Boundary-locked event-triggered mechanism-based adaptive flocking control for multi-USV systems

Area of Science:

• Marine robotics
• Control systems engineering
• Artificial intelligence

Background:

• Unmanned surface vehicle swarms (USVS) are vital for marine operations but face coordination challenges.
• Flocking cohesion, communication, and uncertainty hinder effective USVS deployment.
• Existing control strategies struggle with dynamic marine environments and communication constraints.

Purpose of the Study:

• To develop an event-based adaptive flocking control law for USVS.
• To enhance flock cohesion and ensure network connectivity.
• To reduce communication overhead and improve system robustness against uncertainties.

Main Methods:

• Designed an integrated dual-mode potential function (APF and DCPF) for flock cohesion and connectivity.
• Implemented a boundary-locked event-triggered (BLET) mechanism with MIET and MTIT to minimize communication.
• Incorporated a reinforcement learning-based echo state network (RLESN) for adaptive control and disturbance rejection.

Main Results:

• The proposed control law effectively enhances flock cohesion and maintains connectivity.
• The BLET mechanism significantly reduces communication load while ensuring system stability.
• The RLESN successfully compensates for unmodeled dynamics and external disturbances, improving robustness.

Conclusions:

• The developed event-based adaptive flocking control law offers a robust and efficient solution for USVS coordination.
• The integrated approach addresses key challenges in flock cohesion, communication, and uncertainty management.
• Simulation results validate the superiority of the proposed methodology over existing methods.