Freeway Travel Time Prediction with Dynamic Neural Networks

A number of approaches including Neural Networks, time series, and traffic simulation modeling have been proposed for short-term travel time prediction. These approaches have achieved varying degrees of success in their abilities to predict travel time. Dynamic Neural Networks comprise a class of neural networks that is particularly suitable for predicting variables like travel time, but has not been adequately investigated for this purpose. This study compares the travel time prediction performance of three different Dynamic Neural Networks topologies with different memory setups. The results show that one Dynamic Neural Networks topology (the time-delay neural networks) out-performed the other two Dynamic Neural Networks topologies for the investigated prediction problem. This topology also performed slightly better than the simple multilayer perceptron neural networks that have been used in a number of previous studies for travel time prediction.

Language

  • English

Media Info

  • Media Type: DVD
  • Features: References; Tables;
  • Pagination: 15p
  • Monograph Title: TRB 89th Annual Meeting Compendium of Papers DVD

Subject/Index Terms

Filing Info

  • Accession Number: 01152847
  • Record Type: Publication
  • Report/Paper Numbers: 10-3105
  • Files: TRIS, TRB
  • Created Date: Mar 19 2010 9:58AM