不同控制策略下安装磁流变阻尼器的模型结构振动台试验与分析

本文对不同控制策略下安装有磁流变(MR)阻尼器的模型结构进行了振动台试验和分析。文中首先介绍了经典最优控制(COC)、瞬时最优控制(IOC)和线性二次高斯最优控制(LQG)等三种控制算法,然后对一首层安装有最近设计制造的MRF-04K型的MR阻尼器的、1/3比例的三层钢框架模型,进行了在两种被动控制和三种半主动控制等不同控制策略下的振动台试验,最后对模型结构地震反应的控制效果以及不同控制策略对控制效果的影响和控制稳定性进行了分析。研究表明,安装有MRF-04K阻尼器的结构控制系统具有良好的控制效果,无论是被动控制还是半主动控制,模型结构各层相对位移峰值均减小了45%左右,其均方根值均减小了70%左右,加速度反应峰值均减小了30%左右,其均方根值均减小了75%左右,从而验证了MRF-04K阻尼器是结构控制工程应用的一种理想的控制装置;研究还表明,在三种半主动控制策略中,基于LQG算法的半主动控制仅需模型结构的加速度反应的反馈信息,比基于IOC算法和COC算法的半主动控制

Shaking table test and analysis of model structure installed magnetorheological damper under different control strategies

The shaking table test and analysis of a model structure installed a magnetorheological (MR) damper were performed under different control strategies. Three control algorithms including classical optimal control (COC), instantaneous optimal control (IOC) and linear quadratic GAUSS control (LQG) are first introduced. Then, the shaking table test of a one third scale steel frame model installed a recently designed and manufactured MRF 04K damper was carried out under the control strategies of two kinds of passive controls and three kinds of semi active controls. Finally, the control effect of the seismic responses of the model structure, the influence of the different control strategies on the control effect and the stability of the control were investigated. The results showed that the control system of structure installed MRF 04K damper has perfect control effect, no matter whether it is passive control or semi active control, the peak values of relative displacements of the model structure are reduced by 45%with its root mean square values decreased by 70%, and the peak values of the accelerations are reduced by 30%with its root mean square values decreased by 75%. This demonstrates that the MRF 04K damper is an ideal control device for engineering application of structural control. The results also showed that within the three kinds of semi active control strategies, the semi active control based on the LQG algorithm only needs the feedback information of acceleration responses of the model structure, so it is more economic and easier to be realized than the semi active control based on the COC and IOC algorithms. Therefore, LQG algorithm is an optimum control algorithm for implementation of semi active control using MR damper.