A novel hierarchical parameter identification method for electrochemical-thermal model of Li-ion battery

Lithium-ion batteries electrochemical-thermal model has promising applications as it can provide an in-depth description of the battery's internal state. However, the model accuracy depends on precise parameter identification. In addition, the study of the electrochemical-thermal characteristics of the battery is helpful to the design of the battery thermal management system. Current methods rarely consider temperature effects on electrochemical model parameters, and the computational efficiency needs to be improved. The electrochemical-thermal characteristics of the battery have not been further studied. To address these issues, this paper considers both temperature and voltage errors in the parameter identification process, and innovatively proposes a hierarchical parameter identification method, enhancing identification accuracy while reducing the identification time. The method was validated through 12 constant current and 6 dynamic (HWFET and US06) current conditions. The optimized battery model was used to analyze electrochemical-thermal characteristics under various ambient temperatures and discharge rates. The results indicate that the average values of the mean relative errors for battery voltage and temperature are 1.43 % and 2.42 %, respectively, across all tested conditions. Compared to traditional method, errors have decreased by 74.76 % (voltage) and 45.45 % (temperature), and the calculation time has decreased by 52.89 %. Additionally, the study revealed that higher discharge rates accelerate changes in electrochemical parameters, while lower temperatures reduce the exchange current density, which significantly affects battery overpotential and internal resistance. This results in increased heat generation and temperature rise. This study may provide a new perspective for battery model parameter identification and analysis of electrochemical-thermal characteristics.