用反馈神经网络对非线性离散系统进行实时建模

Modeling a nonlinear discrete-time system by recurrent neural network in real-time

利用时变反馈神经网络的概念及二维线性系统理论给出了非线性离散系统的一种实时建模方法 .理论及仿真实验结果均表明 ,这种实时模型能够任意逼近非线性离散系统 .

A real time modeling approach for nonlinear discrete time system is presented by using the concept of time variant recurrent neural network and the theory of two dimensional system. The theory and simulation results show that the derived mathematical model can approximate the nonlinear dynamics to any degree of accuracy.