A novel online efficient SOC estimation strategy for mining lithium-iron phosphate batteries

Accurate estimation of the state of charge is the cornerstone for battery management systems to achieve effective battery management and control. However, due to the challenging nature of directly measuring real-time SOC, it is typically obtained indirectly by processing external parameters. This paper introduces an innovative SOC estimation method called the VFFRLS-UKF algorithm in detail. Based on the classical Thevenin battery model, this algorithm first employs the variable forgetting factor recursive least squares (VFFRLS) algorithm for online model identification. The SOC is then accurately estimated in conjunction with the unscented Kalman filter (UKF). In order to verify the practicality and accuracy of the VFFRLS-UKF algorithm, a comprehensive validation was carried out under two working conditions, BBDST and DST, respectively.