A study on variations of car-following behavior at sag sections and the impact of introducing ACC system
Abstract
A sag section is a road segment on an expressway in which the vertical slope increases at a small but constant rate. Sag sections are the main traffic bottlenecks on Japanese expressways. Variations in individual car-following... [ view full abstract ]
A sag section is a road segment on an expressway
in which the vertical slope increases at a small but constant
rate. Sag sections are the main traffic bottlenecks on Japanese
expressways. Variations in individual car-following behavior
are generally recognized as the key cause of traffic breakdown.
This research was conducted to develop a quantitative
description of variations in individual car-following behavior
by obtaining car-following model parameter values from actual
trajectories. A simple and reliable model calibration
framework considering the effect of sag section is presented.
Analysis of the parameter values obtained showed that 48.87%
of drivers are affected by the vertical slope change. The
parameter values were found to be concentrated in certain
ranges rather than being evenly distributed. The dynamic
performance of car-following behavior was also analyzed. The
results were compared with the results of experiments
conducted with a commercial Adaptive Cruise Control (ACC)
system. Of the drivers in the analyzed trajectories, 28.83%
were found to outperform the commercial ACC system in
preventing perturbation in the leading vehicle’s motion from
being amplified in the vehicle platoon. The distribution of the
parameter values for these drivers was analyzed to identify
potential improvements to the ACC system.
Authors
-
Yan Yang
(Department of Civil Engineering, The University of Tokyo)
-
Kentaro Wada
(Institute of Industrial Science, The University of Tokyo)
-
Takashi Oguchi
(Institute of Industrial Science, The University of Tokyo)
-
Miho Iryo-asano
(Institute of Industrial Science, The University of Tokyo)
Topic Areas
Data Mining and Data Analysis , Driver Assistance Systems , Human Factors , Simulation and Modeling , Traffic Theory for ITS , Freeway Traffic Control , Traffic Flow Modelling and Control
Session
Tu-D04 » WS04 Advancing the Microscopic Traffic Simulations Towards Realistic Modelling of Driver Behaviour (15:15 - Tuesday, 15th September, Tenerife)
