Neural network disturbance observer-based distributed finite-time formation tracking control for multiple unmanned...

Area of Science:

• Robotics
• Control Systems Engineering
• Aerospace Engineering

Background:

• Unmanned helicopters require robust control for formation tasks.
• External disturbances and model uncertainties challenge precise formation keeping.
• Multiple-time scale dynamics complicate helicopter control system design.

Purpose of the Study:

• To address the distributed finite-time formation tracking control problem for multiple unmanned helicopters.
• To maintain formation positions of follower helicopters despite external interferences.
• To develop a control strategy that guarantees finite-time convergence.

Main Methods:

• A novel finite-time multivariable neural network disturbance observer (FMNNDO) using radial basis function neural networks (RBFNN) to estimate disturbances and uncertainties.
• Dynamic compensation of neural network approximation errors via adaptive laws.
• Design of distributed finite-time formation tracking and attitude tracking controllers using nonsingular fast terminal sliding mode (NFTSM).
• Development of a finite-time sliding mode integral filter for estimating the second derivative of virtual desired attitude signals.

Main Results:

• The proposed FMNNDO effectively estimates external disturbances and model uncertainties.
• The designed controllers achieve finite-time formation and attitude tracking.
• Lyapunov analysis and multiple-time scale principles confirm the finite-time convergence.
• Numerical simulations validate the effectiveness of the FMNNDO and controllers.

Conclusions:

• The developed control strategy ensures robust finite-time formation tracking for multiple unmanned helicopters.
• The integration of FMNNDO and NFTSM controllers provides a viable solution for complex aerial formation tasks.
• The study demonstrates the feasibility of achieving precise formation control under uncertain and dynamic conditions.