Development of Optimal Ramp Metering Strategies

Investigators:

Dr. Angshuman Guin, Senior Research Engineer, Georgia Tech
Dr. Jorge Laval, Assistant Professor, Georgia Tech
Rama Chilukuri, Ph.D. student, Georgia Tech

Project Overview:

The Federal Highway Administration and several state DOTs have identified ramp metering as one of the key congestion mitigation strategies. According to the Texas Transportation Institute’s 2009 Mobility Report, ramp metering in the 25 cities reduced delay by 39.8 million person hours in 2007. However, the potential of ramp metering is currently under-utilized with metering rates based on local conditions, limited consideration to linking all the on-ramps along a corridor, and no consideration for the type of bottlenecks such as merge, diverge, etc. Thus, sub-optimal operation results in excessively long queues on the entrance ramps, and insufficient reduction in system delay.

This project is developing optimal ramp metering strategies using the recent advances in traffic flow theory and simulation. A recently developed theory that can explain stop-and-go oscillations, capacity drop during congestion, and traffic dynamics under ramp metering, will be utilized to perform realistic simulations to evaluate new strategies.

The System-Wide Adaptive Ramp Metering (SWARM) algorithm will be studied and a set of ramp-metering strategies for congestion mitigation will be proposed. Although different strategies are expected to have different impacts under different traffic conditions and network configuration scenarios, the research will focus on peak demands to design the strategies. The microscopic simulation program that will be used for this purpose is being built at Georgia Tech.  It is essential to use a specialized simulation package instead of an off-the-shelf commercial simulation tool since none of the off-the-shelf packages model the traffic dynamics at congested ramp merges and diverges with sufficient accuracy. [For an online demo of the proposed simulation model please visit http://trafficlab.ce.gatech.edu and click on “Freeway ramp-metering” on the left.].  The final report will contain guidelines and the methodological steps that are required to optimize operations on a corridor and eliminate a significant portion of the recurrent congestion in the network.   

This project is part of GTI/UTC project 08-05, Integration of Real Time, Fixed Sensor Data and Simulation/Development