Surrogate based multi-objective optimization of J-type battery thermal management system

This paper proposes a novel and flexible J-type air-based battery thermal management system (BTMS), by integrating conventional Z-type and U-type BTMS. With two controlling valves, the J-type BTMS can be adaptively controlled in real time to help balance the temperature uniformity and energy efficiency under various charging/discharging situations (especially extreme fast changing). Results of computational fluid dynamics simulations show that the J-type system performs better than the U-type and Z-type systems. To further improve the thermal performance of the proposed J-type BTMS, a surrogate-based multi-objective optimization is performed, with the consideration of the two major objectives, i.e., uniformity and energy efficiency. The concurrent surrogate selection (COSMOS) framework is adopted in this paper to determine the most suitable surrogate models. Optimization results show that: (i) the uniformity of the temperature distribution is improved by 38.6% compared to the benchmark, (ii) the maximum temperature is reduced by 19.1%, and (iii) the pressure drop is decreased by 14.5%.