State of charge estimation for LiFePO4 battery via dual extended kalman filter and charging voltage curve

Abstract The state-of-charge (SOC) estimation method currently ignores the measurement error caused by the Battery Management System (BMS). In this paper, the characteristic of LiFePO4 battery is deeply studied to explore the relationship between open-circuit-voltage (OCV) and SOC. By the analysis of the characteristic of the curve, the results show that the curve does not change with the battery aging by the capacity correction. Meanwhile, the feature of the charging voltage curve is also analyzed. It is pointed out that the ohmic internal resistance and capacity can be obtained by the transformation of the charging voltage curve, which reduces the workload of the dual extended kalman filter (DEKF) algorithm. Based on the DEKF algorithm, the SOC under constant current and dynamic discharge conditions are estimated. The results show that the estimation error is within 3%. The influence of battery voltage and current measurement noise on the estimation accuracy of the SOC is then analyzed. It is found that the measurement noise increases the SOC estimation deviation. Finally, the open circuit voltage in measurement equation is replaced by the charging voltage. And a new method of combining DEKF algorithm and charging voltage curve for SOC estimation is proposed. The results of the experiments under constant current and dynamic discharge conditions show that the proposed method can eliminate the measurement noise and ensure the accuracy of SOC estimation.