AbstractOperational modal identification is indispensable in bridge health monitoring because it estimates modal parameters only from the multiple-channel responses of bridges under ambient excitation. Multiple-channel responses are often slightly asynchronous due to sensor errors, which is generally ignored in response monitoring but introduces significant uncertainty to modal identification. Considering that calibrating the asynchronous responses during the sensing or data acquisition process is not economical, an identification method based on the natural excitation technique (NExT) is proposed for asynchronous responses. In the NExT method, the power spectra of monitoring responses are first calculated and then transformed into correlation functions by inverse Fourier transform. The phase characteristics of power spectra and the linearly dependent characteristics of modal components in correlation functions are compared for synchronous and asynchronous responses, respectively, with consideration of the noise level and asynchronous time. Then the enhanced NExT is proposed, in which the multiple-channel responses are synchronized by minimizing the phase slope and maximizing the linear dependency of modal components. Finally, the responses of a numerical example and a highway bridge are taken to verify the method. The results show that the multiple-channel asynchronous responses of bridges under ambient excitation can be effectively aligned to obtain high-precision mode shapes.

Source link

Leave a Reply

Your email address will not be published. Required fields are marked *