Switching Event-Triggered Adaptive Resilient Dynamic Surface Control for Stochastic Nonlinear CPSs With Unknown...

Area of Science:

• Control Systems Engineering
• Cyber-Physical Systems Security
• Stochastic Systems

Background:

• Cyber-physical systems (CPSs) are vulnerable to unknown deception attacks, compromising state availability.
• Designing controllers for uncertain nonlinear stochastic CPSs with attacks is complex due to unknown nonlinearities and perturbations.
• Existing methods struggle with the combined challenges of system uncertainty, stochasticity, and adversarial attacks.

Purpose of the Study:

• To design an adaptive resilient dynamic surface controller for uncertain nonlinear lower triangular stochastic CPSs.
• To address unknown deception attacks by developing a robust control strategy.
• To improve resource efficiency through a switching threshold event-triggered mechanism.

Main Methods:

• Utilized radial basis function (RBF) neural networks to handle unknown nonlinearities.
• Implemented dynamic surface control (DSC) to mitigate the 'explosion of complexity' issue in backstepping design.
• Developed a novel coordinate transformation and attack compensators to manage deception attacks.
• Employed a switching threshold event-triggered mechanism for efficient control signal transmission.

Main Results:

• All closed-loop signals remain bounded in probability under the proposed control scheme.
• Stabilization errors converge to an adjustable neighborhood of the origin in probability.
• The controller effectively handles unknown nonlinearities, stochastic perturbations, and deception attacks.

Conclusions:

• The proposed adaptive resilient dynamic surface control scheme successfully stabilizes uncertain nonlinear stochastic CPSs under deception attacks.
• The integration of RBF neural networks, DSC, and event-triggered control offers a robust and efficient solution.
• Simulation results on a double chemical reactor system validate the effectiveness of the proposed control design.