Urban Travel Time Estimation by Incorporating System Recognition into K-Nearest-Neighbor Method

The importance of providing accurate and timely traffic information to the traveling public is widely recognized, and has become an increasingly important function for the traffic management authorities. This paper proposes an urban travel time estimation model derived by incorporating System Recognition into K Nearest Neighbor models. Since the K Nearest Neighbor models ignore causal relationship between travel time and traffic demand, this paper attempts to use System Recognition to explore the inherent nature of urban travel time. The proposed model is evaluated by the data collected from single loop inductive detectors installed on Kruithuisweg, the Netherlands, by Regiolab Delft. Results generated by this hybrid model and their accuracy are compared with two existing models. The experimental results reveal that the proposed model gives promising performance.

Language

  • English

Media Info

  • Media Type: CD-ROM
  • Features: Figures; References; Tables;
  • Pagination: 19p
  • Monograph Title: TRB 86th Annual Meeting Compendium of Papers CD-ROM

Subject/Index Terms

Filing Info

  • Accession Number: 01043561
  • Record Type: Publication
  • Report/Paper Numbers: 07-0648
  • Files: TRIS, TRB
  • Created Date: Feb 8 2007 5:14PM