Event-Triggered-Based Adaptive Neural Network Secure Strategy for Stochastic Networked Nonlinear Systems Under DoS Attacks
Junsheng Zhao, Xin Zhang, Zong‐Yao Sun, Weihai Zhang
Abstract
An adaptive neural networks event-triggered secure strategy is investigated to explore stochastic networked nonlinear systems with Denial-of-Service (DoS) attacks and unknown dead zone. In order to alleviate the negative effects of DoS attacks, a state estimator is constructed to approximate the immeasurable states. And the unknown dead zone input function is described as a bounded disturbance and a time-varying nonlinear function to offset the unknown dead zone effect. To defend against DoS attacks while ensuring system stability, an adaptive event-triggered prescribed performance controller is designed, ensuring that all signals of the closed-loop system are bounded in probability, and the tracking error tends to the designed performance bound within a predefined finite time. Meanwhile, this secure strategy can completely eliminate potential Zeno behavior. Subsequently, the modified average dwell time (ADT) approach was integrated with Lyapunov stability theory to establish the stability of the system. Eventually, two simulation results are utilized to prove the effectiveness of the developed strategy.