基于MIMO信号降噪的模态参数识别研究  

Modal parameters identification based on denoising MIMO signals

作　　者：包兴先[1] 熊丛博[2] 李翠琳[3] 田玉芹[4] 

机构地区：[1]中国石油大学(华东)石油工程学院,山东青岛266580 [2]国家海洋局第一海洋研究所,山东青岛266061 [3]中国科学院海洋研究所中国科学院海洋地质与环境重点实验室,山东青岛266071 [4]青岛黄海学院交通与船舶工程学院,山东青岛266427

出　　处：《振动与冲击》2015年第19期157-162,202,共7页Journal of Vibration and Shock

基　　金：国家自然科学基金(51309239;41206042);青岛市科技计划项目(13-1-4-216-jch)

摘　　要：发展了一种基于多输入多输出(MIMO)信号降噪的模态参数识别方法。首先对实测的MIMO脉冲响应数据构建block-Hankel矩阵,然后通过模型阶次指标确定矩阵的秩,进而基于结构矩阵低秩逼近(SLRA)计算得到降噪后的信号,最后通过多参考点复指数法(PRCE)识别结构的模态参数。数值算例和模型实验结果表明,该方法对实测MIMO信号有很好的降噪作用,识别效果较好。A modal identification scheme based on denoising MIMO signals was proposed here. With this scheme,the measured MIMO impulse response data were firstly used to construct a block-Hankel matrix,and the rank of the matrix was determined based on the model order indicator,then the structured low rank approximation（ SLRA） method was implemented to achieve the denoised data. Finally,the modal parameters were estimated by using the PRCE method from the denoised MIMO signals. The effectiveness of the proposed scheme was verified with numerical examples and model test results.

关 键 词：多输入多输出 模型阶次 低秩逼近 block-Hankel矩阵 模态参数识别 

分 类 号：TU317[建筑科学—结构工程]