Dynamic Analysis Of Several Classes Of Discrete Neural Network Models

Recently, some papers are concerned with a class of discrete-time neural network with three neurons, but they almost concerned a class of discrete neural network with three neurons on a line.In this thesis, by using of the Lyapunov function. Lasalles invariance principle and the analytical technique to construct return mapping. We describe some important properties of the dynamic behaviors of several classes of discrete-time networks model with three-cell cellular neural and a class of delay discrete-time networks model with two-cell cellular neural, which includes stability, existence of periodic solutions. It is composed of four chapters.In chapter 1, the history and development of neural networks are reviewed briefly. The background of the model of cellular neural networks discussed in this thesis is interpreted and the problems to be resolved, the purpose and innovation of the investigation are denoted.In chapter 2, a class of discrete-time model of neural networks with three-cell cellular on a line and with symmetric template are investigated. We obtain some results about the stability, the existence of 2-periodic solutions.In chapter 3, a class of network with a ring of three identical neurons which has a symmetric template are investigated. We obtain some results about the existence of 2-periodic solutions.In chapter 4, a class of delay discrete-time two-cell cellular networks which has a disproportion message function. We obtain some results about convergence of the system when the threshold is different values.