Fixed-time Synchronization Research Of Two Types Of Variable-coefficient Fuzzy Neural Networks

There are a variety of synchronization types of neural network,and quite a lot of literatures have good results on the problem of system synchronization in limited time.However,there are few researches on fixed-time synchronization,and it has more practical value in the actual background.Therefore,this thesis carries out research on fixed-time synchronization of system.Its essence is to estimate the upper bound of the time consumed by the master-slave system synchronization,and the estimation formula obtained is only related to the parameters of the constructed controller.Therefore,we can pre-set the time spent for master-slave system synchronization by adjusting the parameters of the controller,and its application is more convenient.The main work of this thesis is summarized as follows:Firstly,the continuous or discontinuous activation functions of fuzzy cellular neural networks with time-varying coefficients and delays are introduced respectively.By using the relevant theories in Filippov framework,Lyapunov functional method and inequality scaling tools,the main conclusions related to fixed-time synchronization are obtained.Finally,the feasibility of the conclusions in this chapter is verified by two examples.Secondly,considering the fuzzy neutral-type inertial neural network with time-varying coefficients and delays,the second-order network is transformed into first-order network by reducing order method,and then using Lyapunov functional method and fixed-time stability theory,appropriate controllers are designed according to different conditions,and some criteria for master-slave system synchronization in limited time are obtained.Finally,two examples are used to verify the feasibility of the conclusions of this chapter.