This article focuses on the global exponential stabilization of a neural network problems and global exponential stability of a class of neutral systems and three types of neural networks,. Based on the Lyapunov-Krasovskii functional and linear matrix inequalities (LMI), obtained the conditions for the establishment of globally exponentially stable or globally exponentially stable.Chapter1is an introduction, mainly on the proposed neural network and the meaning of stability, global exponential stability of the neutral model results, the stability theory of dynamical systems, and describes the main work of this paper.Chapter2, we cleverly construct a new Lyapunov-Krasovskii function, with random items leap God cellular neural network model of neural network globally exponentially stable, with random jumping God Neural Network for the global exponential stability of the criterion of cellular neural networks. Finally, numerical simulation, we use MATLAB LMI toolbox with two examples of the paper the results of the feasibility.Chapter3, based on linear matrix inequalities, Lyapunov-Krasovskii stability theory, to study the stability of the neutral model neural network Global Exponential stability criterion. Finally with the help of MATLAB toolbox use numerical simulation shows the effectiveness of the proposed criterion.Chapter4deals with distributed delay discrete neural network. Based on the Lyapunov-Krasovskii function and linear matrix inequality (LMI) technology, application of Halanay inequality and nonlinear measure two methods we have distributed delay discrete neural network system is globally asymptotically stable in order to achieve a stable neural network.Chapter5, based on the Lyapunov stability theory, with the aid of homeomorphism mapping theorem and linear matrix inequality proved cellular neural networks equilibrium point of the existence and uniqueness and stability, and the use of MATLAB numerical examples verify the validity of the conclusion of chapter.Chapter6, based on the Lyapunov stability theory, using linear matrix inequality analysis techniques, puts forward the C-G stochastic neural networks with discrete and distributed delays adequacy criterion, and using LMI toolbox to verify this chapter conclusion validity. |