Scalable cooperative lane-change management for connected autonomous vehicles using MPC-based decision coordination

The advancement of connected autonomous vehicles (CAVs) enables cooperative decision-making for traffic efficiency and safety. This study proposes a centralized lane-change decision coordination framework for multiple CAVs on highways, extending cooperative driving beyond traditional car-following strategies. The controller jointly optimizes speed adaptation and lane-change decisions over a finite prediction horizon to maximize overall traffic utility while ensuring collision-free maneuvers. Assuming reliable vehicle-to-infrastructure (V2I) communication, the planner computes acceleration, braking, and lane-change commands for all vehicles simultaneously. The underlying decision process is formulated as a mixed-integer optimization problem, which is computationally prohibitive for real-time deployment. To address this challenge, a priority-aware search strategy is developed to evaluate only the most promising lane-change combinations at each time step, integrated within a Model Predictive Control (MPC) framework enhanced with Artificial Potential Fields (APFs) for safety assurance and motion guidance. Simulation results demonstrate that the proposed framework effectively balances safety, mobility, and system-level coordination while achieving real-time feasibility through significantly reduced computational complexity. The approach offers a scalable solution for future intelligent transportation infrastructures and real-time traffic management applications.