Citation

Alluri P, Sando T, Kitali AE, Monyo DE, Wang H. Lehman Center for Transportation Research, Florida International University. Tallahassee, FL, USA: Florida Department of Transportation, 2022.

Copyright

(Copyright 2022, Lehman Center for Transportation Research, Florida International University)

The full text is unavailable online. Please contact the publisher.

Abstract

The primary goal of this research was to develop a comprehensive approach to identify and mitigate secondary crashes on the Florida Turnpike System in real time. The specific research objectives were to: (1) identify secondary crashes; (2) identify significant factors contributing to the occurrence of secondary crashes; (3) develop an algorithm that predicts the likelihood of secondary crashes in real time; and (4) explore the potential of connected vehicle (CV) applications in mitigating secondary crashes.

This research used four types of data: (1) incident data; (2) high-resolution traffic data; (3) roadway geometric data; and (4) high-resolution rainfall data. These data were collected from January 2014 to June 2019. The study corridors were primarily selected from the Florida's Turnpike Mainline and Turnpike Extension.

A data-driven approach was developed to accurately estimate the incident impact area and identify secondary crashes within the impacted area. The Least Absolute Shrinkage and Selection Operator (LASSO) penalized logistic regression model, fitted using the bootstrap resampling approach, was used to identify factors influencing the risk of secondary crashes. Traffic flow, incident, temporal, weather, and roadway geometric attributes were considered to influence the likelihood of secondary crashes. An algorithm was developed as a proof-of-concept to predict the likelihood of secondary crashes in real time. This algorithm uses incident data from SunGuide®, traffic data from Regional Integrated Transportation Information System (RITIS), rainfall data from National Oceanic and Atmospheric Administration (NOAA), and the collected roadway geometric characteristics data. The algorithm estimates the impact area of an incident and predicts the likelihood of secondary crashes in real time.
Finally, the microsimulation approach was used to explore the potential of CV applications in mitigating secondary crashes. The following simulation scenarios were considered: blockage of the inner lane, one outer lane, and two outer lanes. A sensitivity analysis was performed by considering varying market penetration rates (MPRs) of CVs. The significance of CV applications in mitigating secondary crashes was assessed using traffic conflicts derived from microsimulation models.