| Title |
Experimental Verification of Mode Decomposition Using Independent Component Analysis Method |
| Authors |
Hwang Jae-Seung ; Park Seong-Cheol |
| Keywords |
Scale Model Structure ; Independent Component Analysis ; Blind Source ; Mode Shape ; Mode Separation ; Hilbert Transform ; Instantaneous Natural Frequency ; Derivative of Covariance ; Output-only System Identification |
| Abstract |
Independent component analysis is one of the linear transformation techniques to separate blind sources from the output signals of system. Recently, the method has been analytically applied for the responses of building structures to obtain the mode shapes and modal responses. The aim of this study is to experimentally verify the feasibility of the blind source separation by ICA method for three storey shear model structure and propose a method to effectively estimate the dynamic properties from the separated modal responses. From the results of experimental verification procedure, it is found that the mode shapes separated by ICA are similar to the mode shapes by the conventional method such as analytic eigen problem or power spectrum method, furthermore, ICA can efficiently and precisely estimate the mode shapes based on the mathematic robustness. In addition, by the proposed method to estimate the dynamic properties in this study, it is shown that the modal properties : natural frequencies and damping ratios of model structure are estimated effectively from the probability distribution function of instantaneous natural frequency determined by Hilbert transform and the differential equation of the output covariance, respectively. Finally, the results in this study also show that the proposed method can be very usefully applied to the output-only system identification. |