Enhanced state-of-charge estimation for lithium-ion iron phosphate cells with flat open-circuit voltage curves

The open-circuit voltage (OCV) forms the basis for many real-time state-of-charge (SOC) estimation algorithms. The OCV-SOC relationship for most battery chemistries often provides a good estimate for SOC. However, for the lithium-ion iron phosphate (LiFePO4) variation of the lithium-ion cell chemistry, the OCV curve is fairly flat over the operational SOC range. Thus, even the smallest error in the OCV obtained from a battery model can lead to divergence in SOC from the actual value. Therefore, as a remedy this paper presents a separated framework for the Extended Kalman Filter (EKF) estimation of SOC, with real-time process noise assessment. In this paper, the states and parameters of a non-linear cell model, namely the two time-constant Randle's model, are also identified using the dual implementation of the EKF algorithm in real time.