| dc.description.abstract | Electric vehicles (EVs) are estimated as the most sustainable solutions for future transportation requirements. However, there are various problems related to the battery pack module and one such problem is invariable high-temperature differences across the battery pack module due to the discharging and charging of batteries under operating conditions of EVs. High-temperature differences across the battery module contribute to the degradation of maximum charge storage and capacity of Li-ion batteries which ultimately affects the performance of EVs. To address this problem, a finite element modeling (FEM) based automated neural network search (ANS) approach is proposed. The research methodology constitutes of four stages: design of air-cooled battery pack module, setup of the FEM constraints and thermal equations, formulating the predictive model on generated data using ANS, and lastly performing multi-objective response optimization of the best fit predictive model to formulate optimum design constraints for the air-cooled battery module. For efficient thermal management of the battery module, an empirical model is formulated using the mentioned methodology for minimizing the maximum temperature differences, standard deviation of temperature across the battery pack module, and battery pack volume. The results obtained are as follows: (1) the battery pack module volume is reduced from 0.003279 m3 to 0.002321 m3 by 29.21%, (2) the maximum temperature differences across the eight cells of battery pack module declines from 6.81 K to 4.38 K by 35.66%, and (3) the standard deviation of temperature across battery pack decreases from 4.38 K to 0.93 K by 78.69%. Thus, the predictive empirical model enhances the thermal management and safety factor of battery module. | |
