Real-Time Data-Based Decision Support System for Arterial Traffic Management

This study developed and evaluated a Traffic Responsive Plan Selection (TRPS) strategy based on supervised and unsupervised machine learning combined with signal timing optimization to address the issues with traditional TRPS. The strategy fills an important gap in providing a proactive traffic control that makes use of Automated Traffic Signal Performance Measures (ATSPM) measures-based data that are becoming available sources, including high-resolution controller data. This study also explored a methodology and evaluated multiple algorithms for the short-term prediction of the traffic state for the next half an hour. The results revealed that the Artificial Neural Network (ANN) algorithm produced the best results in terms of accuracy and areas under the curve. Thus, this study used the ANN prediction model in the remaining task to implement and evaluate the prediction to support the activation of the signal timing plans in real-time operations. The study then assessed the performance of the predictive TRPS based on clustering and prediction by evaluating five different scenarios of signal timing plan selection. The results of the project case study showed that the predictive TRPS method can decrease the travel time by 7 percent compared to existing traffic signals, 4% compared to optimizing for a fixed signal timing plan based on a signature day for the whole database, and 17% compared to optimizing signal timing for a random day in the data. This shows that the TRPS based on traffic pattern identification and prediction has the potential to improve traffic performance compared to other assessed optimization scenarios.

• Record URL:
  • https://stride.ce.ufl.edu/wp-content/uploads/sites/153/2023/12/STRIDE-Project-J2-Hadi_Elefteriadou-Formatted.pdf
• Record URL:
  • https://rosap.ntl.bts.gov/view/dot/72835
• Supplemental Notes:
  • This document was sponsored by the U.S. Department of Transportation, University Transportation Centers Program.
• Corporate Authors:

  Florida International University

  Civil and Environmental Engineering
  10555 W. Flagler Street, EC 3680
  Miami, FL  United States  33174

  University of Florida, Gainesville

  Department of Civil and Coastal Engineering
  Gainesville, FL  United States  32611-6580

  Southeastern Transportation Research, Innovation, Development and Education Center (STRIDE)

  University of Florida
  365 Weil Hall
  Gainesville, FL  United States  32611

  Office of the Assistant Secretary for Research and Technology

  University Transportation Centers Program
  Department of Transportation
  Washington, DC  United States  20590

  Alabama Department of Transportation

  1409 Coliseum Boulevard
  Montgomery, AL  United States  36130-3050
• Authors:
  • Hadi, Mohammed
  • 0000-0003-2233-8283
  • Elefteriadou, Lily
  • 0000-0003-4045-3365
  • Mata, Hector
  • Amini, Ehsan
• Publication Date: 2023-4-20

Language

• English

Media Info

• Media Type: Digital/other
• Edition: Final Report
• Features: Figures; Photos; References; Tables;
• Pagination: 57p

Subject/Index Terms

• TRT Terms: Decision support systems; Highway traffic control; Machine learning; Neural networks; Optimization; Traffic forecasting; Traffic signal timing
• Subject Areas: Highways; Operations and Traffic Management; Planning and Forecasting;

Filing Info

• Accession Number: 01902937
• Record Type: Publication
• Report/Paper Numbers: STRIDE Project J2
• Contract Numbers: 69A3551747104
• Files: UTC, NTL, TRIS, ATRI, USDOT, STATEDOT
• Created Date: Dec 21 2023 4:35PM