Human-Machine Cooperative Trajectory Planning for Advanced Driving Assistance System in Lane Changing Scenarios

This paper presents a human-machine cooperative trajectory planning system for advanced driving assistance systems in lane-changing scenarios. Firstly, it establishes a vehicle dynamics model and provides evaluation indicators for safe and comfortable driving under lane-changing conditions. Secondly, it constructs a human-machine collaborative driving framework, designs a trajectory pre-generation strategy, and develops a learning-based weight generation model for human-machine collaborative driving. Finally, trajectory planning is implemented based on fuzzy logic theory and PID control. The system is validated through a Simulink/Carsim co-simulation experimental platform with driver-in-the-loop manner. The results demonstrate that the human-machine cooperative driving approach offers better performance in driving stability and safety compared to both autonomous driving and human driving alone.