Two potential fields fused adaptive path planning system for autonomous vehicle under different velocities

Area of Science:

• Robotics and Control Systems
• Artificial Intelligence
• Autonomous Vehicle Navigation

Background:

• Path planning is crucial for autonomous vehicles (AVs) but struggles with dynamic velocities and diverse obstacles.
• Existing methods face challenges in adapting to both static (e.g., road boundaries) and dynamic obstacles simultaneously.
• The need for robust path planning systems that ensure safety and efficiency in complex driving scenarios is paramount.

Purpose of the Study:

• To develop an adaptive path planning system for autonomous vehicles (AVs) capable of handling varying velocities and multiple obstacle types.
• To enhance the adaptability and robustness of AV path planning algorithms.
• To improve the safety and performance of autonomous navigation systems.

Main Methods:

• A two potential fields fused adaptive path planning system (TPFF-APPS) was developed, integrating a potential field fusion controller and an adaptive weight assignment unit.
• A novel potential velocity field, incorporating velocity information, was fused with a traditional artificial potential field.
• An adaptive weight assignment unit was designed to dynamically adjust the influence of different potential fields based on environmental conditions.

Main Results:

• The proposed TPFF-APPS demonstrated excellent performance in path planning simulations.
• The system effectively adapted to different velocities, improving navigation responsiveness.
• The system successfully handled various obstacle types, including road boundaries and dynamic obstacles, ensuring safe navigation.

Conclusions:

• The TPFF-APPS provides a robust solution for autonomous vehicle path planning under diverse conditions.
• The fusion of potential fields and adaptive weighting significantly enhances navigation capabilities.
• This approach contributes to the development of safer and more efficient autonomous driving systems.