01594120

Component

Traffic congestion on roadways seriously affect travel experience and cause economic and environmental problems. Part of the recurrent congestion is due to roadway bottlenecks such as lane drops or exit/entry ramps. Another major type of congestion is induced by traffic incidents such as traffic crashes. The former can be remedied by removing physical bottlenecks through the improvement of roadway capacity, geometry etc. However, the latter usually randomly occurs due to the high level of stochasticity of incident events. Therefore, it is a challenge to capture these non-recurrent congestion hot spots due to incidents. Nevertheless, the availability of real-time traffic sensor data provides the opportunity to address this issue through the use of data-driven solutions. Thus, the main objective of this study is to develop an automated approach to quantify incident induced congestion using sensor data. A practice-ready data-driven non-recurrent congestion quantification algorithm is developed and its implementation is demonstrated through real-world case study. It has been shown that the proposed automated approach can be used to efficiently identify incident-induced congestion.

This paper was sponsored by TRB committee AHB20 Standing Committee on Freeway Operations.

01584066

16-5943

English

13p

2016

Transportation Research Board 95th Annual Meeting

Digital/other

Figures; References; Tables

Algorithms; Automatic incident detection; Case studies; Real time information; Sensors; Traffic congestion; Traffic incidents

Data and Information Technology; Highways; Operations and Traffic Management; Planning and Forecasting

Transportation Research Board Annual Meeting 2016 Paper #16-5943

TRIS, TRB, ATRI

Jan 12 2016 6:38PM