Event-triggered H_∞ Control And Filtering Of Networked Systems Under Cyber Attack

For decades,networked control systems(NCSs)have been a research hot spot in the field of control engineering.Its sensors,controllers,actuators and other nodes are connected through the network,and the relevant signals are transmitted and exchanged through the network,so as to realize resource sharing,remote control and operation.It has been widely used in process industry,automobile manufacturing,aerospace and military.Due to the introduction of communication network,the problems of network security and limited network resources have been increasingly prominent.Therefore,it is very important to study the analysis and synthesis of networked systems with the event-triggered scheme and cyber-attacks.This thesis is concerned with event-triggered H∞ control and filtering for networked systems under cyber-attacks.By using more advanced dynamic or decentralized event-triggered scheme,the number of transmitted sampled data packets is greatly reduced,so as to effectively save system resources.Based on the event-triggered scheme,some closed-loop control systems or filtering error systems are established by considering cyber-attacks,signal quantization and sensor saturation.Based on the Lyapunov stability theory,the analysis and design methods of H∞ controller and filter are presented,respectively.The main contents are as follows:(1)The stability analysis of delayed neural networks based on the event-triggered scheme is discussed.By introducing the event-triggered scheme,the closed-loop system is established for the delayed neural network.A more advanced Lyapunov functional is constructed by making full use of the system information of delayed neural networks.Based on the Lyapunov stability theory.a stability criterion is given to ensure the closed-loop system is asymptotically stable,and the result is applied to the case without event-triggered scheme.Finally,simulation examples are given to illustrate the effectiveness of the results.(2)The decentralized event-triggered H∞ networked control for delayed neural networks is studied with stochastic cyber-attacks.Firstly,a new dynamic event-triggered scheme is introduced to monitor the transmission of sampled data,and two independent Bernoulli distribution variables are used to describe stochastic cyber-attacks.Secondly,based on the networked control,the closedloop system is constructed under stochastic cyber-attacks and limited network bandwidth.Then,using the Lyapunov-Krasovskii functional method,a less conservative sufficient condition is given to ensure the closed-loop system is mean-square asymptotically stable and achieves the prescribed H∞ performance.Based on the condition,the state feedback H∞ controller design method is given.Finally,two simulation examples are given to illustrate the effectiveness of the results.(3)The event-triggered quantized H∞ control for NCSs is studied subject to multi-sensor saturations and stochastic cyber-attacks.Firstly,a new decentralized event-triggered scheme is proposed to save system resources.Secondly,the closed-loop system is established under the decentralized event-triggered scheme,sensor saturation,quantization and stochastic cyber-attacks in a unified framework.Then,using the Lyapunov-Krasovskii functional method,a new sufficient condition is established to ensure that the closed-loop system is mean-square asymptotically stable and achieves the prescribed H∞ performance.Based on the condition,a new controller design method is proposed.Finally,the effectiveness of derived results is verified by two simulation examples.(4)The dynamic event-triggered H∞ filter analysis and design is studied for networked systems with stochastic cyber-attacks and quantization.First,the dynamic event-triggered scheme is used to determine whether the sampled data should be sent out or not.Second,by considering stochastic cyber-attacks and quantization,the filtering error system is established.Then,based on the Lyapunov-Krasovskii functional method,a sufficient condition is given such that the filtering error system is mean-square asymptotically stable and satisfies the prescribed H∞ performance.Based on the condition,a filter design method based on the linear matrix inequality is proposed by using the reversible linear transformation method.Finally,the effectiveness of obtained results is verified by two simulation examples.