Bearing-Based Adaptive Neural Formation Scaling Control for Autonomous Surface Vehicles With Uncertainties and Input...

Area of Science:

• Robotics
• Control Systems Engineering
• Marine Engineering

Background:

• Maintaining formation and collision avoidance for autonomous surface vehicles (ASVs) is challenging, especially when transitioning between waterways of different sizes.
• Scaling formation size while preserving shape is a key strategy for adaptable ASV operations.
• Existing control methods often struggle with uncertainties and actuator limitations in ASV formations.

Purpose of the Study:

• To develop a novel adaptive neural formation scaling control scheme for ASVs.
• To enable ASVs to scale formation size while maintaining shape and avoiding collisions.
• To address uncertainties and input saturation in ASV control systems.

Main Methods:

• Utilizing bearing rigidity theory to select leader ASVs and program their trajectories for formation scaling.
• Employing adaptive neural techniques for follower ASVs to track leaders, allowing size adjustments solely by leaders.
• Simplifying neural network weight updates to one-parameter estimation and introducing auxiliary systems to mitigate actuator limitations.

Main Results:

• The proposed scheme achieves the desired formation scaling maneuver for ASVs under infinitesimal bearing rigidity conditions.
• Formation errors are proven to be uniformly ultimately bounded, ensuring stable control.
• The method simultaneously satisfies directional, computational, and actuator constraints, outperforming existing approaches.

Conclusions:

• The developed bearing-based adaptive neural formation scaling control scheme is effective for ASVs.
• The approach offers robustness against uncertainties and actuator saturation.
• This research provides a significant advancement in coordinated control for multi-ASV systems.