Observer-based adaptive affine formation maneuver control for underactuated multi-ASV systems

Area of Science:

• Robotics and Control Systems
• Marine Engineering
• Artificial Intelligence

Background:

• Autonomous surface vehicles (ASVs) require robust control for coordinated maneuvers.
• Limited sensing and unknown disturbances pose significant challenges in ASV formation control.
• Existing methods often struggle with real-time adaptation and estimation of unmeasured states.

Purpose of the Study:

• To develop an observer-based control strategy for underactuated multi-ASV systems.
• To enable ASV formations to perform affine maneuvers (scaling, rotation, translation) in complex environments.
• To address challenges of limited neighbor state information and unknown environmental disturbances.

Main Methods:

• A distributed super-twisting sliding mode observer estimates unavailable neighbor velocities.
• An observer-based affine formation controller is designed for maneuver execution.
• A radial basis function neural network with adaptive weights compensates for unknown disturbances.
• Stability analysis guarantees uniform ultimate boundedness of tracking and estimation errors.

Main Results:

• The proposed strategy effectively estimates unavailable velocities using the observer.
• The controller enables stable affine formation maneuvers.
• The neural network successfully compensates for unknown disturbances.
• Numerical simulations demonstrate high formation tracking accuracy with low mean absolute error (0.310) and root mean square error (1.201).

Conclusions:

• The developed observer and controller are effective for affine formation maneuver control of underactuated ASVs.
• The strategy provides robust performance in the presence of limited sensing and unknown disturbances.
• The approach enhances the navigation capabilities of multi-ASV systems in challenging environments.