应用自递归神经网络(SRNN)预测结构响应
|
摘要
本文提出了一种新的自递归神经网络结构.这种网络结构由全递归网络改造而成,只有一个隐层,而且隐单元仅存在自递归.研究了这种网络的学习算法.为了保证快速学习收敛,应用Lyapunov函数得到一种自适应学习率方法.用这种方法对一两层建筑结构响应进行在线预测.计算机仿真结果表明,这种网络学习算法是有效的,并且是可行的
A new neural paradigm called Self-Recurrent Neural Network (SRNN) is presented here. The architecture of SRNN is a modified model of the fully connected recurrent neural network with one hidden layer, and the hidden layer is comprised of self-recurrent neurons. A generalized dynamic back-propagation algorithm (DBF) is established. To guarantee convergence and faster learning, an approach using adaptive learning rates is developed by introducing a Lyapunov function. Convergence theorem for the adaptive back-propagation algorithm is developed for on-line prediction of the response of a two-story building excited by external force. Results form computer simulation studies demonstrate that the new DBP is valid and feasible in on-line predicting structural response.
A new neural paradigm called Self-Recurrent Neural Network (SRNN) is presented here. The architecture of SRNN is a modified model of the fully connected recurrent neural network with one hidden layer, and the hidden layer is comprised of self-recurrent neurons. A generalized dynamic back-propagation algorithm (DBF) is established. To guarantee convergence and faster learning, an approach using adaptive learning rates is developed by introducing a Lyapunov function. Convergence theorem for the adaptive back-propagation algorithm is developed for on-line prediction of the response of a two-story building excited by external force. Results form computer simulation studies demonstrate that the new DBP is valid and feasible in on-line predicting structural response.
引文
[1]Kobri,T.,Koshika,N.,Yamada,K.Ikeda,Y.:Seismic Response Controlled Structure with Active Mass Drive System,Part Ⅰ:Design,Part Ⅱ:Verfication,Earthquake Engrg and Struct.Dynamics,Vol.20,1991,133-166.
[2]Chang,I.C.H.,Soong,T.T.:Structural Control Using Active Tuned Mass Damper,J.of Engrg.Mech.ASCE Vol.106,No.6,1980,1091-1098.
[3]Roorda,J.:Tendon Control in Tall Structures,J.of struct.ASCE VoL.101,Mar.1975,505-521.
[4]Chung,L.L.,Lin,R.C.,Soong,T.T.,Reinhorn,A.M.:Experimental Study of Active Control of MDOF Structures under Seismic Excitations,J.of Engrg.Mech.ASCE Vol.1l5,No.8,l989,l609-l627.
[5]Yang,J.N.,Akbarpour,A.,Ghaemmaghamic,P.:New Optimal Control Algorithm for Structural Control,J.of Engrg.Mech.Vol.113,1987,1369-1386.
[6]He Yu’ao,Hu Xianzhong,Zhan Sheng:Predicting Seisic Response of Structures by Artificial Neural Network,transactions of Tianjin Univereity Vol.2,No.2,1996,36-39.
[7]徐植信、胡再龙:《结构地震反应分析》,高等教育出版社,1993年.
[8]Marios M.Polycarpou,Petros A.Loannou:Learning and Convergence Analysis of Neural Type Structural Networks,IEEE Trans.on Neural Network Vol.3,No.1,1992,39-50.
[9]何玉敖、吴建军:预测结构动态响应的自递归神经网络(SRNN)方法,第二届全国建筑振动学术会议论文集,中国建筑工业出版社,1997年10月.
[10]何玉敖、吴建军:线性结构振动的人工神经网络控制,第二届全国建筑振动学术会议论文集,中国建筑工业出版社,1997年10月.
[2]Chang,I.C.H.,Soong,T.T.:Structural Control Using Active Tuned Mass Damper,J.of Engrg.Mech.ASCE Vol.106,No.6,1980,1091-1098.
[3]Roorda,J.:Tendon Control in Tall Structures,J.of struct.ASCE VoL.101,Mar.1975,505-521.
[4]Chung,L.L.,Lin,R.C.,Soong,T.T.,Reinhorn,A.M.:Experimental Study of Active Control of MDOF Structures under Seismic Excitations,J.of Engrg.Mech.ASCE Vol.1l5,No.8,l989,l609-l627.
[5]Yang,J.N.,Akbarpour,A.,Ghaemmaghamic,P.:New Optimal Control Algorithm for Structural Control,J.of Engrg.Mech.Vol.113,1987,1369-1386.
[6]He Yu’ao,Hu Xianzhong,Zhan Sheng:Predicting Seisic Response of Structures by Artificial Neural Network,transactions of Tianjin Univereity Vol.2,No.2,1996,36-39.
[7]徐植信、胡再龙:《结构地震反应分析》,高等教育出版社,1993年.
[8]Marios M.Polycarpou,Petros A.Loannou:Learning and Convergence Analysis of Neural Type Structural Networks,IEEE Trans.on Neural Network Vol.3,No.1,1992,39-50.
[9]何玉敖、吴建军:预测结构动态响应的自递归神经网络(SRNN)方法,第二届全国建筑振动学术会议论文集,中国建筑工业出版社,1997年10月.
[10]何玉敖、吴建军:线性结构振动的人工神经网络控制,第二届全国建筑振动学术会议论文集,中国建筑工业出版社,1997年10月.
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心