Event-Based Adaptive Neural Tracking Control for Discrete-Time Stochastic Nonlinear Systems: A Triggering Threshold...

Area of Science:

• Control Systems Engineering
• Nonlinear Dynamics
• Stochastic Systems

Background:

• Discrete-time strict-feedback nonlinear systems face challenges with stochastic noises and limited communication.
• Event-triggered (ET) mechanisms aim to reduce communication load by transmitting control signals only when necessary.

Purpose of the Study:

• To develop an event-triggered tracking control strategy for discrete-time nonlinear systems.
• To design a novel adaptive neural controller that addresses stochasticity and communication constraints.
• To improve tracking accuracy and communication efficiency in control systems.

Main Methods:

• A systematic framework using backstepping procedure for adaptive neural controller design.
• An event-triggered mechanism with a fixed triggering threshold.
• A novel event-triggered actuator incorporating a threshold compensation operator based on hyperbolic tangent and sign functions.

Main Results:

• The proposed framework overcomes noncausality and singularity problems, avoiding virtual control law approximation.
• The ET-based actuator with compensation improves tracking accuracy, especially at triggering instants.
• The compensation operator mitigates communication load and enlarges the ET threshold range.

Conclusions:

• The developed ET control strategy effectively handles stochastic noises and communication limitations in discrete-time nonlinear systems.
• The novel ET-based actuator with threshold compensation offers significant advantages over traditional approaches.
• The approach demonstrates improved performance and efficiency, validated by numerical and practical examples.