Citation

Sun D, Ai Y, Sun Y, Zhao L. PLoS One 2020; 15(1): e0227609.

Affiliation

Beijing Institute of New Technology Applications, Beijing, China.

Copyright

(Copyright © 2020, Public Library of Science)

DOI

10.1371/journal.pone.0227609

PMID

31935238

Abstract

In order to quantitatively analyze the influence of different traffic conditions on highway crash risk, a method of crash risk assessment based on traffic safety state division is proposed in this paper. Firstly, the highway crash data and corresponding traffic data of upstream and downstream are extracted and processed by using the matched case-control method to exclude the influence of other factors on the model. Secondly, considering the weight of traffic volume, speed and occupancy, a multi-parameter fusion cluster method is applied to divide traffic safety state. In addition, the quantitative relationship between different traffic states and highway crash risk is analyzed by using Bayesian conditional logistic regression model. Finally, the results of case study show that different traffic safety conditions are in different crash risk levels. The highway traffic management department can improve the safety risk management level by focusing on the prevention and control of high-risk traffic safety conditions.

Language: en