Stabilization And Synchronization For Two Classes Of Delayed Neural Networks

Over the past twenty years, because of the important practical applications of the delayed neural networks in pattern recognition, signal processing, artificial intelligence, digital image processing, global optimization as well as data mining, the dynamical problems of the delayed neural networks have attracted the worldwide attention and great research interest. Based on the existing results, this thesis mainly focuses on the stabilization and synchronization for two classes of delayed neural networks(namely, chaotic delayed neural networks and delayed BAM neural networks). A series of significant results will be derived, which may enrich and improve the existing results to a certain extent. Specifically, the main contents and achievements of this thesis are as follows: 1. Complete synchronization of chaotic delayed neural networks via a new type of intermittent control.We introduce a new type of intermittent control which has two switches in a control period and apply it to complete synchronization of chaotic delayed neural networks. Through constructing a suitable Lyapunov function and making use of some inequality techniques, we obtain the sufficient conditions for complete synchronization of chaotic delayed neural networks. The computer simulations of two numerical examples are also given to demonstrate the effectiveness of the obtained theoretical results. 2. Stabilization and synchronization of delayed BAM neural networks via a matrix measure approachBased on the analysis on the current relevant results, different from the traditional method of Lyapunov function, a new approach which is called matrix measure is presented and applied to the analysis of stabilization and synchronization for delayed BAM neural networks. Several criteria for the global exponential stability of the equilibrium point of delayed BAM neural networks and several sufficient conditions for the global exponential synchronization of drive-response delayed BAM neural networks are derived by using state feedback control. We also present the computer simulations of a numerical example to illustrate the effectiveness of the obtained theoretical results and turn it out to be more practical and effective compared with the previous results of pertinent literature.