|
CSME 2018/02
Volume 39 No.1
:
113-120
|
|
Study on Relation Between Noise and Matrix Dimension of Data-driven Stochastic Subspace Identification Method
Jun-feng Xina, Yong-bo Zhangb and Jin-lu Shengc
aDepartment of Naval Architecture, Qingdao University of Science & Technology, Qingdao 266100, China
bNational Ocean Graphic Center, Qingdao 266071, China
cCollege of Traffic and Transportation, Chongqing Jiaotong University, Chongqing 400074, China
|
Abstract:
As a linear system identification method, the data-driven stochastic subspace method can effectively obtain modal parameters from the structural signal under ambient excitation. The noise reduction ability of the method is related with its Hankel matrix dimension. The relation between the noise and Hankel matrix of data-driven subspace identification method was introduced theoretically. And a verification procedure was proposed to justify that the noise can be eliminated properly by the data-driven subspace identification method with selected Hankel matrix. The procedure includes SVD, stability diagram and finite element result ( FE). The results of numerical study and jacket platform vibration test demonstrate that the data-driven stochastic subspace identification method with non-square Hankel matrix is of better capability of denoising and can estimate modal parameters with higher accuracy. Effective modal identification method can obtain the relevant parameters of the structure, which can accurately get the basic health status of large structures. Many scholars at home and abroad (Ewins et al., 1984; Juang et al.;1985; Overschee et al., 1993) proposed modal parameter identification method based on time-domain response, such as time series, random decrement method, the natural excitation technique, data-driven stochastic subspace method. Among them, the data-driven stochastic subspace method is by far one of the more advanced modal parameter identification methods under environmental incentives. It can extract modal parameters of large structures such as offshore platforms more accurately. However, to determine the Hankel matrix dimension is the key to the effective application of this method based on Peeters(1999), and different Hankel matrix dimensions will result in different noise-canceling capabilities of data-driven stochastic subspace method(Maia et al., 1997; Peter et al.,1996; Li et al.,2011). How to determine the dimensions of Hankel matrices? Current research on this issue is still rarely reported. In response to these problems, we studied the theoretical relationship between Hankel matrix dimension and the data-driven stochastic subspace noise-canceling method, then we proposed Hankel matrix dimensions selection method of data-driven stochastic subspace identification method. And we discussed the relationship between data-driven stochastic subspace noise-canceling capabilities and different Hankel matrix build ways.
|
|
Keywords: hankel matrix; denoising; data-driven subspace identification method
|
Download PDF
|
|
*Corresponding author; e-mail:
|
©
2018
CSME , ISSN 0257-9731
|