The Research On Stability Of Neutral Stochastic Neural Networks

The study of neural network stability has been for many years and has made many striking achievements so far.Neural network can be used to solve many problems,so it is great significant to study the stability of neural networks.But in real life,the uncertainty factors exist objectively,using deterministic methods to describe uncertain systems may result in a great deviation.Therefore,in the study of nondeterminacy systems,nondeterminacy when considering the uncertainty have important significance.Uncertainty factors include randomness,fuzziness,unconsistency,instability,etc.In this thesis,we study the stability of two classes of stochastic neural networks with time delay.The first chapter introduces the development background and the general situation of the neural network and some basic theories.Finally,it introduces the research work of this paper.In the second chapter,by using the linear matrix inequality,Lyapunov function and martingale convergence theorem,the p order ?? stability and almost sure?stability of stochastic neutral Cohen-Grossberg neural networks with time-varying delays are obtained,spreaded several the existing exponential stability results.In the third chapter,we use the linear matrix inequality,Lyapunov function and martingale convergence theorem to study the p order ?? stability and almost sure?stability of fuzzy cellular neural networks with time-varying delays and neutral terms.,spreaded some known exponential stability results.Finally,we summarizes the work of the thesis with prospect for the future research.