Online Deployment of Dynamic Traffic Assignment: Evaluation and Lessons

Route guidance has the potential to reduce delays and improve travel time reliability. Guidance based on short-term predictions of network state is expected to perform better than information derived from instantaneous traffic conditions. State-of-the-art Dynamic Traffic Assignment (DTA) models have been developed to generate consistent, anticipatory route guidance. This paper presents an on-line evaluation of DynaMIT -- a real-time DTA system. DynaMIT has been integrated with real-time data from the traffic management center in Los Angeles, California. The system operates in a rolling horizon to estimate and predict network state in real-time, and generate route guidance. Critical components of the deployment are outlined, and an on-line system evaluation with real data is presented. The results indicate that DynaMIT is able to replicate and predict the conditions of a real-world network. Several key lessons were learned, suggesting directions for the future: sensor status, traffic signals, and incidents are critical information that need to be collected in a timely fashion to ensure accuracy; continuous evaluation and on-line calibration are necessary for system enhancement; suitable resources should be available for system sustainability over extended periods of time; additional tests should be conducted to evaluate the impact of route guidance.

Language

  • English

Media Info

  • Media Type: DVD
  • Features: Figures; Maps; References; Tables;
  • Pagination: 23p
  • Monograph Title: TRB 87th Annual Meeting Compendium of Papers DVD

Subject/Index Terms

Filing Info

  • Accession Number: 01099282
  • Record Type: Publication
  • Report/Paper Numbers: 08-2955
  • Files: TRIS, TRB
  • Created Date: May 21 2008 7:05AM