Citation

Haque N, Hadiuzzaman M, Rahman F, Siam MRK. J. Transp. Saf. Secur. 2020; 12(10): 1312-1337.

Copyright

(Copyright © 2020, Southeastern Transportation Center, and Beijing Jiaotong University, Publisher Informa - Taylor and Francis Group)

DOI

10.1080/19439962.2019.1597001

PMID

unavailable

Abstract

This paper endeavours to develop a model that estimates head-on crash probability from classified vehicle trajectory. The model formulation considered (1) drivers' overtaking decision (OD) and (2) time-to-collision (TTC) on two-lane undivided highways. Drivers' overtaking decision was modelled using nonlinear random parameter multivariate binary logistic regression. It considered traffic and drivers' characteristics (i.e., aggressiveness) variables. In contrast, TTC was determined using a new formulation adding vehicles' dynamic acceleration in consideration. Incorporation of overtaking importance factor (OIF) and crash frequency parameter (CFP) enabled the estimation of crash probability combining OD and TTC. Background subtraction technique along with Kalman filter was used to obtain vehicle trajectories from real-time video. Variable inputs required for calibrating the OD model were generated by constructing adjacency matrices among the vehicles. Exploiting these inputs, Metropolis-Hastings algorithm from Markov chain Monte Carlo (MCMC) method was applied to obtain calibrated parameters of the OD model for different types of vehicle. Calibration result showed that subject vehicle speed and the subject-opposing vehicle play an important role in influencing the overtaking decision. Moreover, the study also found that bus has maximum crash probability. Finally, the nomographs established in this paper ensures easy determination of the crash probability for practical applications.

Language: en

Keywords

head-on crash; nomograph; Overtaking model; two-lane highway; vehicle classification; vehicle detection; vehicle trajectory