Lithium Battery SOC Estimation Method Based on AFSCKF Algorithm

Aim of this study is to improve the accuracy of the method of estimating the state of charge(SOC) based on the fractional equivalent circuit model(FECM). Based on the fractional model, an adaptive fractional square root cubature Kalman filter(AFSCKF) is designed to estimate SOC. The experiments were compared with fractional cubature Kalman filter(FCKF) under four dynamic cycle conditions. Under the four dynamic cycle conditions, the maximum absolute error(MAE) of AFSCKF is not more than 0.014, the root mean square error(RMSE) is not more than 0.0064, and the MAE of FCKF is not more than 0.02, the RMSE is not more than 0.01. AFSCKF adds a noise estimator based on FCKF, and the transmission of error information is changed from the error covariance matrix to its square root factor. Experimental results show that AFSCKF has better accuracy and robustness, and provides an accurate and reliable method for estimating SOC based on a fractional model.