A reduced-order electrochemical battery model for wide temperature range based on Pareto multi-objective parameter identification method

Abstract

Lithium-ion batteries (LIBs) are critical components of electric vehicles and energy storage systems. However, low ambient temperatures can significantly slow down the electrochemical reaction rate and increase polarization within the battery, resulting in a reduction in capacity and power. In this paper, an accurate reduced-order electrochemical model is developed targeting for a wide temperature range (−20 to 40 °C). The model considers the excess driving force of Li + (de)intercalation in the charge transfer reaction for ion-intercalation materials by adopting adjustment in the Butler-Volmer (BV) equation. Moreover, concentration-dependent solid-phase diffusion coefficients are utilized to improve the accuracy of the model in the voltage recovery session under different charge/discharge rate conditions. To address the multi-objective optimization challenge in parameter identification across a wide range of operating conditions, the Pareto optimization method is employed. The parameters of the proposed model are identified using experimental data under different discharge conditions, including 0.2C, 0.33C, 0.5C, 1C CC discharge, and the UDDS driving cycle. To further validate the model, three dynamic conditions for testing are selected, and the model agrees well with real-world data with an average RMSE of 20 mV at different temperatures and test cycles, exhibiting its capability and robustness in predicting the battery performance under various conditions and temperatures.