Optimization of Thermal Management in Modern Electric Vehicle Battery Cells Employing Genetic Algorithm

Abstract

Optimization of thermal performance processes using genetic algorithm (GA) combined with some commercial software or other soft computing methods like artificial neural networks are common in many heat transfer applications with the exception of battery thermal management. In this article, a novel and innovative approach for single-objective optimization using GA combined with in-house developed finite volume method (FVM)-based code is investigated. Three important thermal and fluid flow performance parameters of modern electric vehicle Lithium–ion battery cells, namely, average Nusselt number (Nuavg), friction coefficient (Cf,avg), and maximum temperature (T¯max) are optimized. The operating parameters considered for optimization include heat generation term (S¯q), Reynolds number (Re), conduction-convection parameter (ζcc), aspect ratio (Ar), and spacing between the cells (W¯ff) varying in some selected range. Optimization in case of internal flow between the battery cells and external flow over the battery cell is performed. Computational time taken by the combined GA and FVM code for 5, 10, 15, and 20 iterations in case of internal and external flow is also presented. From the complete optimization analysis, it is found that for higher charging/discharging rates at which the heat generation is very high, T¯max can be kept within the safe limit, Nuavg to maximum and Cf,avg to a minimum with a slight compromise in pumping power requirement to circulate the coolant in internal flow. For external flow analysis, Re and ζcc in a selected medium range will provide optimized thermal and fluid flow situations.