Citation

Zhang M, Ma Y, Ma Y. China Saf. Sci. J. 2023; 33(8): 190-197.

Copyright

(Copyright © 2023, China Occupational Safety and Health Association, Publisher Gai Xue bao)

DOI

10.16265/j.cnki.issn1003-3033.2023.08.1084

PMID

unavailable

Abstract

In order to improve the crossing efficiency at unsignalized intersections and the safety of pedestrian crossings, K⁃means++ algorithm was used to cluster the conflict indicators to determine the severity of vehicle⁃pedestrian conflict. Severe multicollinearity and insignificant influencing factors were eliminated by multi⁃collinearity test and Pearson correlation analysis, and then the influencing factors were ranked. Multiple ordered Logistic regression algorithm was used to compare and analyze the specific influence of each significant independent factor on the severity of vehicle⁃pedestrian conflicts. The results show that the severity of vehicle⁃pedestrian conflicts can be divided into severe, general and minor conflicts by taking time to collision, gap time and deceleration to safety time as conflict indicators. The influence of gender, distracted behavior, waiting time, average crossing speed and whether accompanied by others on the severity of conflict decreased successively among the pedestrian characteristics. Headway, driving intention and vehicle type have significant effects on the severity of conflict among the vehicle characteristics. © 2023 China Safety Science Journal. All rights reserved.

Language: zh