Abstract: Accurate battery model is the basis for the state estimation and energy management of lithium-ion power battery systems. Benefit from its wide application and high accuracy, the electrochemical mode is urgently needed for the next generation battery management systems. However, as the electrochemical model has many parameters that are difficult to acquired, calibration of the electrochemical model has large influence on model accuracy. To solve this problem, an efficient approach for electrochemical-thermal coupled modeling of large-format lithium-ion battery is proposed. First of all, the parameters of the electrochemical model are classified into several categories. The measurable/identifiable parameters (geometric dimensions, initial stoichiometry coefficient of cathode and anode, and maximum lithium intercalation concentration of solid particles) are accurately measured and identified through tests and post-mortem analysis. Furthermore, the pulsed-charge tests at different temperatures are performed to calibrate the solid-phase diffusion coefficient D_s and the reaction rate constant k of the electrodes. Battery thermal model is also built to simulate battery heat generation, and the electrochemical-thermal coupled model is finally established with the consideration of temperature effects. Also, the electrochemical-thermal coupled model is validated by the following three tests:charging and discharging tests at different rates, pulse-discharge tests at different temperatures and dynamic stress test(DST). The results show that the model has high accuracy and adaptability at different C-rates and temperature, with the average error of voltage is less than 10 mV, and the average error of temperature is lower than 1.1℃. Finally, the modeling analysis results show that the decrease of the solid-phase diffusion coefficient D_s will increase the difference of lithium ion concentration inside the electrode particles, leading to the higher particle surface potential and thus earlier termination of discharge process and lower capacity. Besides, the reduction of reaction rate constant k mainly affects battery resistance, which will cause a overall downward movement of the battery discharge voltage curve and also an increase of heat generation. The modeling method proposed can efficiently establish an electrochemical-thermal model coupled with high-accuracy, easy-applicability, and fair-economy.

Key words: lithium-ion battery, battery management, electrochemical-thermal coupled model, parameter identification, sensitivity analysis of parameter