Applied Mathematics Colloquium: Lili Du (UF)

Title: Platoon Centered Control (PCC) for Cooperative Driving Automation Built upon MPC, Online Learning and Distributed Optimization

Abstract: The recent advancement of connected and autonomous vehicle (CAV)

technologies provides tremendous opportunities to mitigate urban traffic

congestion through innovative traffic flow control and operations. Particularly,

supported by advanced sensors, communication, and portable computing devices, CAVs can sense, share, and process real-time mobility data and conduct cooperative or coordinated driving. This has led to a surging interest in self- driving technology. However, the self-driving technology focusing on assisting individual driving maneuvers mainly ensures neighborhood traffic efficiency and an individual vehicle’s safety; it does not always ensure traffic flow efficiency when a group of self-driven vehicles is packed in a complex freeway system with multiple lanes, on-ramps, and off-ramps, etc. Motivated by this view, our research team seeks to develop an innovative Platoon Centered Control (PCC) approach for CAV platooning driving, leveraging online learning and distributed optimization. 

This talk will present our recent research on coordinated vehicle platooning control under various scenarios, such as local platoon formation, car following, lane change, and passing signal intersections in a complicated transportation system. The methodology development combines traffic flows, MPC, optimization, machine learning, and distributed computation. Our numerical experiments confirmed that this platoon-oriented driving will significantly improve traffic efficiency, smoothness, and safety.