Secondary Crash Identification Using Linear Referencing System

Authors: Dongxi Zheng, Madhav Chitturri, Andrea Bill, and David Noyce

Abstract

[Place holder]

Introduction

Secondary crashes are a major type of highway incidents. In recent researches, a secondary crash is commonly recognized as a crash occurred within a highway traffic queue caused by a previous roadway incident (1-4). [show some stats to proof secondary crashes as a major crash type]

Secondary crash rate is a good but expensive indicator to the performance of traffic incident management. Studies have shown that shorter incident clearance time resulted in lower risk of secondary crashes (5-8). However, secondary crash rates are hard to measure because of the difficulty in identifying secondary crashes. The main idea of secondary crash identification is to measure the spatial-temporal distances between any two incidents, so when two incidents were closer than certain thresholds (static or dynamic), the later incident might be a secondary crash. Among the literatures, different methods have been proposed focusing on improving the accuracy of secondary crash identification by choosing more appropriate spatial-temporal thresholds (1-5, 9, 10). Commenly used data sources for such purposes were police incident reports, in-database crash records, highway detector data, video records, etc. As a results, the cost and the effectiveness of these methods heavily rely on the desired data quantify and the actual data quality, respectively. More unfortunately, when the study scope expends (e.g., from a major arterial to an arterial network), the complexity and the data demand of these methods can explode exponentially. This paper shows a relatively fundamental improvement to the effeciency of secondary crash identification, allowing different previous methods to be built upon and reserve their advantages on accurary.

The essential idea is to utilize a linear referencing system to speed up the calculation of spatial distances. [Explain what a general linear referencing system is and what is the case in seconday crash identification]. Based on the linear referencing system, a spatial search algorithm is proposed to significantly reduces the total amount of calculation needed for a large network with many crash points. A program is implemented according to the algorithm. Previous static-threshold methods can be readily run on the program to get desired output. For dynamic-threshold methods, only several simple extra steps are needed before and after running the program.

The following of this paper is organized in five sections. ...

Literature Review

Methodology

[rephrase of the algorithm]

[how dynamic threshold can be inplanted]

Data Collection

[WisTransPortal data retrieval]

Results and Analyses

Verification results

[Madhav's test case]

[error correction?]

Large network performance measure

[In compare with ArcGIS?]

Conclusion and Recommendations

References

1. Moore, I., James E., G. Giuliano, and S. Cho. Secondary Accident Rates on Los Angeles Freeways. In Journal of Transportation Engineering, Vol. 130, No. 3, American Society of Civil Engineers, Reston, Virginia, 2004, pp. 280-285.
2. Sun, C., and V. Chilukuri. Use of Dynamic Incident Progression Curve for Classifying Secondary Accidents. In TRB 85th Annual Meeting Compendium of Papers CD-ROM, Transportation Research Board of the National Academies, Washington, D.C., 2006, pp. 1-12.
3. Zhan, C., A. Gan, and M. Hadi. Identifying Secondary Crashes and Their Contributing Factors. In Transportation Research Record: Journal of the Transportation Research Board, Vol. 2102, Transportation Research Board of the National Academies, Washington, D.C., 2009, pp. 68-75.
4. Chou, C., and E. Miller-Hooks. Simulation-Based Secondary Incident Filtering Method. In Journal of Transportation Engineering, Vol. 136, No. 8, American Society of Civil Engineers, Reston, Virginia, 2010, pp. 746-754.
5. Karlaftis, M., S. Latoski, N. Richards, and K. Sinha. ITS Impacts on Safety and Traffic Management: An Investigation of Secondary Crash Causes. In ITS Journal, Vol. 5, No. 1, Gordon and Breach Science Publishers, Berks, England, 1999, pp. 39-52.
6. Khattak, A. J., X. Wang, and H. Zhang. Are Incident Durations and Secondary Incidents Interdependent? In Transportation Research Record: Journal of the Transportation Research Board, Vol. 2099, Transportation Research Board of the National Academies, Washington, D.C., 2009, pp. 39-49.
7. Zhang, H., and A. J. Khattak. Analysis of Cascading Incident Event Durations on Urban Freeways. In Transportation Research Record: Journal of the Transportation Research Board, Vol. 2178, Transportation Research Board of the National Academies, Washington, D.C., 2010, pp. 30-39.
8. Vlahogianni, E. I., M. G. Karlaftis, J. C. Golias, and B. M. Halkias. Freeway Operations, Spatiotemporal-Incident Characteristics, and Secondary-Crash Occurrence. In Transportation Research Record: Journal of the Transportation Research Board, Vol. 2178, Transportation Research Board of the National Academies, Washington, D.C., 2010, pp. 1-9.
9. Raub, R. A. Occurrence of Secondary Crashes on Urban Arterial Roadways. In Transportation Research Record: Journal of the Transportation Research Board, No. 1581, Transportation Research Board of the National Academies, Washington, D.C., 1997, pp. 53-58.
10. Hirunyanitiwattana, W., and S. P. Mattingly. Identifying Secondary Crash Characteristics for California Highway System. In TRB 85th Annual Meeting Compendium of Papers CD-ROM, Transportation Research Board of the National Academies, Washington, D.C., 2006, pp. 1-18.