01049567

Component

Quick and reliable detection of incidents is of great importance for incident management. In this paper we assess if individual vehicle data gathered by double loop detectors can be used for the purpose of incident detection. To achieve this, loop detector data for a large number of neighboring detectors on three different Dutch freeways for two months in 2005 have been collected. Furthermore, a database is available that contains information about all reported incidents on the considered freeways. Using these data, first of all we verify if we can identify lane and traffic regime specific upperbounds for time headways under normal conditions such that a single headway exceeding these values can be used as an indicator for an incident. Second of all, we test if it is possible to detect an incident on the basis of changes in headway distributions. A similar analysis is carried out for the speeds at which individual vehicles pass a detector. The results show that it is possible to define useful upperbounds for the time headways, that can be used for incident detection, given that the traffic volume is larger than 900 veh./hour/lane. Also in comparing the headway and speed distributions under normal and incident conditions, we find clear differences between both conditions. Based on the data analysis we propose three incident detection methods using headway measurements. One method is based on upperbounds for time headways, and the other two are based on headway distributions. We find promising results in testing and cross-comparing these methods.

01042056

07-0744

English

14p

2007

Transportation Research Board 86th Annual Meeting

CD-ROM

Figures (1) ; References (7) ; Tables (2)

Freeway operations; Freeways; Headways; Highway operations; Incident detection; Incident management; Loop detectors; Traffic incidents; Traffic volume

Traffic regimes

Netherlands

Highways; Operations and Traffic Management; Planning and Forecasting; I72: Traffic and Transport Planning

Transportation Research Board Annual Meeting 2007 Paper #07-0744

TRIS, TRB

Feb 8 2007 5:19PM