Integration of a Real-time Traffic State Estimation and a Decentralized Game-Theoretic Traffic Signal Controller

Details

• Creators:
  Shafik, Amr ; Rakha, Hesham A.
• Corporate Creators:
  Virginia Tech Transportation Institute ; Morgan State University. Sustainable Mobility and Accessibility Regional Transportation Equity Research Center (SMARTER) Region 3 University Transportation Center (UTC)
• Corporate Contributors:
  United States. Department of Transportation. University Transportation Centers (UTC) Program ; United States. Department of Transportation. Office of the Assistant Secretary for Research and Technology
• Subject/TRT Terms:
  Game Theory Kalman Filtering Optimization Probe Vehicles Traffic Data Traffic Flow Traffic Signal Controllers Traffic Signal Phases
• Publication/ Report Number:
  SM28
• Resource Type:
  Tech Report
• Geographical Coverage:
  United States
• Edition:
  Final, September 2024 - August 2025
• Corporate Publisher:
  Morgan State University. Sustainable Mobility and Accessibility Regional Transportation Equity Research Center (SMARTER) Region 3 University Transportation Center (UTC)
• Abstract:
  The report presents an integrated approach to adaptive traffic signal control by combining advanced traffic state estimation and adaptive optimization techniques. A two-stage adaptive Kalman filter (KF) algorithm is proposed to estimate turning movements and refine upstream density and queue sizes using probe vehicle data and detector measurements. Tested on data from Orlando, the method improves estimation accuracy significantly—reducing turning movement estimation error by up to 50% and queue size errors by 32.8%. The estimation algorithm is then integrated with a Decentralized Nash-Bargaining (DNB) traffic signal controller to optimize signal timings, even with incomplete vehicle trajectory data. Evaluations in Toronto and Orlando show that the DNB-KF system outperforms traditional traffic control strategies under limited data conditions. Additionally, the study explores two more optimization methods: the LDR cycle length adjustment and an enhanced DNB controller. Both methods use real-time traffic states to reduce delays and fuel consumption, achieving up to 37.7% delay reduction and 7.4% fuel savings. Overall, this research demonstrates that combining accurate traffic state estimation with adaptive, game-theoretic signal control leads to more efficient and sustainable urban traffic management, even with partial traffic data.
• Format:
  PDF
• Funding:
  69A3552348303
• Collection(s):
  University Transportation Centers
• Main Document Checksum:
  urn:sha-512:31a1c162193f29ba2588e72312ca7dab9f7cc2eea867245d507fc75f46ef89c794e744def637d3ea88280584809d5f3b532d41b1db8ae3d16249520a18d6da51
• Download URL:
  https://rosap.ntl.bts.gov/view/dot/86632/dot_86632_DS1.pdf
• File Type:
  [PDF - 2.53 MB ]