Surrogate based multidisciplinary design optimization of lithium-ion battery thermal management system in electric vehicles

Abstract

A battery thermal management system (BTMS) is a complex system that uses various heat removal and temperature control strategies to keep battery packs at optimal thermal conditions, thereby improving the lifetime and safety of lithium-ion battery packs in electric vehicles (EVs). However, an optimal design of BTMS is still challenging, due to its large number of sub-systems and/or disciplines involved. To address this challenge, an air-based BTMS is hierarchically decoupled into four sub-systems and/or sub-disciplines in this paper, including the battery thermodynamics, fluid dynamics, structure, and lifetime model. A high-fidelity computational fluid dynamics (CFD) model is first developed to analyze the effects of key design variables (i.e., heat flux, mass flow rate, and passage spacing size) on the performance of BTMS. Aiming to perform the multidisciplinary design optimization (MDO) of BTMS based on the high-fidelity CFD model, surrogate models are developed using an automatic model selection method, the Concurrent Surrogate Model Selection (COSMOS). The surrogate models represent the BTMS performance metrics (i.e., the pressure difference between air inlet and outlet, the maximum temperature difference among battery cells, and the average temperature of the battery pack) as functions of key design parameters. The objectives are to maximize the battery lifetime and to minimize the battery volume, the fan’s power, and the temperature difference among different cells. The MDO results show that the lifetime of the battery module is significantly improved by reducing the temperature difference and battery volume.