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

Area of Science:

• Intelligent Transportation Systems
• Autonomous Vehicle Control
• Traffic Management

Background:

• Connected autonomous vehicles (CAVs) offer potential for enhanced traffic efficiency and safety through cooperative decision-making.
• Existing cooperative driving strategies primarily focus on car-following, limiting broader traffic coordination.
• Real-time optimization of complex traffic scenarios involving multiple CAVs presents significant computational challenges.

Purpose of the Study:

• To develop a centralized framework for coordinating lane-change decisions among multiple CAVs on highways.
• To optimize joint speed adaptation and lane-change maneuvers for maximizing traffic utility while ensuring safety.
• To address the computational complexity of real-time decision-making in cooperative driving.

Main Methods:

• Formulation of a mixed-integer optimization problem for joint decision-making.
• Development of a priority-aware search strategy to reduce computational load.
• Integration of Model Predictive Control (MPC) with Artificial Potential Fields (APFs) for safety and guidance.
• Utilizing vehicle-to-infrastructure (V2I) communication for centralized planning.

Main Results:

• The proposed framework effectively balances safety, mobility, and system-level coordination.
• Demonstrated real-time feasibility through significantly reduced computational complexity.
• Successful optimization of acceleration, braking, and lane-change commands for multiple CAVs simultaneously.
• Validation of collision-free maneuvers and enhanced traffic utility in simulations.

Conclusions:

• The developed framework provides a scalable and computationally efficient solution for cooperative lane-change decisions in CAVs.
• This approach advances intelligent transportation systems by enabling sophisticated real-time traffic management.
• The methodology supports the future deployment of advanced autonomous driving functionalities on highways.