Artificial neural network-based multi-objective optimization of cooling of lithium-ion batteries used in electric vehicles utilizing pulsating coolant flow

Abstract

Due to the enhancement of production and development of electric vehicles, their improved performance has attracted much attention. Lithium-ion battery packs that supply the energy needed for electric vehicles require cooling to enhance their function, reduce costs and increase lifetime. The current work investigates the effect of using pulsating flow instead of steady flow on batteries' maximum temperature and temperature difference. In this respect, two different arrangements have been utilized with a commonly used cylindrical lithium-ion battery named 18650. The results demonstrate that pulsating flow decreases the maximum temperature and maximum temperature difference by 27% and 50%, respectively, in an optimal setting. The influence of three main characteristics of pulsating flow on the battery pack cooling system has been studied. These characteristics are the Strouhal number and the oscillation amplitude, while the Reynolds number is taken as a parameter. In both arrangements, by increasing the amplitude, maximum temperature and maximum temperature difference are reduced. As expected, increasing the Reynolds number would improve cooling performance. The effect of the Strouhal number on the maximum temperature and maximum temperature difference indicates the existence of extremum points in both arrangements. Moreover, the optimum points in both battery packs are different, which means the optimum point depends on the geometry. A double-objective optimization was carried out using a surrogate model obtained by artificial intelligence to find the Pareto front for minimizing both the maximum temperature and the maximum temperature difference. Moreover, for the best points on the Pareto front, the optimal Strouhal number in the staggered and in-line arrangement of batteries are 0.5457 and 0.369, respectively.