基于瞬时最优控制神经网络的建筑结构主动控制研究

英文篇名：RESEARCH ON ACTIVE CONTROL OF ARCHITECTURE CONSTRUCTION BASED ON INSTANTANEOUS OPTIMAL CONTROL NEURAL NETWORK
中文刊名：振动与冲击
英文刊名：Journal of Vibration and Shock
作者：黄永安 ; 邓子辰
英文作者：Huang Yongan 1 Deng Zichen 1 ; 2 (1.Department of Engineering Mechanics ; Northwestern Polytechnical University ; Xi'an 710072 ; 2.State Key Laboratory of Structural Analysis of Industrial Equipment ; Dalian University of Technology ; Dalian 116023)
中文关键词：人工神经网络 ; 瞬时最优控制 ; 主动控制
英文关键词：artificial neural networks ; instantaneous optimal control ; active control
出版日期：2005-04-30
机构：西北工业大学工程力学系,西北工业大学工程力学系西安710072 ,西安710072,大连理工大学工业装备结构分析国家重点实验室,大连116024
年：2005
期：02
出版单位：振动与冲击

摘要

改进一种基于瞬时最优控制的神经网络训练算法。本方法以瞬间最优控制价值函数最小化为训练目标,考虑了地震输入的能量,利用最速下降梯度法计算权值的改变量,并对敏感度矩阵进行近似处理,可解决神经网络控制中神经网络控制器难以获得的训练输入/输出样本对的难题。该方法适合多输入/多输出结构体系,整个推导过程都是针对此体系进行的。文中通过对一个三层框架结构体系进行有效的仿真计算,说明了算法的有效性。
The neural network training algorithm is improved, which is based on instantaneous optimal control method, taking into consideration of the energy of earthquake. The training object of the algorithm is minimization of the cost function in optimal control, variations of the weight values are obtained by the steepest descend gradient method and moreover an approximate treatment is made on the sensitivity matrix evaluation. The method can overcome the difficulties of obtaining the training input/output samples for neuro-controller in neural networks control. The algorithm can be used in MIMO system, and the simulation of active control of a three-story frame structure subjected to ground excitation shows that the proposed method is effective.

引文

1　吴建军,何玉傲.结构振动控制中神经网络应用的新进展.世界地震工程,2000,16(1)11:10—19
    2　HousnerGW,BergmanLA,etal.StructureControl:Past,Present,andFuture.JournalofEngineeringMechanics,ASCE,1997,123(9):879—971
    3　GhaboussiJ,JoghataieA.ActiveControlofStructuresusingNeuralNetworks.JournalofEngineeringMechanics,ASCE,1995,121(4):555—567
    4　Ju-TaeKim,HyungJoJung,In-WonLee.OptimalStructur alControlUsingNeuralNetworks,JournalofEngineeringMe chanics,ASCE,2000,126(2):201—205
    5　HoshiyaM,SaitoY.PredictionControlofSDOFSystem.Jour nalofEngineeringMechanics,ASCE,1995,121(10):1049—1055
    6　SatoT,TokiK,HashimotoM.(1991).PredicativeControlofSeismicResponseofStructurewithTimeDelay,Proc,JapanSocofCivilEng.,Tokyo.Japan,No.428:193—202(inJa pan)
    7　何玉彬,李新忠.神经网络控制技术及其应用.北京:科学出版社,2000
    8　Shin-LinHung,CYKao,JCLee.ActivePulseStructureControlUsingArtificialNeuralNetworks.JournalofEngineer ingMechanics,ASCE,2000,126(8):839—849
    9　王永骥,涂　健.神经元网络控制.北京:机械工业出版社,1999