A Dual-Layer Hybrid-A* Path Planning Algorithm for Unstructured Environments Based on Phase Windows

Area of Science:

• Robotics
• Artificial Intelligence
• Computer Science

Background:

• Mobile robots require robust path planning for autonomous navigation.
• Complex terrain poses challenges to robot stability and safety.
• Existing algorithms struggle with dynamic terrain and robot motion constraints.

Purpose of the Study:

• To propose a novel dual-layer Hybrid-A* algorithm for enhanced path planning.
• To improve robot stability and safety in unstructured environments.
• To address limitations of traditional path-planning methods.

Main Methods:

• Developed a dual-layer Hybrid-A* algorithm with dynamic phase windows.
• Integrated real-time feedback and adaptive adjustments for path refinement.
• Employed bicubic interpolation for elevation map refinement and posture estimation.
• Utilized dynamic cost functions, adaptive expansion radius, pruning, and phase-window activation.

Main Results:

• The proposed algorithm outperforms traditional methods in complex terrain.
• Generated optimized paths that satisfy robot motion constraints.
• Demonstrated improved efficiency and safety in real-world applications.

Conclusions:

• The dual-layer Hybrid-A* algorithm offers a significant advancement in mobile robot path planning.
• The approach enhances robot navigation capabilities in challenging environments.
• This method ensures both efficiency and safety for autonomous systems.