Fast identification method for thermal model parameters of Lithium-ion battery based on discharge temperature rise

Abstract An accurate thermal model of lithium-ion battery is extremely important for the safe operation of electric vehicles. The entropy coefficient is a key thermal characteristic of the battery, which is usually measured in advance. However, traditional measurement methods require a long test time or expensive equipment. In this paper, a novel thermal model parameter identification method with fast and continuous entropy coefficient identification is proposed, in which thermal parameters are obtained with lower time and equipment costs. First, a lumped thermal equivalent circuit model is established to describe the dynamic behaviors of battery temperature. Second, the thermal model parameters are identified based on experiments and calculations. The entropy coefficient is calculated based on temperature rise responses with different discharge currents. Finally, the identification results of the entropy coefficient are compared with traditional methods, which are found to be in good agreement. Furthermore, the thermal models are thoroughly verified under both galvanostatic discharge tests and dynamic profiles over the temperature range from 0 °C to 40 °C. Three kinds of thermal models with different entropy coefficient treatments are compared in terms of temperature errors, which indicates the superiority of the model with dynamic entropy coefficient, especially at high temperatures.