ITSC 2025 Paper Abstract

Paper FR-EA-T35.3

Wang, Leizhen (Monash University), Duan, Peibo (Monash University), Lyu, Cheng (Technical University of Munich), Wang, Zewen (Southeast university), Zheng, Nan (Monash University), Ma, Zhenliang (KTH Royal Institute of Technology)

Scalable and Reliable Multi-Agent Reinforcement Learning for Traffic Assignment

Scheduled for presentation during the Regular Session "S35b-Optimization, Control, and Learning for Efficient and Resilient ITS" (FR-EA-T35), Friday, November 21, 2025, 14:10−14:30, Surfers Paradise 2

2025 IEEE 28th International Conference on Intelligent Transportation Systems (ITSC), November 18-21, 2025, Gold Coast, Australia

This information is tentative and subject to change. Compiled on October 18, 2025

Keywords Transportation Optimization Techniques and Multi-modal Urban Mobility, Data Analytics and Real-time Decision Making for Autonomous Traffic Management

Abstract

Escalating urbanization and increased travel demand impose stringent benchmarks on traffic assignment methodologies. Multi-agent reinforcement learning (MARL) approaches outperform traditional methods in emulating adaptive travel routing without requiring explicit system dynamics, which is beneficial for real-world implementation. Nevertheless, MARL frameworks face challenges in scalability and reliability when managing extensive networks with substantial travel demand, which restricts their practical use in solving large-scale traffic assignment problems. This research introduces an innovative MARL framework for traffic assignment, redefining agents as origin-destination (OD) routers instead of individual travelers, enhancing scalability. Additionally, a specialized action space formulation using a proposed Dirichlet-based strategy and a reward formulation based on the local relative gap is crafted to efficiently reach optimal solutions, increasing model reliability. Experiments demonstrate the proposed MARL framework effectively handles medium-sized networks with extensive and varied city-level OD demand, surpassing existing MARL methods. Applied to the SiouxFalls network, the method achieves better assignment outcomes in fewer steps, reducing the relative gap by 83.6% compared to traditional techniques.