Traffic Light Control

Traffic light control aims to optimize traffic flow and reduce congestion by coordinating signal timings across intersections. Current research heavily utilizes deep reinforcement learning (DRL), employing architectures like transformers and graph attention networks to learn optimal control policies from real-time traffic data, often incorporating vehicle-to-infrastructure communication and even UAV-based monitoring for enhanced situational awareness. This focus on DRL, coupled with efforts to improve model transferability and data efficiency, seeks to create more adaptable and scalable solutions for managing increasingly complex urban traffic networks. The resulting improvements in traffic efficiency have significant implications for reducing travel times, fuel consumption, and environmental impact.

Papers

April 18, 2024

X-Light: Cross-City Traffic Signal Control Using Transformer on Transformer as Meta Multi-Agent Reinforcement Learner
Haoyuan Jiang, Ziyue Li, Hua Wei, Xuantang Xiong, Jingqing Ruan, Jiaming Lu, Hangyu Mao, Rui Zhao
Transformer Based Multi Agent Reinforcement Learning Traffic Signal Control Meta Level Traffic Light Decision Path Traffic Light Control

September 26, 2023

Semantic Map Learning of Traffic Light to Lane Assignment based on Motion Data
Thomas Monninger, Andreas Weber, Steffen Staab
Autonomous Vehicle Semantic Map Motion Forecasting Motion Pattern Traffic Light Traffic Light Control

September 18, 2023

Deep Reinforcement Learning for the Joint Control of Traffic Light Signaling and Vehicle Speed Advice
Johannes V. S. Busch, Robert Voelckner, Peter Sossalla, Christian L. Vielhaus, Roberto Calandra, Frank H. P. Fitzek
Deep Reinforcement Learning Traffic Management System Joint Control Traffic Light Control Speed Advisory Intelligent Traffic Light

September 10, 2023

AVARS -- Alleviating Unexpected Urban Road Traffic Congestion using UAVs
Jiaying Guo, Michael R. Jones, Soufiene Djahel, Shen Wang
Traffic Data Urban Traffic Traffic Light Control

August 28, 2023

Traffic Light Control with Reinforcement Learning
Taoyu Pan
Reinforcement Learning Deep Q Network Deep Q Learning Traffic Light Control

February 23, 2023

Inequity aversion reduces travel time in the traffic light control problem
Mersad Hassanjani, Farinaz Alamiyan-Harandi, Pouria Ramazi
Deep Reinforcement Learning Travel Time Traffic Light Control Inequity Aversion

February 4, 2023

Deep Reinforcement Learning for Traffic Light Control in Intelligent Transportation Systems
Xiao-Yang Liu, Ming Zhu, Sem Borst, Anwar Walid
Deep Reinforcement Learning Deep Q Network Road Network Intelligent Transportation System Traffic Light Control Intelligent Traffic Light

August 1, 2022

Model-based graph reinforcement learning for inductive traffic signal control
François-Xavier Devailly, Denis Larocque, Laurent Charlin
Task Transferability Graph Reinforcement Learning Adaptive Traffic Signal Control Road Topology Dynamic Traffic Traffic Light Control

November 15, 2021

ModelLight: Model-Based Meta-Reinforcement Learning for Traffic Signal Control
Xingshuai Huang, Di Wu, Michael Jenkin, Benoit Boulet
Reinforcement Learning Meta Reinforcement Learning Traffic Signal Control Traffic Light Control Intelligent Traffic Light