Battery Thermal Management System for Electric Vehicles: Design, Optimization, and Control

We are witnessing a fast-growing demand in vehicle electrification nowadays due to the widespread environmental consciousness, stringent emission regulations, and carbon neutrality implementation. As one of the most promising energy storage and electrification solutions, lithium-ion battery has been widely employed for electric vehicles (EVs) due to its excellent properties like high energy density, low maintenance, and long cycle life. However, there still exist multiple critical challenges in using lithium-ion battery at large scale as the major power source, such as reliability issues, safety concerns, and especially the range anxiety. Several promising solutions have been explored in the EV industry to mitigate the drawback of range anxiety, such as larger capacity with high energy density and ultra-fast charging. All these approaches challenge the temperature sensitive battery system as a side effect by bringing in extra overburdened waste heat. Given these concerns, battery thermal management system (BTMS) plays an indispensable role in maintaining the maximum temperature and temperature uniformity for EVs. This dissertation proposes a novel J-type air-based cooling structure via re-designing conventional U- and Z- type structures. Aiming to further improve the thermal performance, a surrogate-based optimization framework with two-stage cluster-based resampling is developed for BTMS structural optimization. Compared with the U- and Z- type, the novel J-type structure is proved with significant advancements. Based on the optimized J-type configuration, an operation mode switching module is designed to mitigate the temperature unbalance by controlling the opening degree of two outlet valves. Tested by an integrated driving cycle, results reveal that the J-type structure with its appropriate control strategy is a promising solution for light-duty EVs using an air cooling technology. Improving the energy efficiency is another potential approach to mitigate range anxiety. In this dissertation, a model predictive control (MPC)-based energy management strategy is developed to simultaneously control the BTMS, the air conditioning system, and the regenerative power. A vehicle velocity forecasting framework is integrated with the MPC-based energy management to further improve the energy efficiency. Deep learning and image-based traffic light detection techniques have been leveraged for velocity forecasting. Results show that the proposed energy management method has significantly improved the overall EV energy efficiency.