Predictive Energy Management in Networked Vehicles: Exploiting Traffic and Terrain Preview for Fuel Saving
This presentation summarizes our recent findings on the role that preview information of terrain, traffic signal timing, and traffic flow can have in fuel savings of “networked’’ vehicles with conventional or hybrid powertrains. Terrain and traffic information is now available through various providers and can be integrated into the vehicle navigation system or into its add-on accessories. However systematic methods for utilizing this rich and dynamic information have not been explored in the past. We present methodologies that plan over time the best vehicle velocity profile (or battery utilization for hybrid vehicles) that reduces fuel consumption and emissions with minimal influence on trip time. We cast these problems in a dynamic constrained optimization framework with partial future information. Standard numerical solution via dynamic programming is viable but requires full information of future events. To complement the partial future information, one can create predictive traffic models that use streaming traffic data to forecast evolving traffic patterns down the road. Receding finite-horizon optimization will be also explored as a mean to tackle missing information and the computational cost of dynamic programming.
Ardalan Vahidi received his Ph.D. in Mechanical Engineering from the University of Michigan, Ann Arbor in 2005. He had obtained his B.S. and M.S. in Civil Engineering in 1996 and 1998 from Sharif University of Technology and his second M.S. in Transportation Safety in 2002 from George Washington University, Washington DC. He is now an Assistant Professor of Mechanical Engineering at Clemson University in South Carolina, USA. His current research interests are in optimization-based control methods and control of vehicular and energy systems.