Autonomous Docking: Trajectory planning and dynamic route adaptation for autonomously docking a marine vessel using Model Predictive Control

Abstract

This thesis presents a comprehensive study on the application of Model Predictive Control (MPC) for the autonomous docking of marine vessels. The research focuses on developing and implementing trajectory planning and dynamic route adaptation algorithms that enable a vessel to autonomously dock in various maritime environments. The research on autonomous docking is important for its potential to enhance safety, efficiency, and reliability in maritime operations, particularly in crowded or challenging docking scenarios. This technology aims to minimize human error and simplify the docking procedure in busy marine environments. The key challenges addressed include path planning, collision avoidance, trajectory tracking, and the integration of real-time dynamic adjustments to account for moving obstacles and environmental changes. Our methodology utilizes MPC to continuously predict and optimize the vessel’s path, thereby ensuring safe and efficient docking maneuvers. A simulation environment created in Python and real-world simulations created in a Unity-based environment were utilized to validate the effectiveness of the proposed algorithm. Simulation results demonstrated a functional trajectory planner capable of successfully following a reference path, avoiding obstacles, and docking a marine vessel in narrow spaces, indicating the potential for using MPC to autonomously dock a boat. Initial tests in a real-world environment were performed to further confirm the potential of the proposed solution. Comparative performance analysis highlights the strengths and limitations of the system during different conditions, demonstrating its potential for real-world application. Future works aim to enhance the complexity of the maritime scenarios and vessel dynamics handled by the algorithms, as well as additional testing in real-world environments to further validate and refine the system. The presented analysis also demonstrates the potential for additional features to improve the stability and reliability of the MPC-based system. This includes the integration of various sensor data for extensive environmental mapping providing real-time updates about the surroundings to ensure a safer docking procedure.