiveness and robustness remains to be raised currently, the paper presents an improved KF integrated with the Coulomb Counting and Open-Circuit Voltage method. Considering that the performance of battery is liable to be influenced by the change of some status parameters, such as temperature, cycle times, self-discharge rate and so on, a fading memory recursive lease squares algorithm is presented to online identify the concerned parameters of battery model which is indispensable on the process of SOC estimation using the KF. By taking advantage of the algorithm, the improved KF is able to be time-varied with battery status. The experimental results verify that the proposed method can effectively increase the precision of SOC estimation, which is significant to optimize the running of energy battery and raise the efficiency of energy transfer.