University of California, Riverside

Department of Electrical and Computer Engineering

Qiu Jin - a Talk on "Eco-Friendly Agent-Based Advanced Traffic Management Techniques in a Connected Vehicle Environment"

Qiu Jin - a Talk on "Eco-Friendly Agent-Based Advanced Traffic Management Techniques....

Qiu Jin - a Talk on "Eco-Friendly Agent-Based Advanced Traffic Management Techniques in a Connected Vehicle Environment"

August 28, 2015 - 10:30 am
Winston Chung Hall, 202

Transportation is responsible for one third of greenhouse gases (GHG), as well as a major source of other pollutants including hydrocarbons (HC), carbon monoxide (CO), and nitrogen oxides (NOX). Existing transportation systems are facing numerous issues resulting from the increased travel demands and limited capacities of roadway infrastructures. As wireless communication advances, agent-based system techniques provide a new perspective to advanced traffic management system (ATMS) and has received increased attention from researchers and engineers in the field of transportation. In both urban arterial and highway networks, vehicles and road infrastructure interact with each other as individual intelligent agents in an integrated environment (called multi-agent system), which can significantly improve the overall traffic performance in terms of safety, mobility and environmental sustainability, due to knowledge sharing and system-wide decision-making. In this dissertation, we propose a variety of environmentally-friendly agent-based advanced traffic management technologies in a connected vehicle environment.

For agent-based arterial traffic management, we examine the concept of ATMS for connected vehicles using a multi-agent systems approach, where both vehicle agents (VAs) and intersection management agents (IMAs) can take advantage of real-time traffic information exchange, and develop an agent-based hierarchical structure for signal-less intersection management system. From the perspective of IMAs, they receive probe vehicle data from VAs, dynamically schedule VAs’ arrival times (by potentially grouping VAs in platoons), reserve intersection time-space occupancies for VAs, and communicate arrival time advices back to VAs. Furthermore, an optimal lane selection algorithm for agent-based traffic management system is developed, which could provide guidance on determining optimal target lanes for individual vehicle agent in order to better regulate traffic flow, thus achieving a system-wide optimal solution in terms of maintaining desired traffic speeds.

On the other hand, vehicle agents use advices to plan their trajectories in order to further minimize energy consumption and pollutant emissions by considering vehicle dynamics (e.g., engine efficiency map), roadway grade and other constraints (e.g., traffic signal status). Two kinds of dynamic longitudinal control algorithms have been developed. Firstly, an Eco-Approach and Departure algorithm is introduced and field test has been conducted in Turner Fairbank Highway Research Center on an automated vehicle. Secondly, a power-based approach is used to develop an optimal vehicle longitudinal control algorithm for individual vehicles with specific engine types in order to maximize fuel economy under a variety of traffic conditions.

For freeway traffic management, we introduce a dynamic driving speed advisory system, where advice is given in real-time to drivers according to changing traffic conditions in the vehicle’s vicinity, by taking advantage of real-time traffic sensing and telematics. A driving simulator study is conducted with truck drivers to evaluate the energy and emissions benefits as well as study the behavioral impact eco-driving may have on truck drivers.



More in Defenses

More Information 

General Campus Information

University of California, Riverside
900 University Ave.
Riverside, CA 92521
Tel: (951) 827-1012

Department Information

Electrical and Computer Engineering
Suite 343 Winston Chung Hall
University of California, Riverside
Riverside, CA 92521-0429

Tel: (951) 827-2484
Fax: (951) 827-2425
E-mail: E-mail/Questions