Yash Mathur
In transitioning from internal combustion engines (ICE) to electric batteries, vehicle thermal management performance requirements have changed substantially. While ICE vehicles rely on a pumped oil loop for engine cooling, battery electric vehicles (BEVs), require active cooling via the vapor compression system (VCS) that is also used to deliver cooling to the passenger cabin. The inherent dynamic coupling of the dual-evaporator VCS, as well as a substantial difference in timescales of the battery dynamics compared to those of the cabin, makes the thermal regulation problem a difficult one to tackle. In this thesis I design a model-based nonlinear model predictive controller (MPC) to optimally control a dual-evaporator VCS to meet performance requirements for maintaining cabin comfort and appropriate thermal conditions for the battery, over standard drive cycles and varied weather conditions. Through two simulated case studies, I demonstrate an optimal approach to mitigating the thermal management problem, particularly during off-nominal operating conditions. I then compare the performance of the proposed optimal controller against a baseline approach that is representative of the current state-of-the-art. The results show that the proposed approach achieves improved setpoint tracking, disturbance rejection, and energy efficiency.