SACI framework-based fixed-time learning control for nonlinear systems with asymmetric constraints.

Area of Science:

• Control Theory
• Machine Learning
• Nonlinear Systems

Background:

• Constrained nonlinear systems present challenges in control design due to asymmetric states and input limitations.
• Existing control methods may struggle with rapid convergence and computational efficiency under such constraints.

Purpose of the Study:

• To develop a fixed-time learning control scheme for nonlinear systems with asymmetric state and input constraints.
• To enhance system performance and stability while minimizing computational and communication overhead.

Main Methods:

• A Sub-Actor-Critic-Identifier framework combined with optimal backstepping techniques.
• Integration of universal barrier functions and prescribed performance functions for constraint handling.
• Utilization of fuzzy logic systems for approximating nonlinear dynamics and compensating input constraints.
• Implementation of a dynamic event-triggered mechanism for controller sampling.

Main Results:

• Demonstrated boundedness and fixed-time convergence of all system signals via Lyapunov stability analysis.
• Effective handling of asymmetric state and input constraints using barrier and fuzzy logic systems.
• Significant reduction in controller sampling frequency through the event-triggered mechanism.

Conclusions:

• The proposed fixed-time learning control scheme effectively manages constrained nonlinear systems with asymmetric limitations.
• The Sub-Actor-Critic-Identifier framework ensures rapid convergence and robust stability.
• The dynamic event-triggered mechanism enhances practical applicability by reducing resource demands.