基于FIR神经网络的热连轧机时滞动态模型辨识

TIMEDELAY DYNAMIC MODEL IDENTIFICATION ON TANDEM HOT ROLLING MILL WITH FIR NEURAL NETWORK

介绍一种FIR神经网络模型及其时变反传算法,并把它用于纯滞后非线性MIMO系统模型辨识;以带钢热连轧受控对象为例,建立2机架的小偏差模型,并进行了仿真;使用FIR神经网络模型对仿真数据进行了模型辨识,辨识结果显示了网络良好的性能。

An FIR neural network,is introduced with its temporal back propagation algorithm and its application to model identification of timedelay nonlinear MIMO system Simulation has been done on the basis of building an partial equation model of 2stands tandem millThe results of model identification is completed with simulation data from FIR neural network The results of identification shown that the network is suitable