Rapid global path planning algorithm for unmanned surface vehicles in large-scale and multi-island marine...

Area of Science:

• Robotics
• Marine Engineering
• Computational Geometry

Background:

• Path planning for Unmanned Surface Vehicles (USVs) is computationally intensive in large, complex marine environments.
• Existing Fast Marching Method (FMM)-based algorithms face efficiency challenges on high-resolution grids.
• Need for efficient global path planning algorithms for USVs with time constraints.

Purpose of the Study:

• To propose an efficient global path planning algorithm for USVs in large-scale, complex marine environments.
• To enhance computational efficiency while maintaining path precision.
• To enable flexible modification of obstacle-avoiding path segments.

Main Methods:

• A two-stage path planning approach using different spatial resolution (SR) grids.
• Initial planning on a low SR grid to identify effective regions.
• Subsequent planning on a high SR grid for high-precision path generation.
• Application of a modified Inshore-Distance-Constraint Fast Marching Square (IDC-FM^2) method for path generation and obstacle constraint.

Main Results:

• The proposed algorithm significantly improves calculation efficiency compared to single-grid methods.
• Smooth and safe global paths are generated for USVs.
• Obstacle-avoiding path segments can be flexibly modified.
• High precision of the planned path is maintained.

Conclusions:

• The dual-resolution grid approach effectively addresses computational efficiency for USV path planning.
• The modified IDC-FM^2 method allows for adaptable and safe navigation around obstacles.
• This algorithm offers a practical solution for time-critical USV operations in challenging marine settings.