摘要
介绍了第3代结构风振控制基准问题的定义。通过观测部分楼层加速度和控制力输出,建立了模糊神经网络控制器,解决了传统控制中有限的传感器数目对系统振动状态估计的困难;利用模糊神经网络预测结构的控制行为,消除了闭环控制系统中存在的时滞;通过模糊神经网络控制器的学习功能,解决了土木工程复杂结构模糊控制中难以依据专家的主观经验来确定模糊控制规则和语言变量隶属函数等困难。以风振控制的基准问题为研究对象,编制了程序对受控系统进行数值仿真分析。分析表明,模糊神经网络控制策略能有效地抑制高层建筑的风振反应。
The definition of a benchmark control problem for wind-induced tall buildings was presented. A fuzzy neural network controller, in which few sensors and no observer were needed, was designed on the basis of the measurement of floor accelerations and control forces. The fuzzy neural network controller to predict control action eliminated the effects of time delay in the control loop. With a fuzzy logic control strategy, it＇s difficult to obtain an appropriate set of rules and membership functions. But it＇s easy to solve these problems utilizing the ability of adaptive learning of the fuzzy neural network controller. Numerical simulation was carried out for analyzing the wind-induced responses of the benchmark building. The analysis results show that the control strategy using fuzzy neural network has good performance in reducing across wind response of the benchmark building.
出处
《地震工程与工程振动》
CSCD
北大核心
2007年第6期211-217,共7页
Earthquake Engineering and Engineering Dynamics
基金
福建省青年科技人才创新项目(No.2006F3008)
福建省教育厅基金项目(No.JA06027)
关键词
基准问题
高层建筑
风振反应
模糊神经网络控制
benchmark problem
tall buildings
wind-induced response
fuzzy neural network control
