Adaptive Neural State-Feedback Tracking Control of Stochastic Nonlinear Switched Systems: An Average Dwell-Time...

Area of Science:

• Control Theory
• Nonlinear Systems
• Stochastic Systems

Background:

• Addressing adaptive neural state-feedback tracking control for stochastic nonstrict-feedback nonlinear switched systems with unknown nonlinearities.
• Overcoming design challenges posed by nonstrict-feedback structures and unknown dynamics.

Purpose of the Study:

• To develop a robust adaptive neural controller for stochastic nonstrict-feedback nonlinear switched systems.
• To ensure semiglobal uniform ultimate boundedness of system signals and convergence of tracking error.

Main Methods:

• Utilizing radial basis function neural networks for approximating unknown nonlinear functions.
• Employing a variable separation technique to handle nonstrict-feedback structures.
• Constructing individual Lyapunov functions for subsystems using bounds of control gain functions.
• Integrating average dwell-time scheme with adaptive backstepping design.

Main Results:

• A novel adaptive neural state-feedback controller design algorithm is presented.
• All signals of the switched closed-loop system are proven to be in probability semiglobally uniformly ultimately bounded.
• Tracking error converges to a small neighborhood of the origin in probability.
• The control scheme's effectiveness is validated through two simulation examples.

Conclusions:

• The developed adaptive neural control scheme effectively addresses the complexities of stochastic nonstrict-feedback nonlinear switched systems.
• The proposed method guarantees system stability and precise tracking performance.
• The approach offers a significant advancement in adaptive control for uncertain nonlinear systems.