California PATH Task Order 4158
Title:   PATH ATMS Testbed Support and Research
Task 2: Proactive Ramp Metering Control
Researcher: Lianyu Chu, Henry Liu
Time: 2002.10 – 2003.12
Project status: On-going

Project Summary: 

This project focuses on developing a proactive ramp metering control strategy based on real-time shockwave estimation and prediction obtained from the time-space analysis of loop detector data.

Project Description:

A shortcoming of existing ramp-metering systems is that they have no ability of traffic flow prediction.  Although traffic patterns often are recurrent, the short-term traffic flow prediction during peak hours is very difficult because of the unstable traffic condition and the frequent disturbances of traffic flow mentioned above.   

The major reasons of freeway congestion are the capacity decrease due to one or more of lane drops, high demands of traffic merging to freeway from entrance ramps and vehicle weaving before off-ramp exits.  For typical urban freeways, there may be several on-ramps and off-ramps within a small stretch of freeway.   These on-ramps and off-ramps are potential bottlenecks if there are high demands of traffic from entrance ramps or exiting to off-ramps.  Freeway bottlenecks always cause the occurrence of a queue, or a backward-moving shockwave; the features of which can be explained by the kinematic wave theory of traffic. The formation and propagation of the backward-moving shockwaves have been observed and analyzed upstream of some freeway bottlenecks.  The propagation of the shockwaves in queued traffic can be treated as a random walk, with predictable statistical variations.  

The proposed method will analyze and predict backward-moving shockwaves based on freeway loop data.  Once a backward shockwave is formed, its corresponding features, such as the speed and its variation, can be estimated based on the current traffic flow condition and the predictive flow pattern.  The time the shockwave reaches each on-ramp can thus be predicted. As a result, the appropriate ramp metering control strategy can be decided.