Analysis of Interdriver Heterogeneity Based on Trajectory Data with K-means Clustering Method

DOI

10.2991/icsnce-16.2016.12How to use a DOI?

Keywords

Driver heterogeneity; K-means clustering; Car-following; NGSIM

Abstract

This paper presents a methodology to study the interdriver heterogeneity by using vehicle trajectory data. Different from the existing studies subjectively dividing the drivers into two or three types, this paper explores a K-means clustering methodology to classify the drivers based on real traffic data. So the classification would be more reasonable. In terms of the vehicle trajectory data extracted from the Next Generation Simulation (NGSIM) project, such microscopic variables as velocity, acceleration, spacing (space headway) and headway (time headway) are selected to represent the heterogeneity among drivers. The findings suggest that headway is the best variable to describe drivers' heterogeneity, and spacing is the second best. Additionally, according to the two selected variables, the drivers are divided into three types: stable driver, timid driver and aggressive driver.

Copyright

© 2016, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Tailang Zhu
AU  - Dongfan Xie
PY  - 2016/07
DA  - 2016/07
TI  - Analysis of Interdriver Heterogeneity Based on Trajectory Data with K-means Clustering Method
BT  - Proceedings of the 2016 International Conference on Sensor Network and Computer Engineering
PB  - Atlantis Press
SP  - 55
EP  - 61
SN  - 2352-5401
UR  - https://doi.org/10.2991/icsnce-16.2016.12
DO  - 10.2991/icsnce-16.2016.12
ID  - Zhu2016/07
ER  -