论文部分内容阅读
主要讨论反应扩散递归神经阿络全局指数稳定的鲁棒分析.给定反应扩散递归神经网络是全局指数稳定,首先,在此神经网络基础上考虑噪音扰动,利用超越方程得到噪音密度的上界,在上界范围内,带噪音的反应扩散递归神经网络仍然是全局指数稳定.进一步,在反应扩散递归神经网络基础上同时考虑噪音扰动和连接权参数不确定,利用超越方程得到连接权参数和噪音密度上界,在两个参数描述的超越方程范围内,带噪音和连接权参数不确定的反应扩散递归神经网络仍然是全局指数稳定.最后给出数值算例证实相关理论的有效性.
In this paper, the robust exponential stability of recursive neural network with reaction-recursive neural networks is mainly discussed. Given a globally exponential stability of recursive neural network with reactive diffusion, we first consider the noise disturbance based on this neural network and use the transcendental equation to obtain the upper bound of noise density. In the upper bound, the recursive neural networks with noisy reaction-diffusion are still global exponential stability.Furthermore, based on the reaction-diffusion recurrent neural network, the parameters of noise disturbance and connection weight are considered simultaneously, and the connection parameters and noise are obtained by using the transcendental equation Density upper bound, recursive diffusion-recursive neural networks with unknown parameters of noise and connection weights are still globally exponential stable over the range of transcendental equations described by the two parameters. Finally, numerical examples are given to demonstrate the validity of the relevant theory.