Innovative lumped-battery model for state of charge estimation of lithium-ion batteries under various ambient temperatures

Ambient temperature alters properties of lithium-ion batteries and affects the accuracy of estimation of state of charge (SOC), which is an important function to ensure the safety and reliability of electric vehicles. An accurate SOC estimation is critical under various temperatures. The existing methods have two problems: 1) need to perform time-consuming pre-experiments to investigate influence of various temperatures on the battery properties, and 2) use of the Thevenin model which is inaccurate at sub-zero temperatures. This study proposes an innovative lumped-battery model to improve the accuracy of both SOC estimation and battery modeling without pre-experiments. Two main causes of modeling errors by the Thevenin model are analyzed. Proposed model parameters are estimated using the recursive least squares, and the extended Kalman filter is used to estimate SOC in real time. Experiments are conducted under time-invariant and time-varying temperature conditions ranging from -10 degrees C to 30 degrees C. The results indicate that relative errors of battery modeling are less than 2.4% and that estimation errors of SOC are at most 0.4% under various temperatures. Therefore, the proposed method can be conveniently and widely applied to all-climate battery management systems to achieve a high accuracy of SOC estimation. (C) 2021 Elsevier Ltd. All rights reserved.

Automated summary

This study proposes an innovative lumped-battery model to improve the accuracy of SOC estimation and battery modeling without pre-experiments. The model parameters are estimated using recursive least squares, and SOC is estimated in real-time using the extended Kalman filter. Experiments show that the proposed method achieves high accuracy in various temperature conditions.