A Novel Extreme Learning Control Framework of Unmanned Surface Vehicles

Area of Science:

• Robotics
• Control Systems Engineering
• Artificial Intelligence

Background:

• Unmanned Surface Vehicles (USVs) require robust control systems to navigate complex environments.
• Unknown system dynamics and external disturbances pose significant challenges for USV tracking control.
• Existing control methods may struggle with accurate real-time identification of nonlinearities.

Purpose of the Study:

• To propose a novel Extreme Learning Control (ELC) framework for USV tracking.
• To address challenges posed by unknown dynamics and external disturbances in USV control.
• To achieve accurate and robust tracking performance for USVs.

Main Methods:

• Utilizing a Single-Hidden-Layer Feedforward Network (SLFN) with random hidden nodes for nonlinear system identification.
• Defining an error surface and transformed states to encapsulate unknown dynamics and disturbances.
• Employing adaptive projection-based laws derived from Lyapunov stability theory for output weight updates.
• Incorporating an error compensator to mitigate approximation residuals.

Main Results:

• The proposed ELC framework accurately identifies lumped nonlinearities without prior system knowledge.
• Adaptive control laws ensure global asymptotic stability of the closed-loop system.
• Simulation results demonstrate high accuracy in both tracking and approximation tasks.
• The ELC system exhibits robustness against unknown dynamics and external disturbances.

Conclusions:

• The developed ELC framework offers an effective solution for USV tracking control under uncertain conditions.
• The integration of SLFNs and Lyapunov-based adaptive control enhances system performance and stability.
• This approach provides a reliable method for achieving precise navigation for unmanned surface vehicles.