A Novel Approach of Dynamic Time Warping for Short-Term Traffic Congestion Prediction

This paper proposes new automated algorithm that can detect traffic patterns leading to congestion in microscopic traffic variables. The approach algorithm, namely Dynamic Time Warping, has many abilities to classify complex time series but requires much smaller training dataset than traditional Artificial Intelligent (AI) algorithms. The performance of the proposed algorithm is assessed using Bayesian algorithm with microscopic traffic simulation environment and real-world data. The result shows that the proposed algorithm has performance comparable to Bayesian algorithm using the standard deviation of speed as algorithm figures to predict traffic congestion patterns. The proposed algorithm also performs very well even when applied to another site in real-world scenarios.

• Corporate Authors:

  Transportation Research Board

  500 Fifth Street, NW
  Washington, DC  United States  20001
• Authors:
  • Hiri-O-Tappa, Kittipong
  • Pan-Ngum, Setha
  • Narupiti, Sorawit
  • Pattara-Atikom, Wasan
• Conference:
  • Transportation Research Board 90th Annual Meeting
  • Location: Washington DC, United States
  • Date: 2011-1-23 to 2011-1-27
• Date: 2011

Language

• English

Media Info

• Media Type: DVD
• Features: Figures; Photos; References; Tables;
• Pagination: 15p
• Monograph Title: TRB 90th Annual Meeting Compendium of Papers DVD

Subject/Index Terms

• TRT Terms: Algorithms; Artificial intelligence; Time series analysis; Traffic congestion; Traffic simulation
• Uncontrolled Terms: Bayesian analysis; Real world data; Traffic patterns
• Subject Areas: Highways; Operations and Traffic Management; Planning and Forecasting; I72: Traffic and Transport Planning;

Filing Info

• Accession Number: 01340064
• Record Type: Publication
• Report/Paper Numbers: 11-3402
• Files: TRIS, TRB
• Created Date: May 18 2011 10:51AM