DISTRIBUTED APPROACH TO OPTIMIZE CONTROL OF STREET TRAFFIC SIGNALS

The paper discusses an economically important domain of computer applications, namely, improving street traffic by using collaborative, distributed, learning expert systems that control street traffic signals. Every processor runs an identical expert system and communicates directly with the four adjacent processors. Messages can reach all indefinitely distant processors, modulated ty the needs of intervening ones. The rule base of the expert system has a natural segmentation, corresponding to different prevailing traffic patterns and the respective control strategies. Details of the system are described, including multidimensional learing programs, measures of effectiveness, and the generation of traffic scenarios. A running prototype simulation program has proved the feasibility and utility of the approach.

• Availability:
  • Find a library where document is available. Order URL: http://worldcat.org/oclc/8674831
• Corporate Authors:

  American Society of Civil Engineers

  345 East 47th Street
  New York, NY  United States  10017-2398
• Authors:
  • Findler, N V
  • Stapp, J
• Publication Date: 1992-1

Media Info

• Features: Appendices; Figures; References; Tables;
• Pagination: p. 99-110
• Serial:
  • Journal of Transportation Engineering
  • Volume: 118
  • Issue Number: 1
  • Publisher: American Society of Civil Engineers
  • ISSN: 0733-947X
  • Serial URL: https://ascelibrary.org/journal/jtepbs

Subject/Index Terms

• TRT Terms: Computer programs; Expert systems; Measures of effectiveness; Optimization; Traffic signal controllers
• Subject Areas: Highways; Operations and Traffic Management; I73: Traffic Control;

Filing Info

• Accession Number: 00621884
• Record Type: Publication
• Files: TRIS
• Created Date: May 31 1992 12:00AM