作　　者： ; ; ;

机构地区： 同济大学土木工程学院土木工程防灾国家重点实验室

出　　处： 《振动工程学报》 2009年第2期156-161,共6页

摘　　要： 随机子空间算法在桥梁断面气动导数识别中表现出了良好的适应性,已能较好地识别出系统的频率和阻尼比,但在低风速条件下模态振型的识别精度尚无法令人满意。气动导数对模态振型相对变化的敏感性分析表明,模态振型对气动导数识别结果影响显著。在传统的基于输出协方差的随机子空间方法(SSI)基础上,引入一种新的稳定图,同时将频率、阻尼比和振型这3种模态参数的相对误差作为形成稳定轴的标准来获取气动导数。为验证该方法的可行性,进行了平板节段模型的数值仿真,结果表明该方法有助于提高模态振型的识别精度,进而提高气动导数的识别精度。 Stochastic subspace technique to identify flutter derivatives of bridge decks is numerically robust and able to give precise estimation of system frequencies and damping ratios, while it is not precise enough for determination of mode shapes in low-speed flow. Firstly, sensitivity of flutter derivatives of the thin plate section model to modal shape is confirmed. Secondly, in combination with a new kind of stabilization diagram which choose system frequencies, damping ratios and mode shapes as stability criterion pole, covariance-driven stochastic subspace identification （SSI） technique is proved to be effective to give a better estimation of mode shapes via a numerical example. The identified flutter derivatives are compared to both Theodorson analytical solutions of the ideal thin plate and results from traditional stochastic subspace identification method, and the precision of the present method is verified.

关 键 词： 桥梁断面 气动导数 随机子空间法 稳定图

领　　域： [建筑科学] [交通运输工程] [航空宇航科学与技术] [航空宇航科学技术]