时滞系统神经网络辨识方法研究

A Research on the Method to Identify the Neural Network with Time Delay System

提出基于神经网络的时滞非线性系统的辨识方法,通过一种带有时滞的神经网络进行在线辨识,神经网络的权值矩阵在线自适应调整,并通过Lyapunov-Krasovskii方法验证了算法的收敛性,表明辨识算法可以对任意的非线性时滞系统进行有效辨识。

This paper proposes the identification method basing on the neural network with the nonlinear time delay system.This method is to make an online identification through the neural system with time dealay,and the online self adjustment of the weight number matrix of the neural network,and to verify the convergence of the calculating method.Consequently it shows that the identification method is valid for any nonlinear time delay system.