Heterogeneous traffic flow: how agent interactions shape collective properties
Research Assistant Professor, Texas Tech University
This seminar took place on November 1, 2019.
Collective properties of traffic flow, such as its equilibrium, aggregate dynamics and stability, are determined by attributes of agents (i.e. drivers/vehicles) as well as how the agents interact. Understanding connections between the two is crucial to control and operations, e.g. towards designing mechanisms to make mixed traffic flow of autonomous and human-driven vehicles self-organize and self-stabilize. In this talk, Jia Li presented recent research in this direction. In the first part, Li provided an explicit characterization of equilibriums attainable by heterogeneous traffic flow in multilane settings, where one class of agents are “type-sensitive”, a property that autonomous vehicles may likely be endowed with. In the second part, Li presented simulation evidence along with a heuristic analysis towards explaining spontaneous platoon formation in heterogeneous traffic flow and the role of opportunistic agent behaviors. Finally, Li discussed implications of these results from a control perspective.
Jia Li is now a Research Assistant Professor at Texas Tech University. His research focuses on the intersection of traffic flow theory, transportation systems modeling, and connected and autonomous vehicles. He received Ph.D. from University of California, Davis in 2013, under the supervision of Professor Michael Zhang. Then he joined University of Texas at Austin and worked with NAE member, Professor Michael Walton. He published twenty peer-reviewed papers and contributed to a dozen of proposals that secured around two-million research funds. His current research is sponsored by a seed grant from Hurricane Resilience Institute and an internal grant from Texas Tech University.