| 自递归神经网络预测结构响应 |
| |
| 引用本文： | 李海岭,马亚丽,马智永.自递归神经网络预测结构响应[J].河南科学,2005,23(6):867-869. |
| |
| 作者姓名： | 李海岭 马亚丽 马智永 |
| |
| 作者单位： | 中国公路工程咨询监理总公司,北京,100101;北京工业大学建筑工程学院,北京,100022;郑州大学土木工程学院,河南,郑州,450002 |
| |
| 摘 要： | 推导了多输出自递归神经网络的学习算法,并基于Lyapunov函数得到保证网络快速收敛的自适应学习率.最后,应用此网络预测一3层建筑结构对于地震的响应.计算机仿真结果表明,网络学习算法的有效性,及此网络预测结构响应的可行性.
|
| 关 键 词： | 自递归神经网络 学习算法 学习率 结构响应 |
| 文章编号： | 1004-3918(2005)06-0867-03 |
| 收稿时间： | 2005-08-10 |
| 修稿时间： | 2005年8月10日 |
Prediction of structural reponses by self-recurrent neural network |
| |
| LI Hai-ling,MA Ya-li,MA Zhi-yong.Prediction of structural reponses by self-recurrent neural network[J].Henan Science,2005,23(6):867-869. |
| |
| Authors: | LI Hai-ling MA Ya-li MA Zhi-yong |
| |
| Institution: | 1. China Highway Engineering Consulting and Supervision General Corperation, Beijing 100101, China; 2. College of Civil and Arehitectere Engineer, Beijing University of Technology, Beijing 100022, China; 3. College of Civil Engineering, Zhengzhou University, Zhengzhou 450002, China |
| |
| Abstract: | Learning algorithms for self-recurrent neural network(SRNN) with multiple output nodes are deduced in this paper.And adaptive learning rates to guarantee convergence are developed based on Lyapunov function.Finally,SRNN is utilized to predicting the responses of a three-storey structure building subjected to seismic wave.Results of the computer simulation show the validity of learning algorithms and the feasibility of forecasting structural responses. |
| |
| Keywords: | self-recurrent neural network learning algorithm learning rate structural response |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
