| 迟滞非线性系统的神经网络建模 |
| |
| 引用本文： | 刘向东,修春波,李黎,刘承.迟滞非线性系统的神经网络建模[J].压电与声光,2007,29(1):106-108. |
| |
| 作者姓名： | 刘向东 修春波 李黎 刘承 |
| |
| 作者单位： | 1. 北京理工大学,自动控制系,北京,100081
2. 天津工业大学,自动化系,天津,300160 |
| |
| 摘 要： | 提出了一种利用神经网络对迟滞非线性系统的建模方法。网络分三层结构,隐层神经元的激励函数采用两个移位的Sigmoid函数组成的迟滞环构成,根据非线性系统的特点,给出了网络结构的确定方法。网络的学习算法采用传统的梯度下降学习算法。利用该网络对压电陶瓷迟滞特性进行了建模实验,实验结果证明了该方法的有效性。
|
| 关 键 词： | 压电陶瓷 神经网络 建模 迟滞 |
| 文章编号： | 1004-2474(2007)01-0106-03 |
| 修稿时间： | 2005-10-12 |
Hysteresis Modeling Using Neural Networks |
| |
| LIU Xiang-dong,XIU Chun-bo,LI LI,LIU Cheng.Hysteresis Modeling Using Neural Networks[J].Piezoelectrics & Acoustooptics,2007,29(1):106-108. |
| |
| Authors: | LIU Xiang-dong XIU Chun-bo LI LI LIU Cheng |
| |
| Affiliation: | 1. Dept. of Automatic Control Beijing Institute of Technology, Beijing 100081,China; 2. Dept. of Automatic Control , .Tianjin Polytechnic University, Tianjin 300160, China |
| |
| Abstract: | A novel neural networks for hysteresis modeling is proposed.The networks has three layers.The activation function of the neuron in the hidden layer is composed by two displacement Sigmoid function.According to the characteristic of the nonlinear systems,the paper give the choice method of structure and learning method.The modeling experiment of the piezoelectric ceramic is realized by the network.The simulation results prove validity of the algorithm. |
| |
| Keywords: | piezoelectric ceramic neural networks modeling hysteresis |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
