Several Types Of Synchronization Studies On Multiple Neural Networks

As a kind of neural network,multiple neural network has attracted more and more scholars' research in recent years because of its unique characteristics.We can see the figure of multi neural network in the migration of fish swarm,unmanned aerial vehicle,computer group,etc.,but there should be more application space for multiple neural network,but this must be based on theoretical research,so it is necessary and meaningful to further study.In this paper,we mainly study the synchronization behavior of several kinds of multiple neural networks.By using Lyapunov-Razumikhin method,LMI,differential inclusion theory and reasonable inequality scaling techniques,the corresponding synchronization criteria are obtained according to the characteristics of multiple neural networks.The main work of this paper is summarized as follows:The input-to-state synchronization of a class of nonlinear coupled neural networks with time-delay under the action of observer based event triggered impulsive controller is studied.By using Lyapunov-Razumikhin method and LMI,an appropriate augmented system based on observation is constructed.When the true real value of the system is unknown,the criterion required for the original system to converge and synchronize is obtained.The effectiveness of the theory is verified by an example.The synchronization of a class of multiple neural networks with time-varying delays under the action of event-triggered impulsive coupling controller is studied.By constructing an observer based augmented system and a suitable Lyapunov function,the sufficient conditions for the synchronization of the original system are obtained under the action of the event-triggered impulsive coupling controller.A numerical example is given to verify the effectiveness of the theoretical results.In this paper,a class of bidirection multiple neural networks with impulsive coupling control is studied.The general single-layer multiple neural networks are extended to the two-layer bidirection multiple neural networks.The impulsive coupling control is adopted.The necessary criteria for synchronization of this kind of multiple neural networks are obtained by Lyapunov-Razumikhin method and convex hull theory.A numerical example is given to prove the effectiveness of the theory.