Multi-Ship Control and Collision Avoidance Using MPC and RBF-Based Trajectory Predictions.

Area of Science:

• Maritime Engineering
• Robotics and Control Systems
• Artificial Intelligence in Navigation

Background:

• Collision avoidance is critical for maritime safety, with increasing interest in autonomous vessel control.
• Existing systems often lack sophisticated prediction capabilities, impacting timely evasive actions.
• Large vessels like container ships have slow dynamics, necessitating proactive collision avoidance strategies.

Purpose of the Study:

• To address the multi-ship control problem in automatic collision avoidance for surface vessels.
• To develop and evaluate a model predictive controller (MPC) utilizing advanced obstacle ship trajectory prediction.
• To enhance decision support for officers and enable autonomous vessel operation.

Main Methods:

• Utilized a model predictive controller (MPC) for multi-ship control.
• Developed obstacle ship trajectory prediction models based on the Radial Basis Function (RBF) framework.
• Trained prediction models using real Automatic Identification System (AIS) data from an open-source database.
• Evaluated the controller's performance in a real-life case study in the Miami port area.

Main Results:

• The proposed MPC with RBF-based trajectory prediction demonstrated effective collision risk inference and timely evasive control.
• The method successfully accounted for the slow dynamics of large vessels, such as container ships.
• Generated trajectories were assessed for safety, economy, and compliance with International Regulations for Preventing Collisions at Sea (COLREGs).
• Performance was benchmarked against an MPC controller using simpler straight-line predictions.

Conclusions:

• Sophisticated trajectory prediction models significantly improve the effectiveness of automatic collision avoidance systems.
• The developed MPC enhances cooperation between controlled vessels and ensures safer maritime operations.
• The approach offers a viable solution for improving navigation safety and autonomy in complex maritime environments.