Kong, Aijing (Tongji University), Wu, Xian (Tongji University), Chen, Xinbo (Tongji University), Zhang, Hao (Tongji University), Hang, Peng (Tongji University), Sun, Jian (Tongji University)

A Distributed Platoon Reconfiguration Framework in Emergency Scenarios with Heterogeneous MARL

Scheduled for presentation during the Invited Session "S20a-Foundation Model-Enabled Scene Understanding, Reasoning, and Decision-Making for Autonomous Driving and ITS" (TH-LM-T20), Thursday, November 20, 2025, 10:30−10:50, 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

Abstract

Autonomous driving platoons, with their tightly coordinated formations, offer significant gains in traffic safety, efficiency, and fuel economy. However, limited reaction time renders them vulnerable to collisions in sudden hazardous scenarios. This paper presents a heterogeneous, distributed platoon reorganization architecture that adaptively coordinates formations to avoid unexpected obstacles. The decision‑making framework is based on multi‑agent reinforcement learning and incorporates a dynamic, knowledge‑based masking mechanism that constrains each agent’s action space by role and embeds prior knowledge to improve learning efficiency and safety. Our framework scales to platoons of up to eight vehicles. Numerical simulations in sudden hazard scenarios demonstrate that, compared with baseline methods, the proposed approach yields faster and more reliable reconfiguration and markedly enhances driving safety.