Formation Learning Control of Multiple Autonomous Underwater Vehicles With Heterogeneous Nonlinear Uncertain...

Area of Science:

• Robotics
• Control Systems
• Marine Engineering

Background:

• Autonomous underwater vehicles (AUVs) require sophisticated control for coordinated tasks.
• Existing formation control methods often struggle with uncertain and heterogeneous vehicle dynamics.
• Real-time identification of AUV dynamics is crucial for adaptive control.

Purpose of the Study:

• Introduce a new formation learning control concept for multi-AUV systems.
• Address the challenge of distributed formation tracking with uncertain nonlinear dynamics.
• Develop a control scheme capable of both formation control and dynamic system identification.

Main Methods:

• Propose a two-layer distributed formation learning control scheme.
• Utilize a distributed adaptive observer in the upper layer.
• Implement a decentralized deterministic learning controller in the lower layer.

Main Results:

• The proposed scheme effectively achieves distributed formation tracking control for heterogeneous AUVs.
• The control protocol accurately identifies nonlinear uncertain dynamics of individual AUVs.
• The system learns from experience to enhance formation control performance.

Conclusions:

• The novel formation learning control scheme offers a robust solution for multi-AUV systems.
• Distributed and adaptive control is achieved without reliance on global information.
• The method successfully integrates formation control with dynamic system identification.