Prescribed Performance Output Feedback/Observer-Free Robust Adaptive Control of Uncertain Systems Using Neural...

Area of Science:

• Control Systems Engineering
• Nonlinear System Dynamics
• Artificial Intelligence in Engineering

Background:

• Designing controllers for uncertain nonlinear systems is challenging due to complex dynamics and external disturbances.
• Existing methods often require detailed system models or observers, limiting their applicability.
• Ensuring both output performance and internal signal stability is crucial for robust control.

Purpose of the Study:

• To develop an output feedback, observer-free continuous controller for multiple-input-multiple-output (MIMO) uncertain nonlinear systems.
• To guarantee prescribed performance bounds on the system output and boundedness of all closed-loop signals.
• To address the impact of additive external disturbances and unmodeled dynamics.

Main Methods:

• Utilized a neural network-based approach for controller design.
• Employed an output feedback strategy, eliminating the need for an observer.
• Incorporated assumptions on system properties, including unboundedness observability and output Lagrange stability of unmodeled dynamics.
• Ensured the nominal system is output feedback equivalent to a strictly passive one.

Main Results:

• Successfully designed a continuous controller capable of handling MIMO uncertain nonlinear systems.
• Guaranteed prescribed performance bounds on the system output.
• Ensured the boundedness of all other closed-loop signals in the presence of disturbances and unmodeled dynamics.
• Demonstrated the controller's effectiveness through simulations on an induction motor system.

Conclusions:

• The proposed neural network controller offers a robust solution for controlling complex nonlinear systems without requiring an observer.
• The controller effectively manages external disturbances and unmodeled dynamics while maintaining performance and stability.
• The observer-free, output feedback approach provides a practical advancement in nonlinear control engineering.