State-of-Charge Estimation Using an EKF-Based Adaptive Observer

• Published in: IEEE transactions on control systems technology Vol. 27; no. 5; pp. 1907 - 1923
• Main Authors: Afshar, Sepideh, Morris, Kirsten, Khajepour, Amir
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.09.2019; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Adaptation models; Adaptive filters; Adaptive observer; Battery charge measurement; Computer simulation; Electric vehicles; electrochemical equations; Energy storage; Extended Kalman filter; Kalman filter (KF); Lithium; Lithium-ion batteries; low-order model; Mathematical model; Model accuracy; Observers; Rechargeable batteries; Solid modeling; State of charge; state-of-charge (SOC) estimation; Storage batteries; variable solid-state diffusivity model
• ISSN: 1063-6536; 1558-0865
• DOI: 10.1109/TCST.2018.2842038

Lithium-ion batteries are used to store energy in electric vehicles. State of charge (SOC) is an important quantity of the battery cells that need to be estimated using limited measurements. In this paper, SOC estimation via an electrochemical model, a physics-based model, is considered. For lithium iron phosphate cells, a variable solid-state diffusivity model provides significantly more accuracy, but this complicates the model further. A previously obtained, simplified but still a physics-based model is used in this paper. An extended Kalman filter (KF)-based adaptive observer is designed via a low-order approximation of this electrochemical model. The predictions of the estimator are compared with the experimental data in simulations. The simulations are efficient and more accurate than a standard KF.