eduzhai > Applied Sciences > Engineering >

Adaptive Coordination Offsets for Signalized Arterial Intersections using Deep Reinforcement Learning

  • king
  • (0) Download
  • 20210506
  • Save

... pages left unread,continue reading

Document pages: 10 pages

Abstract: One of the most critical components of an urban transportation system is thecoordination of intersections in arterial networks. With the advent ofdata-driven approaches for traffic control systems, deep reinforcement learning(RL) has gained significant traction in traffic control research. Proposed deepRL solutions to traffic control are designed to directly modify either phaseorder or timings; such approaches can lead to unfair situations -- bypassinglow volume links for several cycles -- in the name of optimizing traffic flow.To address the issues and feasibility of the present approach, we propose adeep RL framework that dynamically adjusts the offsets based on traffic statesand preserves the planned phase timings and order derived from model-basedmethods. This framework allows us to improve arterial coordination whilepreserving the notion of fairness for competing streams of traffic in anintersection. Using a validated and calibrated traffic model, we trained thepolicy of a deep RL agent that aims to reduce travel delays in the network. Weevaluated the resulting policy by comparing its performance against the phaseoffsets obtained by a state-of-the-practice baseline, SYNCHRO. The resultingpolicy dynamically readjusts phase offsets in response to changes in trafficdemand. Simulation results show that the proposed deep RL agent outperformedSYNCHRO on average, effectively reducing delay time by 13.21 in the AMScenario, 2.42 in the noon scenario, and 6.2 in the PM scenario. Finally, wealso show the robustness of our agent to extreme traffic conditions, such asdemand surges and localized traffic incidents.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×