Surrogate model-based multiobjective design optimization for air-cooled battery thermal management systems

A well-designed battery thermal management system (BTMS) can achieve optimal cooling performance with less power consumption than a poorly-designed system. However, it is difficult to use the computational fluid dynamics (CFD) method perform an effective and design of BTMSs when there are several structural parameters multiple evaluation criteria. In this paper, instead CFD, compound surrogate model based on mixture experts (MoE) developed accurately approximate BTMS different configurations. Then, criteria transformed into multiobjective optimization (MOO) problem, which solved by nondominated sorting genetic algorithm II (NSGA-II). To address nonuniqueness solutions contradiction between criteria, entropy weight (EWM) importance through intercriteria correlation (CRITIC) applied analyze each criterion. Finally, obtained for corresponding weights. The results show that surrogate-based MOO find meets expectations, approach provide guidelines BTMSs.