Evaluation of a model predictive control framework for motorway traffic involving conventional and automated vehicles

Georgia Perraki*, Claudio Roncoli, Ioannis Papamichail, Markos Papageorgiou

*Corresponding author for this work

Research output: Contribution to journalArticleScientificpeer-review

16 Citations (Scopus)
40 Downloads (Pure)

Abstract

A Model Predictive Control (MPC) strategy for motorway traffic management, which takes into account both conventional control measures and control actions executed by vehicles equipped with Vehicle Automation and Communication Systems (VACS), is presented and evaluated using microscopic traffic simulation. A stretch of the motorway A20, which connects Rotterdam to Gouda in the Netherlands, is taken as a realistic test bed. In order to ensure the reliability of the application results, extensive speed and flow measurements, collected from the field, are used to calibrate the site's microscopic traffic simulation model. The efficiency of the MPC framework, applied to this real sizable and complex network under realistic traffic conditions, is examined for different traffic conditions and different penetration rates of equipped vehicles. The adequacy of the control application when only VACS equipped vehicles are used as actuators, is also considered, and the related findings underline the significance of conventional control measures during a transition period or in case of increased future demand.

Original languageEnglish
Pages (from-to)456-471
Number of pages16
JournalTransportation Research Part C: Emerging Technologies
Volume92
DOIs
Publication statusPublished - 1 Jul 2018
MoE publication typeA1 Journal article-refereed

Keywords

  • Aimsun
  • Connected and autonomous vehicles
  • Microscopic traffic simulation
  • Model predictive control
  • Motorway traffic control

Fingerprint Dive into the research topics of 'Evaluation of a model predictive control framework for motorway traffic involving conventional and automated vehicles'. Together they form a unique fingerprint.

Cite this