Prediction Model of Thermal Behavior of Lithium Battery Module Under High Charge-Discharge Rate

Abstract

In order to achieve accurate thermal prediction of lithium battery module at high charge and discharge rates, experimental and numerical simulations of the charge-discharge temperature rise of lithium battery cells at lower rates of 1C, 2C, and 3C have been conducted firstly to verify the accuracy of the NTGK model (Newman, Tiedemann, Gu, and Kim, NTGK) at low rates. The experimental results of temperature rise at a 5C rate has matched the experimental results and prove the accuracy of the model, and the maximum temperature prediction difference of the GPR-NTGK model (Gaussian process regression, GPR) has been reduced by 4.86K compared with the original model. Further thermal prediction and temperature rise experiments have been conducted for the battery module at 1C, 2C, 3C and 5C rates, and the simulation results are in good agreement with the experimental results, which verified the applicability of the GPR-NTGK model for the battery module. The maximum temperature at 5C rate is 334.32K, which is beyond the safe working temperature range of Li-ion batteries, and the high temperature area is concentrated in the lower center of the module, with poor thermal uniformity and a great risk of thermal runaway.

Keywords: charge-discharge at high rate, heat prediction, lithium-ion battery, electrochemical-thermal model, gaussian process regression