Event-triggered Synchronization Of Complex Dynamic Networks

Complex dynamic network is a kind of large-scale network model which is composed of many independent nodes connected to each other.A?ected by the dynamics of individual nodes and the coupling characteristics of the network,complex network has a variety of di?erent evolution and dynamic properties.Network synchronization,as one of the most important clustering behaviors in network system,is a state that achieved through its own dynamic evolution and the coupling between nodes,or even through external control.In the study of synchronous control in complex dynamic networks,although fruitful results have been achieved for some classical models,there are still some weak areas waiting for breakthroughs in the face of the rapid development of science and technology.For example,as complex network systems are widely used in real production,factors such as delay,network attack,data loss,quantization and system failure in network communication are not uniform,which puts more demanding requirements on the synchronization control of the system.It is worthwhile to design more e cient,stable and energy-e cient synchronous control schemes for complex dynamic networks with communication constraints such as system time delays,controller failures and network attacks.In this thesis,we address the above issues with the following main work:The first chapter mainly introduces the background knowledge and research status of complex dynamic network,synchronization control strategy and network security control.In the second chapter,a class of complex dynamic network synchronization probvi lem with time delay and controller faults are studied.Firstly,we establish a kind of complex dynamic network model with system time delay and controller failure,and a fault-tolerant control system based on dynamic event-triggered control is proposed.By constructing internal variables to design a class of dynamical event-triggered mechanisms,we can e?ectively reduce the number of triggers in the system network communication resources.Then,by using Lyapunov stability theory,some su cient conditions for complex dynamic networks to achieve global exponential synchronization are given.It is rigorously illustrated that the minimum trigger interval of the system under the event-triggered mechanism is more than a positive number,which demonstrates that the system does not produce Zeno behavior.Finally,the reliability of the results is verified by simulation examples.In the third chapter,the security control of a class of complex dynamic networks subjected to random network attacks is studied.Firstly,a kind of complex network model with spoofing attacks is established,which takes into account that the attacker can launch malicious attacks on the system at any time,and the attack signal strength changes in real-time.Secondly,based on the spoofing attack model,to reduce unnecessary data transmission and reduce the probability of malicious attacks,a combination of dynamical event-triggered control and impulsive control method is adopted.Then according to Lyapunov stability theory and linear matrix inequality technique,the sufficient conditions for the system to reach the mean-square bounded synchronization are given,and it is verified that the system can avoid Zeno phenomenon under the eventtriggered mechanism.Finally,the validity of the conclusions is verified by numerical simulation.The fourth chapter summarizes the main work of this thesis and discusses the current deficiencies and future research directions.