Rule-adaptive lane-changing trajectory planning method for autonomous vehicles driven by dynamic risk information.

Area of Science:

• Robotics
• Artificial Intelligence
• Automotive Engineering

Background:

• Autonomous driving technology faces challenges in balancing safety and efficiency in dynamic traffic.
• Existing trajectory planning methods often use fixed safety constraints, limiting driving efficiency.
• Vehicle suspension characteristics significantly impact driving safety and stability.

Purpose of the Study:

• To propose a dynamic risk-information-driven adaptive trajectory planning method for autonomous vehicles (AVs).
• To enhance driving efficiency by adaptively adjusting safety constraints based on risk assessment.
• To ensure driving safety is maintained while improving efficiency.

Main Methods:

• Developed collision and instability risk assessment indices using a three-way-coupled dynamic model.
• Implemented a safety risk assessment module to evaluate potential risks.
• Integrated adaptive safety constraint adjustment into trajectory planning and optimization.
• Utilized a cost function for trajectory selection and optimization.

Main Results:

• The proposed method adaptively adjusts safety constraints based on real-time risk assessment.
• Driving efficiency improved by 55.9% in instability constraint scenarios.
• Driving efficiency improved by 27.86% in collision constraint scenarios.
• Driving safety was maintained throughout the experiments.

Conclusions:

• Dynamic risk assessment and adaptive safety constraints significantly improve AV driving efficiency.
• The proposed method offers a viable solution for enhancing AV performance in complex environments.
• Future research can explore further integration of vehicle dynamics and risk assessment for AVs.