Abstract: A frequency and spatial domain decomposition (FSDD) method for operational modal analysis (OMA) is presented in this paper, which is an extension of Complex Mode Indicator Function (CMIF) method for experimental modal analysis (EMA). Singular value decomposition is adopted to separate the signal space from the noise space. The singular values plot is a good indicator for locating modal frequencies. The singular vectors are employed as a weight function to obtain an enhanced power spectrum density (PSD), from which accurate modal frequency and damping ratio are got through the least square fitting in frequency domain. A simulation case of a two-storey building and an application case of the famous Z24 highway bridge are employed to validate the FSDD algorithm.

Key words: modal parameters identification, operational modal analysis, frequency and spatial domain, least square fitting