Nonlinear Kalman Filtering Algorithms for On-Line Calibration of Dynamic Traffic Assignment Models

This paper provides an online calibration approach to dynamic traffic assignment (DTA) models that works through jointly estimating demand and supply parameters. Empirical testing has been conducted on the presented design and thereupon validated the design. When calibration is formulated as a nonlinear state-space model, Kalman filtering cannot be used and thus nonlinear extensions must be taken into account. The three nonlinear extensions discussed herein are first the extended Kalman filter (EKF), a limiting EKF (LimEKF), and the unscented Kalman filter. With a test site located on a freeway network in Southampton in the UK, researchers provide a solution algorithm that is applied to an on-line calibration for the DTA model. It was found that the LimEKF consistently performed as well as the best EKF algorithm while being orders of magnitude less than those algorithms.

• Availability:
  • Find a library where document is available. Order URL: http://worldcat.org/oclc/41297384
• Authors:
  • Antoniou, Constantinos
  • Ben-Akiva, Moshe
  • Koutsopoulos, Haris N
• Publication Date: 2007-12

Language

• English

Media Info

• Media Type: Print
• Features: Figures; Photos; References; Tables;
• Pagination: pp 661-670
• Serial:
  • IEEE Transactions on Intelligent Transportation Systems
  • Volume: 8
  • Issue Number: 4
  • Publisher: Institute of Electrical and Electronics Engineers (IEEE)
  • ISSN: 1524-9050
  • Serial URL: http://ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber=6979

Subject/Index Terms

• TRT Terms: Algorithms; Calibration; Dynamic traffic assignment; Kalman filtering; Mathematical models; Optimization
• Subject Areas: Highways; Operations and Traffic Management; I71: Traffic Theory;

Filing Info

• Accession Number: 01087699
• Record Type: Publication
• Source Agency: UC Berkeley Transportation Library
• Files: BTRIS, TRIS
• Created Date: Jan 30 2008 7:34AM