Influence of uncertainties in a battery pack with air cooling for electric vehicles on temperature difference and volume of battery module

초록

Heat generation in the battery packs in electric vehicles affects the battery's overall performance. Researches available focus on predicting the temperature difference between the cells and the volume of the battery module for different arrangements of battery cells. However, the actual dimensions of the battery cells may have some uncertainties while constructing them compared to what is planned. Large deviations in the geometric properties may affect the overall performance of the battery module. The present work predicts the influence of uncertainties in the geometry of the battery cells and the mass flow rate of cooling air over the maximum temperature difference of the cells in the battery (TD) and the volume of the whole battery module (V). At first, the surrogate model is trained to predict the values of the TD and V in the framework of the Gaussian Process Regression (GPR) machine learning algorithm, followed by the artificial introduction of noise in the dataset using Bootstrapping. The noisy data is then fed to the trained GPR to predict the influence of the noise on TD and V. The performance of the battery under stochastic conditions is most affected by the dimension of the battery cells from the lower board. Also, the bootstrapping is found to be computationally efficient compared to the MonteCarlo simulations for accessing the uncertainties in the battery pack.

키워드