Real-time Optimal Battery Thermal Management System Controller for Electric and Plug-in Hybrid Electric Vehicles

Nonlinear Model Predictive Control

Masoudi, Yasaman

University of Waterloo

关键词: Nonlinear Model Predictive Control;   

Others  :  https://uwspace.uwaterloo.ca/bitstream/10012/11269/1/Masoudi_Yasaman.pdf

瑞士|英语

来源: UWSPACE Waterloo Institutional Repository

【 摘 要 】

The objective of this thesis is to propose a real-time model predictive control (MPC)scheme for the battery thermal management system (BTMS) of given plug-in hybrid electricand electric vehicles (PHEV/EVs). Although BTMS control in its basic form can be wellrepresented by a reference tracking problem, there exists only little research in the literaturetaking such an approach. Due to the importance of a prediction component in thermalsystems, here the BTMS controller has been designed based on MPC theory to addressthis gap in the literature. Application of the controller to the baseline vehicles is thenexamined by several simulations with di erent optimization algorithms.By comparing the results of the predictive controller with those of the standard rulebased(RB) controller over a variety of driving scenarios, it is observed that the predictivecontroller signi cantly reduces the power consumption and provides a better tracking behaviour.Integrating trip prediction into the control algorithm is particularly important incases such as aggressive driving cycles and highly variable road-grades, where the standardBTMS scheme does not perform as e ectively due to the load current pro le.Moreover, based on the simulation results, the designed controller is observed to have aturnaround time between 10s to 1 ms, and is thus applicable to the real-time automotivesystems.Prosperity of the proposed BTMS control methodology paves the way for the use ofmodel-based (MB) thermal management techniques, not only in future