Connected Automated Traffic Systems

Organizers:  H. Yu

Abstract:

Being the lifeblood of modern cities, transportation systems play an essential role that integrates economic, social and environmental dimensions of modern society. Over the recent decades, the development of transportation system is undergone profound and rapid disruptions, led by several revolutions taking place in parallel, in particular, automation and connectivity. The proposed invited session focuses control and estimation problems of traffic flow systems that are empowered by various connected and automated technologies.

The traffic flow systems refer to vehicular traffic flow. The interactions of the vehicles and drivers and their collective effects are described with traffic flow dynamics. The systems are usually categorized into either macroscopic Partial Differential Equation (PDE) models or microscopic Ordinary Differential Equation (ODE), with respect to aggregation level and mathematical structure. From the macroscopic level of traffic flow systems, control and estimation problems focus on developing various control methods to mitigate traffic congestion, via road-based traffic management infrastructures such as ramp metering and varying speed limits, as well as traffic sensing infrastructures like loop detectors. There is rich literature on road-based control methods that are developed for PDE traffic flow models, ranging from the early feedback rules to sophisticated control frameworks. From the microscopic vehicle level of traffic flow systems, the recent technological advancements towards autonomous and connected vehicles (CAV) bring new opportunities for traffic flow control and estimation via vehicle-based measures. There are surging research interests in developing new control and estimation algorithms for CAVs which are employed as moving sensors and actuators to regulate traffic systems.

The contributed papers not only consist of model-based control methods but also data- driven methods for stabilizing traffic flow. In addition, both the theoretical results and experimental validation are presented in the proposed invited session. Therefore, the invited session provides a comprehensive view for both the traffic flow systems and methodologies for control and estimation problems.

The proposed invited session will bring together a shared research interest of applying various control methodologies to intelligent traffic management applications, and will provide diverse perspectives on this topic, with respect to the choice of control methods, traffic models and objectives of interest. The proposed invited session will be of great interest to two types of conference attendants: the ones who are mainly interested in control theory and are looking for relevant applications for motivation; and the ones who focus on solving various transportation management problems and connected and automated vehicles.