Deep Learning-Based Decentralized Traffic Signal Control and Dynamic Perimeter Adjustment for Traffic Optimization

The Transport and Traffic Engineering Laboratory (LICIT-ECO7) is a Joint Research Unit which is placed under the supervision of University Gustave Eiffel and ENTPE (Post-graduate School of Transport and Civil Engineering).

LICIT-ECO7 is also part of the University de Lyon and Lyon Urban School.

The laboratory is located on both the ENTPE Campus and the University Gustave Eiffel Bron Campus in Lyon.

LICIT-ECO7 is part of the JRT GEST (Joint Research Team on Energy Management and Storage for Transport).

LICIT-ECO7’s main research theme is the Dynamic Modelling, Monitoring and Control of Mobility Networks as well as New Vehicles and New Components Usage, Ageing and Control.

To avoid traffic conflicts at intersections, traffic signal control allocates green times to various vehicle movements at signalized intersections. However, when poorly optimized, these strategies can lead to severe congestion, increased energy consumption, and higher pollution levels. Optimizing traffic signal strategies across a network is particularly challenging in dense urban areas, where it is crucial to address disruptions caused by congestion, accidents, or equipment failures.

In such urban grid networks, local traffic signal control strategies often result in gridlocks under oversaturated conditions. Perimeter Control (PC) has been proposed as a solution to protect a specific region (the Protected Network) and mitigate the spread of congestion within this region . While existing approaches typically assume a fixed perimeter for the protected region, the real challenge lies in dynamically identifying and adjusting this perimeter in real time as traffic conditions evolve, ensuring a well-fitted and effective traffic flow management. The objective is to avoid over-reacting to gridlock and applying well-suited network protection, when necessary.

Recent advancements have explored the potential of data-driven methods, particularly Deep Reinforcement Learning (DRL) algorithms, for decentralized traffic control systems, demonstrating strong scalability and adaptability.

In this context, this internship will focus on leveraging Deep Reinforcement Learning (DRL) with Graph Neural Networks (GNN) to integrate the local road network configuration and interaction in the state description of the agent dedicated to optimising the traffic crossing at one intersection. The goal is to develop a strategy where traffic control systems consider the interactions and impacts of traffic flows over time and across space. The integration of attention mechanisms within the graph modeling, coupled with reinforcement learning, will be explored to assess conditions both at individual intersections and across neighboring areas, creating a more adaptive approach. Moreover, this strategy will be tested as a key element for dynamically determine when a protected zone is required and define its boundaries.

The ideal candidate should preferably have:

•  Experience with Deep Learning and familiarity with concepts like reinforcement learning, Graph neural networks and attention mechanism.
• Proficiency in Python programming and experience with deep learning libraries such as PyTorch.
• Interest in urban mobility and traffic control systems.