Citation

Wang X, Wang X, Cai B, Liu J. Transp. Res. C Emerg. Technol. 2019; 104: 172-183.

Copyright

(Copyright © 2019, Elsevier Publishing)

DOI

10.1016/j.trc.2019.04.027

PMID

unavailable

Abstract

Excessive longitudinal deceleration and acceleration reduce driving comfort and increase safety risk. There is a dearth of research, however, on how geometric design characteristics, especially complex alignments and their adjacent segments, affect deceleration and acceleration. The Tongji University driving simulator was used to collect vehicle operation data. Deceleration and acceleration values were measured and classified as deceleration, near-cruising and acceleration. A random parameter multinomial logistic regression model was used to reflect the relationships between drivers' deceleration and acceleration and the geometric alignments of a mountainous freeway. Vehicle speed was treated as a random slope parameter to account for the heterogeneity among subjects. Model results showed deceleration and acceleration were significantly influenced by the fixed effects of maximum curvature in the following 50 m, difference between maximum and minimum slope in the preceding 100 m, maximum slope change in the following 400 m, and maximum curvature in the preceding 50 m. These findings can assist in developing safer design and evaluation standards for mountainous freeways.

Language: en

Keywords

Acceleration; Combined alignment; Deceleration; Driving simulator; Random parameter multinomial logistic model