Citation

Nagahama A, Yanagisawa D, Nishinari K. Transportmetrica B: Transp. Dyn. 2020; 8(1): 22-48.

Copyright

(Copyright © 2020, Hong Kong Society for Transportation Studies, Publisher Informa - Taylor and Francis Group)

DOI

10.1080/21680566.2019.1710002

PMID

unavailable

Abstract

The dynamics of mixed traffic were analysed by examining the driving features of various vehicles based on their car-following behaviours in a test-driving circuit. By applying decision tree analysis to the observed data, we successfully compared the acceleration, velocity difference and distance gap of different vehicles during driving trials and quantitatively extracted the features of these physical values from the respective driving phases. Our investigation resulted in a short-distance gap for motorcycles in the steady phase, while a difference in the distance gap between cars and trucks appeared in the early deceleration phase. We also obtained the characteristic behaviours of these vehicle types in different driving phases. Our findings indicate that the car-following characteristics of various types of vehicles are affected by the recognition of vehicle agility. Furthermore, we propose the necessary model parameters for replicating these features in different driving phases.

Language: en

Keywords

car-following characteristics; decision tree; Mixed traffic; shapelet