Two computer algorithms for automatic, freeway-incident detection under low-volume conditions were developed. The first approach uses a time-scan process. The second approach, considered to be superior to the first, operates on an event-scan principle. Computer simulations produced a family of curves that are useful in determining sensor- spacing requirements for an operational system using the event-scan algorithm. The results indicate that, when detector spacings of 1,000 ft (304.8 m) are used, all incidents on a 3-lane freeway section can be detected within 3 min for volumes up to 500 vehicles per hour. When volumes approach 1,000 vehicles per hour, 85 percent of the incidents can be detected within 3 min. Faster detection capabilities at the higher volumes would require closer detector spacings. Incident-detection operational considerations, particularly the manner in which software can be developed to recognize and compensate for vehicle-count errors produced by semitrailers and lane-change maneuvers, also are discussed.

Media Info

  • Media Type: Print
  • Features: Figures; References; Tables;
  • Pagination: pp 34-47
  • Monograph Title: Freeway entry, flow and control
  • Serial:

Subject/Index Terms

Filing Info

  • Accession Number: 00099661
  • Record Type: Publication
  • ISBN: 030902384X
  • Files: TRIS, TRB
  • Created Date: Nov 5 1975 12:00AM