Structural Modal Parameter Identification Using Wavelet Transform and Singular Value Decomposition under Ambient Excitation

闵志华,孙利民,孙智,淡丹辉

DOI: https://doi.org/10.3969/j.issn.1004-4523.2009.02.006

2009-01-01

Abstract:Structural modal parameters were identified using the method of wavelet transform and singular value decomposition under ambient excitation.Assuming that the structural system was time-invariant and the ambient excitations were white noise,the covariance response of structural accelerations were calculated from response data measured at multiple locations of the structure;and wavelet transform was imposed to obtain the wavelet coefficient matrix of the covariance response.Then the singular value decomposition was adopted to decompose the wavelet coefficient matrix at each wavelet scale,and the modal parameters could be identified exactly from the singular value and singular vector.A numerical example of three-degree-of-freedom structure was used to analysis the noise resistance of the method.Results show that the method does not seen to be affected at the presence of 15 dB measurement noise and at same time the accuracy was better than that using a conventional wavelet transform method.Finally the result of a cable-stayed bridge validated the method.