Performance of state of charge estimation model-based via adaptive extended Kalman filter

Accurate estimation of the state of charge (SOC) of batteries is an essential task of the battery management system (BMS). The effectiveness of the adaptive extended Kalman filter (AEKF) model-based observer for SOC estimation of the dynamic behavior of the battery is investigated. The SOC is a reflexion of the chemistry of the cell; it is the key parameter for the BMS. In this paper, three equivalent circuits models (ECMs) have been established and their parameters were identified by applying the least square method. However, the relationship between open circuit voltage (OCV) and SOC have been proposed by four mathematical functions model-based. In fact, the SOC estimation accuracy of the battery depends on the model and the efficiency of the algorithm. The AEKF method is used to estimate the SOC of Lead acid battery. The experimental data is employed to identify the parameters of the three models and used to build different open circuit voltage-state of charge (OCV-SOC) functions relationship. The results show that the SOC estimation based-model on high order polynomial and third-order equivalent circuit can effectively limit the error, thus, guaranteeing the accuracy and robustness. © JES 2019.