Stability Analysis For Two Classes Of Neural Networks With Mixed Time-delays

This paper studies the stability for two classes of the neural networks: one is about the stochastically asymptotic stability problem for a class of neutral-type neural networks with mixed time-delays and Markovian jumping parameters under impulsive perturbations in the mean square,and the other is novel delay-dependent stability criteria for fuzzy neural networks using delay-decomposition approach.Based on the virtue of the Lyapunov functional method,by effective combination of Jensen integral inequality with reciprocally convex and quadratic convex combination method,several novel sufficient conditions are derived to ensure the asymptotic stability of the considered neural networks.The main contents are as follows:The first chapter is an introduction,which mainly introduces the research achievement of in recent years and significance of neural network stability.Then we give the stability theory of dynamical systems.Finally,it gives a brief account of the main work in this paper.The second chapter,combined with some methods of the free weight matrix and linear matrix inequalities,we have applied the Lyapunov stability theory to make a further study of the stability of considered neural network model.By constructing a new Lyapunov-krasovkii functional,we propose a sufficient condition of considered neural network with global asymptotic stability.Finally,with the help of the MATLAB we give two numerical simulations to verify the feasibility and correctness of the conclusion.The third chapter,based on Lyapunov stability theory,a set of new Lyapunov function is constructed,and a criterion for the stability of fuzzy neural networks with mixed time-varying delays is proposed.With the using of LMI tools of MATLAB,we verify the validity of the conclusion of this chapter.