一类二元时滞神经网络模型解的定性研究

Qualitative Analysis of Solutions to a Class of Two-Neuron Network

对一类具有自反馈信号的二元时滞神经网络模型的解进行定性研究,针对网络参数的不同取值,分步地求解一个在一定初始函数空间中给定初始值的泛函微分方程组,再讨论模型解的收敛性.我们的分析表明网络参数在神经网络模型研究中具有极其重要的作用.

Qualitative analysis is made on solutions to a class of two - neuron network model with delayed self - feedback signal. Referring to some different values of the network parameters, the step method is adopted to solve functional differential equations with a given initial value in a certain initial function space. The convergence of the solutions to the model is then studied. It is shown through the analysis that the network parameters play vital roles in the research of the neural network model.