Intermittent Synchronization Control For Several Delayed Complex Networks

Synchronization for complex networks(CNs)has become the major research topic owing to its wide applications in image processing,multi-agent systems,secure communication and other fields.It has been observed that time delays inevitably exist.In the study of synchronization for delayed complex networks(DCNs),such issuses as how to quantitatively describe the synchronization capability of networks and how the topological structure of networks affects the synchronization behavior are vital both in theory and in practice.Without the external forces,a network cannot achieve synchronization by itself.Developing some control strategies to drive the network synchronize is of great necessity.Intermittent control(IC)outperforms the continuous control in terms of saving energy and flexible operation.However,the theoretical analysis for adopting IC to study synchronization of DCNs is few.Hence,this thesis devotes to investigate the synchronization problem for DCNs via IC.By using new analysis tools,a series of synchronization criteria are established,and the IC can have a wider practical application.The main contributions are presented as follows.Firstly,the periodically IC for synchronization of CNs with mixed time delays is investigated.Differ from the previous works,by designing periodically IC strategy,some sufficient criteria ensusing synchronization for underlying systems is derived from the view of topology structure of network.The obtained results show that,the synchronization for CNs with mixed time delays can be achieved via periodically IC,and there is a relationship between the synchronization and the topology structure of network.Secondly,the periodically IC for synchronization of stochastic DCNs is considered.By exploiting graph theory,Lyapunov function method and stochastic analysis technique,some synchronization conditions are established under periodically IC.Furthermore,the obtained results are applied to the second-order Kuramoto oscillators,some synchronization conditions for second-order Kuramoto oscillators are presented.These results show that the DCNs under stochastic disturbance can be synchronized effectively by adjusting the control width of periodically IC.Next,the aperiodically IC for synchronization of stochastic Markov switching CNs is studied.Considering that the internal delay and coupling delay are not the same and independent of each other.Based on aperiodically IC,some sufficient conditions ensuring pth-moment exponential synchronization are derived with the help of M-matrix theory and graph theory.These conditions are less conservative since the differentiability of time delay function is not required,and the restriction between control width and time delay is also removed.It is shown that,without additional restrictive conditions on time delay,the network can achieve synchronization when the control width and the transition rates(TRs)of Markov chain meets certain conditions.Then,the aperiodically IC for synchronization of multi-weighted CNs with Markov switching and reaction diffusion effects is discussed.A aperiodically IC strategy is proposed according to the concept of master-slave system synchronization.By the technique of network split,the synchronization of multi-weighted CNs with Markov switching and reaction diffusion effects is analyzed,and the corresponding criteria are given as well.Based on the obtained theoretical results,some conditions ensuring the synchronization of Cohen-Grosser neural networks are derived.The obtained results suggest that the control strengths and coupling strengths of multiple sub-networks have affect on synchronization property,but it need not to require that all the sub-networks are strongly connected.Finally,aperiodically adaptive IC for the synchronization of Semi-Markov switching CNs is concerned.Based on aperiodically IC,an adaptive scheme is introduced which can adjust the control strength on basis of the synchronization process.The synchronization conditions for Semi-Markov switching CNs under aperiodically intermittent adaptive control(Ap IAC)are established via the assumption that the dwell time of Semi-Markov switching obeys Weibull distribution and the uncertain TRs are bounded.These results show that by Ap IAC,the synchronization of Semi-Markov switching DCNs can be effectively realized in case of the unknown TRs,moreover,the synchronization conditions are closely related to the maximum uncontrolled ratio of IC and the bounds of TRs.In this thesis,therefore,based on periodically IC,the synchronization issues of CNs with mixed time delays and stochastic DCNs are addressed.In addition,the synchronization criteria of stochastic Markov switching CNs and multi-weighted CNs,as well as Semi-Markov switching CNs are,respectively,obtained via aperiodically IC.