Citation

Wang C, Coifman BA. IEEE Trans. Intel. Transp. Syst. 2008; 9(3): 523-535.

Copyright

(Copyright © 2008, IEEE (Institute of Electrical and Electronics Engineers))

DOI

10.1109/TITS.2008.928265

PMID

unavailable

Abstract

This paper investigates the linearity of empirically observed spacing–speed relations for various drivers in the context of car-following theory and how lane-change maneuvers perturb the relation. It is shown that the impacts of lane-change maneuvers are not balanced, and the response time to an exiting vehicle is much longer than the response time to an entering vehicle. This accommodation imbalance will propagate upstream, and as discussed herein, it appears to be a source of speed and flow fluctuations (or oscillations) within a queue. This paper is motivated by Newell's simplified car-following theory, namely, that during congested periods, the trajectory of a given vehicle is essentially identical to the preceding vehicle's trajectory, except for a translation in space and time. One of the basic assumptions in Newell's presentation is that spacing and speed are linearly related. While other researchers have found macroscopic evidence supporting Newell's theory, they have also found that it fails in the presence of frequent lane-change maneuvers. This paper takes a microscopic approach, employing vehicle trajectory data. This paper provides support for Newell's assumed linear relation between spacing and speed over a large range of speeds when vehicles are not impacted by lane-change maneuvers. It also offers a possible explanation for the degraded performance of Newell's theory in the presence of heavy lane-change maneuvers. Although the focus is on Newell's simplified car-following theory, the empirical results of this paper have similar implications for many other car-following theories as well.