REAL TIME AREA WIDE LINK FLOW INFERENCE FROM LINK DETECTION

This paper presents a network flow based model which can infer link flows from detected flows in quasi real-time. Applying Operational Research methods, the model considers flow continuity. A Southern California city has been surveyed to provide the test bed data, which demonstrates the reliability of the model. Such real time flow data can be applied to real time flow monitoring and incident detection. Observed flow data provides the basis for the simulation of detectors which are then supplied as model input. Detector failure is simulated through the removal of pseudo-detectorized links, and the model is shown to estimate unknown link flows and representations of so called failed detectors alike.

Language

  • English

Media Info

  • Pagination: p. 281-285

Subject/Index Terms

Filing Info

  • Accession Number: 00712625
  • Record Type: Publication
  • Files: TRIS
  • Created Date: Oct 24 1995 12:00AM