Research on Lane-Changing Decision Making and Planning of Autonomous Vehicles Based on GCN and Multi-Segment...

Area of Science:

• Autonomous Systems
• Robotics
• Artificial Intelligence

Background:

• Navigating dynamic driving environments requires sophisticated decision-making and trajectory planning for autonomous vehicles.
• Understanding and modeling interactive effects between vehicles is crucial for safe and efficient lane changes.

Purpose of the Study:

• To propose an advanced method for autonomous vehicle lane-changing behavior decision-making and trajectory planning.
• To leverage graph convolutional networks (GCNs) and multi-segment polynomial curve optimization for enhanced performance.

Main Methods:

• Hierarchical modeling of dynamic driving environments using graph-structured data.
• Application of graph convolutional neural networks (GCNs) for processing interaction information and generating behavior commands.
• Optimization-based multi-segment polynomial curve trajectory planning to ensure collision-free motion.

Main Results:

• The proposed method demonstrated superior performance compared to traditional approaches in simulations and real-world experiments.
• Achieved good robustness, real-time performance, and strong scenario generalization capabilities.
• Successfully generated collision-free, dynamically constrained motion trajectories for lane-changing maneuvers.

Conclusions:

• The GCNs and polynomial curve optimization method offers a robust and efficient solution for autonomous vehicle lane-changing.
• The approach effectively handles complex interactions in dynamic driving environments.
• Validated effectiveness through comprehensive simulations and on-road vehicle experiments.