Traffic Flow Simulation by Neuro-Fuzzy Approach

Traffic flow characteristics, which are the basic parameters of the traffic stream, play a crucial role in analyzing and modeling traffic flow. Simulation software is wildly utilized to predict the parameters such as average speed, travel time, capacity, etc. In essence, such programs are used to characterize the model input parameters for achieving an appropriate matching between the model and actual data. In this study, the actual traffic data is used for the evaluation of the susceptibility of adaptive neuro-fuzzy inference methodology to simulate the nonlinear mapping between flow and speed parameters. In the neuro-fuzzy model, the Sugeno type of inference system is selected for fuzzy approximate reasoning, and the model parameters are adopted by the hybrid-learning algorithm. Results denote that a neuro-fuzzy system can easily characterize such a nonlinear relationship including a certain amount of uncertainty. This study is only for the demonstration of the application of the proposed methodology; nevertheless, additional parameters of traffic flow must also be considered in such a system to better characterize real-world problems.

• Record URL:
  http://www.ijtte.com/uploads/news_files/ICTTE_Belgrade_2014-Proceedings.pdf
• 
• Availability:
  • Find a library where document is available. Order URL: http://worldcat.org/isbn/9788691615321
• Supplemental Notes:
  • Abstract republished with permission of City Net Scientific Research Center, Belgrade.
• Corporate Authors:

  City Net Scientific Research Center Ltd.

  Belgrade,   Serbia 
• Authors:
  • Seitllari, Aksel
  • 0000-0001-8476-5841
• Conference:
  • Second International Conference on Traffic and Transport Engineering (ICTTE)
  • Location: Belgrade , Serbia
  • Date: 2014-11-27 to 2014-11-28
• Publication Date: 2014

Language

• English

Media Info

• Media Type: Digital/other
• Features: References; Tables;
• Pagination: pp 97-102
• Monograph Title: Proceedings of Second International Conference on Traffic and Transport Engineering (ICTTE)

Subject/Index Terms

• TRT Terms: Fuzzy systems; Neural networks; Statistical inference; Traffic flow; Traffic simulation
• Uncontrolled Terms: Adaptive neuro-fuzzy inference systems
• Subject Areas: Highways; Planning and Forecasting;

Filing Info

• Accession Number: 01600722
• Record Type: Publication
• ISBN: 9788691615321
• Files: TRIS
• Created Date: May 31 2016 9:14AM