State of charge estimation of lithium-ion phosphate battery based on weighted multi-innovation cubature Kalman filter

State of charge (SoC) is one of the most important parameters of battery manage system (BMS). To obtain a better result of SoC estimation, an accurate battery model and effective SoC estimation algorithm are indispensable. In this work, a Thevenin model is established and linear Kalman filter (LKF) is used for online parameters identification. Results show that convergence of parameters identification of LKF is better than that of the recursive least squares (RLS) and recursive least squares with forgetting factor (FFRLS). In addition, the equilibrium potential equation (EPE) is used to fit the relationship between open circuit voltage (OCV) and SoC instead of 9th order polynomial. Then, by the weighted calculation of the innovation vector based on error distribution and time distribution, the weighted multi-innovation cubature Kalman filter (WMICKF) is proposed for the SoC estimation. The experimental data are obtained based on a prismatic LiFePO4 battery, which is tested under urban dynamometer driving schedule (UDDS) test and new European driving cycle (NEDC) test at room temperature. The results show that the WMICKF outperforms the classical cubature Kalman filter (CKF) and the multi-innovation CKF (MICKF). The SoC estimation error of WMICKF can be limited within 1% (0.91%), which is better than 1.30% of CKF and 2.71% of MICKF. In addition, the mean absolute error (MAE), root mean square error (RMSE) and coefficient of determination (R-square) are calculated for comprehensive assessment of the proposed method. Furthermore, the robustness of WMICKF is verified under different initial SoC error and different types of noise disturbance. Results show that under these uncertain factors, the proposed WMICKF is still reliable for accurate SoC estimation.