Decentralized adaptive neural prescribed performance control for high-order stochastic switched nonlinear...

Area of Science:

• Control Theory
• Artificial Intelligence
• Stochastic Systems

Background:

• High-order stochastic nonlinear systems present significant control challenges due to unknown dynamics and interconnected nonlinearities.
• Ensuring prescribed performance and stability under arbitrary switching signals is a complex problem in decentralized control.

Purpose of the Study:

• To develop a decentralized adaptive neural backstepping control strategy for high-order stochastic nonlinear systems.
• To address unknown interconnected nonlinearities and arbitrary switching signals while guaranteeing prescribed performance.
• To reduce the number of learning parameters and overcome over-parameterization issues.

Main Methods:

• Prescribed Performance Control (PPC) is utilized to ensure tracking performance without initial error constraints.
• Radial Basis Function (RBF) neural networks are employed to approximate unknown system dynamics.
• A common Lyapunov stability method is applied to design the decentralized adaptive neural controller.

Main Results:

• The proposed controller ensures that all signals in the closed-loop system are semi-globally uniformly ultimately bounded (SGUUB).
• Prescribed tracking control performance is guaranteed even under arbitrary switching conditions.
• The method effectively handles unknown system dynamics and interconnected nonlinearities.

Conclusions:

• The developed decentralized adaptive neural control scheme is effective for high-order stochastic nonlinear systems with unknown dynamics and arbitrary switchings.
• The approach successfully guarantees prescribed performance and system stability, validated by simulation results.
• This work contributes a robust control strategy with reduced parameter learning for complex nonlinear systems.