Neural network-based optimal adaptive output feedback control of a helicopter UAV

Area of Science:

• Robotics and Control Systems
• Artificial Intelligence in Engineering
• Aerospace Engineering

Background:

• Helicopter unmanned aerial vehicles (UAVs) are critical in military and civilian sectors.
• Designing high-performance controllers for underactuated nonlinear helicopter UAVs is challenging.
• Existing control methods may struggle with complex dynamics and real-time adaptation.

Purpose of the Study:

• To develop an optimal output-feedback controller for trajectory tracking of helicopter UAVs.
• To utilize a neural network (NN) for enhanced control performance and stability.
• To address the challenges posed by the underactuated nonlinear nature of helicopter UAVs.

Main Methods:

• An output-feedback control system employing backstepping methodology with kinematic and dynamic controllers.
• Integration of a neural network (NN) observer for state estimation.
• An online approximator-based dynamic controller that learns the Hamilton-Jacobi-Bellman equation for optimal control input calculation.
• Utilizing a single NN for cost function approximation and optimal tracking.

Main Results:

• The proposed controller demonstrates effective trajectory tracking for helicopter UAVs.
• The closed-loop system stability is rigorously proven using Lyapunov analysis.
• Simulation results validate the superior performance of the NN-based optimal control design.

Conclusions:

• The developed optimal output-feedback controller effectively achieves trajectory tracking for helicopter UAVs.
• The use of a neural network observer and controller enhances system performance and stability.
• This approach offers a promising solution for controlling complex underactuated aerial robotic systems.