BP网络结构和算法对非线性系统进行辨识  

BP Network Structure and Algorithm for Nonlinear System Identification

作　　者：丁伟[1] 陈羽鲲 Ding Wei;Chen Yukun(Chonfar Engineering and Technology Co., Ltd.,Shenyang 110026,China)

机构地区：[1]化工部长沙设计研究院沈阳分院,辽宁沈阳110026

出　　处：《山东化工》2019年第2期137-139,共3页Shandong Chemical Industry

摘　　要：利用MATLAB神经网络工具箱进行了基于BP神经网络的非线性系统辨识仿真研究,针对实际问题选取BP网络结构和BP算法,并对网络输入、输出、隐层节点的个数选取作了探讨,选取一个非线性系统,通过改变隐层节点数和训练函数,找出适当的BP网络结构和算法对非线性系统进行辨识。The nonlinear system identification simulation based on BP neural network is carried out by using MATLAB neural network toolbox.BP network structure and BP algorithm are selected for practical problems.The selection of network input,output and hidden layer nodes is discussed.A nonlinear system is selected by changing the number of hidden layer nodes and training function.We find out the suitable BP network structure and algorithm to identify the nonlinear system.

关 键 词：神经网络 系统辨识 BP网络 BP算法 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]