Motion Plan of Maritime Autonomous Surface Ships by Dynamic Programming for Collision Avoidance and Speed...

Area of Science:

• Marine Engineering and Technology
• Robotics and Autonomous Systems
• Artificial Intelligence in Navigation

Background:

• Maritime Autonomous Surface Ships (MASS) are increasingly sophisticated, necessitating advanced control systems.
• Safe and efficient navigation is paramount for autonomous vessels operating alongside conventional ships.
• Existing motion planning algorithms may face challenges in complex, dynamic maritime environments.

Purpose of the Study:

• To investigate optimal motion planning strategies for MASS to ensure safe and timely task completion.
• To develop and evaluate algorithms that can effectively manage potential conflicts with other vessels.
• To enhance the reliability and efficiency of autonomous ship navigation systems.

Main Methods:

• Development of velocity obstacle models for both dynamic and static obstacles to define conflict-free regions.
• Proposal of a greedy interval-based motion-planning algorithm using the Velocity Obstacle (VO) model.
• Introduction of a way-blocking metric to assess collision risk and improve the greedy algorithm.
• Development of a novel Dynamic Programming (DP) method for optimal multi-interval motion planning, assuming constant velocities of surrounding ships.

Main Results:

• The greedy approach, while based on VO, demonstrated potential failures in avoiding collisions across successive intervals.
• The proposed way-blocking metric aimed to mitigate collision risks, improving upon the basic greedy strategy.
• Extensive simulations confirmed that the Dynamic Programming (DP) algorithm achieved the lowest overall collision rate.
• The DP algorithm exhibited superior sailing efficiency compared to the greedy approaches.

Conclusions:

• The Dynamic Programming (DP) method offers a more robust and effective solution for optimal motion planning in MASS.
• Advanced algorithms like DP are crucial for enhancing the safety and efficiency of autonomous maritime navigation.
• The study highlights the limitations of simpler greedy methods and the benefits of more sophisticated planning techniques for MASS.