Hybrid Controls Development and Optimization of a Fuel Cell Hybrid Powertrain

• Main Author: Koch, Alexander Karl
• Language: en
• Published: 2012
• Subjects: Hydrogen; Fuel Economy; Fuel Consumption; PHEV; HEV; EV; Batteries; Lithium Ion; Fuel Cells; Powertrain; Controls; Software; Vehicle; Optimization; Modeling; Simulations; Electric Motors; DCDC Converters; Mechanical Engineering
• Online Access: http://hdl.handle.net/10012/7115

The University of Waterloo Alternative Fuels Team’s participation in EcoCAR: The Next Challenge provided an unparalleled opportunity to execute advanced vehicle technology research with hands on learning and industry leading mentoring from practicing engineers in the automotive industry. This thesis investigates the optimization of the hybrid operating strategy on board the EcoCAR development vehicle. This investigation provides the framework to investigate the pros and cons of different hybrid control strategies, develop the model based design process for controls development in a student team environment and take the learning of this research and apply them to a mule development vehicle. A primary controls development model was created to simulate software controls before releasing to the vehicle level and served as a tool to evaluate and compare control strategies. The optimization routine was not directly compatible with this model and so a compromise was made to develop a simplified vehicle model in the MATLAB environment that would be useful for observing trends but realizing that the accuracy of the results may not be totally consistent with the real world vehicle. These optimization results were then used to create a new control strategy that was simulated in the original vehicle development model. This new control strategy exhibited a 15% gain in fuel economy over the best case from the literature during an Urban Dynamometer Driving Schedule (UDDS) drive cycle. Recommendations for future work include adding charge depletion operation to the simulation test cases and improving the accuracy of the optimization model by removing the simplifications that contributed to faster simulation time. This research has also illustrated the wide variability of drive cycles from the mildly aggressive UDDS cycle having 5 kilowatts average propulsion power to the very aggressive US06 cycle having 19 kilowatts average propulsion power and their impact on the efficiency of a particular control strategy. Understanding how to adapt or tune software for particular drive cycle or driver behaviour may lead to an interesting area of research.