Trajectory-based traffic management inside an autonomous vehicle zone

Gongyuan Lu, Yu(Marco) Nie*, Xiaobo Liu, Denghui Li

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

16 Scopus citations

Abstract

This paper studies a trajectory-based traffic management (TTM) problem for the purpose of managing traffic in a road facility reserved exclusively for autonomous vehicles (AV). The base TTM model aims to find optimal trajectories for multiple AVs while resolving inter-vehicle conflicts in the most generic way. The model is formulated as a mixed integer program (MIP) that can be solved using off-the-shelf solvers. To improve computational efficiency, a specialized algorithm based on the rolling horizon approach is also developed. We then show that the base TTM model can be easily extended to first accommodate scheduling decisions (the TTMS model) and to further impose equity constraints (the TTMSE model). For the simplest network and homogeneous users, solutions to TTMS and TTMSE are similar, respectively, to system optimal (SO) and user equilibrium (UE) solutions of Vickrey's bottleneck model. Numerical experiments highlight TTM's ability to simultaneously generate optimal trajectories for multiple vehicles. They also show that, while solving TTM exactly is computationally demanding, obtaining good approximate solutions can be accomplished efficiently by the rolling horizon algorithm.

Original languageEnglish (US)
Pages (from-to)76-98
Number of pages23
JournalTransportation Research Part B: Methodological
Volume120
DOIs
StatePublished - Feb 2019

Keywords

  • Autonomous vehicles
  • Mixed integer program
  • Rolling horizon algorithm
  • Trajectory-based traffic management

ASJC Scopus subject areas

  • Civil and Structural Engineering
  • Transportation

Fingerprint Dive into the research topics of 'Trajectory-based traffic management inside an autonomous vehicle zone'. Together they form a unique fingerprint.

Cite this