Quadratic programming-based autonomous cruise control of an intelligent tugboat

Area of Science:

• Marine Engineering
• Robotics
• Control Systems

Background:

• Autonomous vessels require sophisticated control for navigation and safety.
• Intelligent tugboats face complex multi-objective control challenges including cruising, collision avoidance, and actuator constraints.

Purpose of the Study:

• To develop a quadratic programming-based control method for autonomous tugboats.
• To ensure autonomous cruising to a target location with specified speed, heading, and path.
• To achieve collision avoidance and adhere to actuation limits.

Main Methods:

• Utilized back-stepping and sliding mode control to derive a desired control input for tracking error stability.
• Transformed positional collision avoidance constraints into input constraints using Nagumo's theorem.
• Synthesized a unified controller via quadratic programming to balance multiple control objectives.

Main Results:

• The proposed method successfully enforced prescribed safety distances and actuator limits.
• Demonstrated avoidance of large or persistent deviations from the reference trajectory.
• Showcased rapid re-establishment of desired speed and heading after collision avoidance maneuvers.

Conclusions:

• The quadratic programming-based method effectively addresses multi-objective control for autonomous tugboats.
• The approach ensures safe and efficient autonomous navigation within operational constraints.
• Validated through simulations, the method offers robust performance for intelligent vessel control.