| 动态神经网络学习算法中的参数估值 |
| |
| 引用本文： | 闫纪红,王子才,史小平.动态神经网络学习算法中的参数估值[J].系统工程理论与实践,1999,19(7):26-29. |
| |
| 作者姓名： | 闫纪红 王子才 史小平 |
| |
| 作者单位： | 哈尔滨工业大学仿真中心 |
| |
| 摘 要： | 以一种简单的动态ＢＰ网络作为并联模型，运用卡尔曼滤波原理，提出了一种新的神经网络辨识算法．该算法的学习速度是由带时间参数的Ｒｉｃｃａｔｉ方程来确定的，从而大大减少了学习的迭代次数．仿真结果表明此算法是有效可行的
|
| 关 键 词： | 辨识 动态BP网络 卡尔曼滤波 学习算法 |
| 修稿时间： | 1997-12-22 |
Parametric Estimation in Learning Algorithm for Dynamic Neural Networks |
| |
| YAN Jihong,WANG Zicai,SHI Xiaoping.Parametric Estimation in Learning Algorithm for Dynamic Neural Networks[J].Systems Engineering —Theory & Practice,1999,19(7):26-29. |
| |
| Authors: | YAN Jihong WANG Zicai SHI Xiaoping |
| |
| Institution: | Simulation Centre, Harbin Institute of Technology |
| |
| Abstract: | Abstract In this paper, a simple and convenient dynamic neural network is used as parallel model.Making use of Kalman filtering, a new neural network based algorithm for system identification is presented. To reduce the iteration number of times, the learning velocity of this algorithm is computed by Riccati difference equation. Simulation results show that the algorithm is feasible and effective. |
| |
| Keywords: | |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
| 点击此处可从《系统工程理论与实践》浏览原始摘要信息 |
| 点击此处可从《系统工程理论与实践》下载免费的PDF全文 |
