Towards Emergency Scenarios: An Integrated Decision-Making Framework of Multi-Lane Platoon Reorganization

Vehicle platoons are considered an effective strategy to enhance road safety and traffic efficiency, yet their capability to respond to sudden emergencies remains insufficiently addressed. Existing studies mainly focus on stable formation control or coordination in normal conditions, while systematic frameworks for dynamic platoon reorganization in high-risk scenarios are lacking. To fill this gap, this paper proposes a platoon distribution reorganization decision-making framework that integrates three layers: platoon distribution, vehicle cooperative decision-making, and vehicle planning and control. First, a reinforcement-learning-based distribution model is developed, in which a risk potential field quantitatively evaluates driving risks and a tailored reward function guides the reorganization process. Second, a coalition-game-based decision-making model captures inter-vehicle cooperation by dynamically forming coalitions and generating optimal actions. Moreover, a novel graph-theory-based Platoon Disposition Index (PDI) is introduced into the reward function to evaluate platoon distribution states and accelerate reorganization. Finally, validation under two high-risk scenarios with random traffic flows demonstrates that the proposed framework significantly reduces collision risk and improves traffic efficiency, while the inclusion of PDI further shortens reorganization time and enhances overall efficiency.