Definition and Measurement of Transportation System Performance

Problem

The research sought to improve traffic incident detection and congestion monitoring for more accurate and reliable measurement of traffic system performance

Method

This project presents a new set of advanced traffic surveillance techniques based on inductive loop detector (ILD) technology currently in widespread use. The emphasis is on applying new algorithms to inductive loop detector data to obtain vehicle "signatures." By correlating the vehicle signature data from two sets of loop detectors, one "upstream" and one "downstream," data can be obtained on the traffic characteristics of a section of the transportation network.

The research included development of a process for extracting vehicle feature vehicle pattern matching information from the inductive loop detector data so that information on traffic density and travel time data could be obtained. A video system was used to validate the inductive loop detector data and vehicle pattern matching accuracy.

Findings

The section-related data generated by the system developed for this research project provides a quantitative measure of traffic density, which can be used as a direct measure of traffic demand. Traffic density also gives a good idea of the "closeness" of vehicles, which can reveal the psychological comfort of drivers. For example, the stress level associated with free-flowing traffic is much less than that of "stop-and-go" traffic.