遗传BP神经网络主动AMD对偏心结构的减震控制
详细信息 本馆镜像全文    |  推荐本文 | | 获取馆网全文
摘要
本文采用在结构顶层水平双向设置AMD主动控制装置的方法,对偏心结构在多向地震作用下的振动进行了遗传BP神经网络主动控制研究。首先论述了遗传BP神经网络的基本理论,然后介绍结构-AMD扭转耦联控制系统微分方程及其状态空间求解方法,最后介绍利用遗传BP神经网络对结构反应进行预测及对结构实施主动控制的方法。数值结果表明,利用遗传BP网络进行结构反应预测和控制是可行且有效的,能对结构的平动和扭转反应起到较好的减震效果。
In this paper, the control approach to irregular building excited by multidimensional earthquake ground motions is presented by using active massed damper (AMD) with the combined method of the genetic algorithm (GA) back propagation (BP) neural network. The basic theory of GABP neural network is firstly introduced. Then, the mechanical model of a multistorey irregular structureAMD system is established and its equation of motion is given. The method of GABP neural network is used to predict the structural response and reduce the earthqake action with AMD. The numerical examples have shown that it is an effective method presented here to control the translational response and rotational response of irregular buildings.
引文
[1] 孙增圻,等.智能控制理论与技术[M].北京:清华大学出版社,1997.
    [2] 王耀南.智能控制系统[M].长沙:湖南大学出版社,1996.
    [3] 周其鉴,等.智能控制及其展望[J].信息与控制,1987(2):38-45.
    [4] 闫石.结构振动智能控制的人工神经网络与模糊逻辑方法研究[D].大连理工大学土木建筑工程学院博士学位论文.2000.
    [5] GhaboussiJ.andJoghataieA.ActiveControlforStructuresUsingNeuralNetworks[J].ASCEJ.ofEngrg.Mech.,1995,121(4):555-567.

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