首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于时频滤波和自回归建模方法的时变模态参数辨识
引用本文:续秀忠,张志谊,华宏星.基于时频滤波和自回归建模方法的时变模态参数辨识[J].上海海事大学学报,2005,26(4):1-5.
作者姓名:续秀忠  张志谊  华宏星
作者单位:1. 上海海事大学,物流工程学院,上海,200135
2. 上海交通大学,振动、冲击、噪声国家重点实验室,上海,200030
基金项目:国家自然科学基金资助项目(10302019)
摘    要:应用非平稳信号的时频滤波及非平稳时间序列的时变自回归(AR)建模方法进行时变线性系统的模态参数辨识,将基于Gabor展开的时频滤波方法引入多自由度线性时变结构模态参数辨识中,提取单模态响应分量。对线性时变系统在白噪声激励下振动响应的单模态响应进行建模来辨识系统的每一阶模态参数,在模型参数的辨识中通过引入基函数将非平稳过程的辨识问题转化为线性时不变过程的辨识,利用时变的伯格(Burg)算法对时变的AR模型系数和时变结构模态频率进行估算。通过对附加质量随时间连续变化的悬臂梁的模态分析验证了辨识方法的有效性。实验结果和理论计算结果比较表明:该方法为参数时变线性系统的模态参数辨识提供一条新途径。

关 键 词:时变线性系统  模态参数辨识  时频滤波  单模态响应  时变自回归建模
文章编号:1672-9498(2005)04-0001-05
收稿时间:2005-04-13
修稿时间:2005-10-20

Identification of time-variant modal parameters by time-frequency filter and autoregressive modeling
XU Xiuzhong,ZHANG Zhiyi,HUA Hongxing.Identification of time-variant modal parameters by time-frequency filter and autoregressive modeling[J].Journal of Shanghai Maritime University,2005,26(4):1-5.
Authors:XU Xiuzhong  ZHANG Zhiyi  HUA Hongxing
Abstract:A new method of modal parameters identification based on the time-frequency filter and autoregressive modeling is proposed. Single modal response of time-variant multi-degree freedom system is extracted by Gabor expansion. Each rank modal parameter is identified by an autoregressive model of single modal response. A new algorithm is presented to estimate the model coefficients and modal frequencies, in which the base functions have shifted the time-variant modeling to a linear time-invariant identification problem. The identification of a cantilever with varying mass is given to validate the proposed method. The identification results have indicated that the proposed method is an effective tool for modal parameters identification of time-variant linear system.
Keywords:time-variant linear system  modal parameters identification  time-frequency filter  single modal response  time-variant autoregressive modeling
本文献已被 CNKI 维普 万方数据 等数据库收录!
点击此处可从《上海海事大学学报》浏览原始摘要信息
点击此处可从《上海海事大学学报》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号