分布时滞动态神经网络的全局指数稳定性

Globally Exponential Stability of Dynamic Neural Networks with Distributed Delays

在没有假定激励函数有界、可微的情况下,研究包含分布时滞的动态神经网络平衡点的存在性、唯一性和全局指数稳定性.根据M-矩阵和拓扑学的有关知识,以及李雅普诺夫稳定性理论,获得该类神经网络平衡点的存在性、唯一性及其全局指数稳定的充分判据.用神经网络的权值矩阵和激励函数满足的条件构造判定矩阵.如果判定矩阵为M-矩阵,则系统存在唯一平衡点,是全局指数稳定的.

The existence, uniqueness and globally exponential stability of the equilibrium point of a dynamic neural network with distributed delays were studied without assumption of boundedness and differentiability of activation functions. Sufficient criteria for existence, uniqueness and global exponential stability of the equilibrium point of such neural networks were obtained based on the knowledge of M-matrix, to and Liapunov stability theory. A test matrix was constructed by the weight matrix and the conditions satisfying activation functions of the neural networks. A neural network has a unique equilibrium point and is globally exponential stable if the test matrix is an M-matrix.