Real Time Signal Control Using Artificial Neural Network For Developing Megacities

Most of the cities in developing or in newly industrializing countries operate their road signal network by time-dependent control. A well-designed coordinated fixed time control can reveal outstanding performance if the traffic pattern remains unchanged. However, commonly occurring incidents, maintenance operations or big events change the daily traffic pattern especially in megacities. The aim of this study to develop a low-cost real-time coordinated traffic signal control system for megacities in developing or newly industrializing countries. The proposed system recognizes the traffic state with a few system detectors and chooses the optimal timing plans, which are previously optimized using network signal control optimization model TRANSYT. The proposed pattern recognition based neural network control system is evaluated based on microscopic traffic flow simulation through a component object model interface, and compared with the optimized pre-timed control strategies. The simulation results indicate the efficiency of this proposed method. With the implementation of neural network controller, the average speed increases at least 25 %, the average delay and average number of stops decrease at least 15% and 50%, respectively. Although this study based on microscopic traffic flow simulation experiments, the algorithm itself already for real world implementations.

• Availability:
  • Find a library where document is available. Order URL: http://itswc.confex.com/itswc/WC2011/webprogram/start.html
• Supplemental Notes:
  • Abstract reprinted with permission from Intelligent Transportation Society of America.
• Corporate Authors:

  ITS America

  1100 17th Street, NW, 12th Floor
  Washington, DC  United States  20036
• Authors:
  • Gundogan, Fatih
  • Fellendorf, Martin
• Conference:
  • 18th ITS World Congress
  • Location: Orlando Florida, United States
  • Date: 2011-10-16 to 2011-10-20
• Publication Date: 2011

Language

• English

Media Info

• Media Type: Digital/other
• Features: Figures; References; Tables;
• Pagination: 10p
• Monograph Title: 18th ITS World Congress, Orlando, 2011. Proceedings

Subject/Index Terms

• TRT Terms: Neural networks; Pattern recognition systems; Real time control; Traffic signal controllers
• Identifier Terms: TRANSYT (Computer program); VISSIM (Computer model)
• Uncontrolled Terms: Traffic patterns; Traffic responsive control
• Subject Areas: Highways; Operations and Traffic Management; I73: Traffic Control;

Filing Info

• Accession Number: 01449814
• Record Type: Publication
• Files: TRIS
• Created Date: Oct 22 2012 9:14AM