State of health estimation of the LiFePO4 power battery based on the forgetting factor recursive Total Least Squares and the temperature correction

Abstract

The decline of the lithium-ion power battery's State of Health (SOH) with usage significantly impacts other state estimation results, such as State of Charge (SOC). Hence, accurate estimation of the lithium-ion power battery's SOH holds vital importance in the battery management system. This paper proposes a SOH estimation method for the lithium-ion power battery, utilizing the Forgetting Factor Recursive Total Least Squares (FFRTLS) and incorporating the temperature correction. The FFRTLS effectively addresses the SOC estimation errors and the terminal current measurement noise simultaneously. The temperature correction method, based on the Arrhenius equation, corrects the influence of the ambient temperature during the SOH estimation process, ensuring that the ambient temperature does not affect the accuracy of the SOH estimation results. Additionally, the capacity convergence coefficient enhances the reliability of the SOH estimation results by preventing abrupt changes of the maximum available capacity. Experimental results on a LiFePO4 power battery under diverse working conditions and varying ambient temperatures, validate the effectiveness of the proposed method. The evaluation indexes, including Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and Maximum Absolute Error (Max-AE), demonstrate the high accuracy of the SOH estimation results, with all indexes below 0.21%, 0.25% and 0.35% respectively.