Global Predefined-Time Adaptive Neural Network Control for Disturbed Pure-Feedback Nonlinear Systems With Zero...

Area of Science:

• Control Theory
• Nonlinear Systems
• Artificial Intelligence

Background:

• Traditional control methods for pure-feedback systems often achieve only semiglobal bounded tracking.
• Achieving global zero tracking error with predefined convergence time remains a significant challenge.

Purpose of the Study:

• To develop a global adaptive neural-network-based control algorithm for disturbed pure-feedback nonlinear systems.
• To guarantee zero tracking error convergence within a user-specified predefined time.

Main Methods:

• A mild semibound assumption for nonaffine functions is introduced to address pure-feedback system complexities.
• Radial basis function (RBF) neural networks and Young's inequality are employed to bound unknown functions and disturbances.
• Finite-time differentiators are utilized to estimate derivatives of virtual control inputs.

Main Results:

• The proposed controller ensures global convergence of tracking error to zero within a predefined time.
• Adaptive parameters and robust control gain are designed to manage uncertainties and disturbances.
• A practical example demonstrates the effectiveness and feasibility of the predefined-time control strategy.

Conclusions:

• The developed algorithm provides a robust and effective solution for global predefined-time control of disturbed pure-feedback nonlinear systems.
• This work advances the state-of-the-art by achieving global zero tracking error with user-defined convergence time.
• The method offers a practical and verifiable approach for complex control applications.