01660326

Component

Highway incidents are responsible for 50% - 70% of total highway delay time nationwide. The fundamental method to reduce the delay time is to detect and verify incidents early, and remove the incident quickly. Many methodologies have been studied and implemented in past 3 decades, however, due to the large spacing between the detection stations – normally 1/3 – 1/4 miles apart, all the methods have had limited capability to detect and verify incidents quickly. The emergence of Connected Vehicle (CV) technology provides a new data source, which would significantly improve the incident detecting and verifying methods. This paper documents the findings of a research project which created a digital simulation testing environment to test the usage of CV data for quick incident detection and verification. A great effort has been made to generate a large amount of incident scenarios data, and the data were used for quick incident detection. A set of new MOEs and algorithms were developed for the study. A comparison was made to conventional incident detection algorithm such as California Algorithm #7. The research outcome indicates that using CV data could significantly reduce the detection time, from minutes to just seconds. Some future studies are suggested in the paper. In the near future, with the increase of CV market penetration, this methodology would be very promising in quick incident detection with low false alarm rate.

This paper was sponsored by TRB committee ABJ35 Standing Committee on Highway Traffic Monitoring.

18-05089

English

5

2018

Transportation Research Board 97th Annual Meeting

Digital/other

References; Tables

Connected vehicles; Data analysis; Incident detection; Mobile communication systems; Traffic incidents

Data and Information Technology; Highways; Operations and Traffic Management

Transportation Research Board Annual Meeting 2018 Paper #18-05089

TRIS, TRB, ATRI

Jan 8 2018 11:16AM