A novel model predictive artificial potential field based ship motion planning method considering COLREGs for complex...

Area of Science:

• Maritime Autonomous Surface Ships (MASS)
• Robotics and Control Systems
• Navigation and Path Planning

Background:

• Ship motion planning is critical for autonomous navigation.
• Traditional Artificial Potential Field (APF) methods face local optima issues.
• Complex encounter scenarios require robust collision avoidance.

Purpose of the Study:

• To propose a novel Model Predictive Artificial Potential Field (MPAPF) motion planning method.
• To address collision avoidance challenges in complex maritime encounters.
• To ensure safe and feasible path generation for MASS.

Main Methods:

• Developed a new ship domain with a closed interval potential field function.
• Utilized a Nomoto model for kinematically conforming path generation.
• Formulated the problem as a non-linear optimization with constraints (maneuverability, rules, waterways).

Main Results:

• The MPAPF method effectively solves local optima problems.
• Demonstrated superior performance in complex encounter scenarios compared to APF, A-star, and RRT.
• Generated feasible motion paths ensuring collision avoidance.

Conclusions:

• The proposed MPAPF algorithm provides a robust solution for MASS motion planning.
• It enhances safety and efficiency in complex maritime navigation scenarios.
• The method offers a significant advancement over existing path planning techniques.