Optimizing Traffic Flow with Adaptive Control and Multi-Agent Learning

The field of traffic management is moving towards more adaptive and intelligent control systems, leveraging advancements in machine learning and multi-agent systems to optimize traffic flow and reduce congestion. Research is focusing on developing frameworks that can learn from driver behavior and adjust traffic signal timings accordingly, as well as coordinating the actions of multiple agents, such as autonomous vehicles and traffic signals, to achieve more efficient traffic flow. Noteworthy papers in this area include:

  • Route Recommendations for Traffic Management Under Learned Partial Driver Compliance, which proposes a route recommendation framework that learns partial driver compliance to optimize traffic flow.
  • Large-Scale Mixed-Traffic and Intersection Control using Multi-agent Reinforcement Learning, which presents a decentralized multi-agent reinforcement learning approach for large-scale mixed traffic control.
  • Joint Pedestrian and Vehicle Traffic Optimization in Urban Environments using Reinforcement Learning, which demonstrates a deep RL framework for adaptive control of traffic signals, jointly optimizing both pedestrian and vehicular efficiency.

Sources

Route Recommendations for Traffic Management Under Learned Partial Driver Compliance

Distributed Mixed-Integer Quadratic Programming for Mixed-Traffic Intersection Control

Large-Scale Mixed-Traffic and Intersection Control using Multi-agent Reinforcement Learning

An Efficient Approach for Cooperative Multi-Agent Learning Problems

Joint Pedestrian and Vehicle Traffic Optimization in Urban Environments using Reinforcement Learning

Safe and Efficient Coexistence of Autonomous Vehicles with Human-Driven Traffic at Signalized Intersections

Federated Hierarchical Reinforcement Learning for Adaptive Traffic Signal Control

Built with on top of