Filtering and smoothing of base loading spectrum using singular spectrum analysis

摘要

Power spectral densities of the wind loading acquired from the high-frequency force balance or the synchronous pressure integration measurements appears to be a saw-toothed shape. Noise filtering and smoothing of base loading spectrum prior to frequency domain analysis are of significance and can be improved using the singular spectrum analysis. The revised singular spectrum analysis tailored for base loading spectrum is briefly described and applied to base loading spectra of tall twin buildings. The filtered and smoothed base loading spectrum may reduce potential spectral errors inherent in estimation of wind-induced response. The singular spectrum analysis has a great potential for filtering and smoothing of wind engineering data.