High-resolution Spectral-analysis for Fundamental Frequency Estimation of High-rise Buildings subjected to Earthquakes

Accurate estimation of the fundamental frequency of civil structures is crucial for many applications; in particular, this estimation becomes more important when it helps to avoid or minimize human, economic, and structural damages as the case of the design and analysis of high-rise buildings subjected to earthquakes. Although many techniques have been proposed for this task, several aspects such as accuracy, noise immunity, and suitable time-frequency resolution have not been adequately addressed. In this work, the short-time multiple signal classification (ST-MUSIC), a high-resolution spectral-analysis technique, is proposed to estimate the fundamental frequency of high-rise buildings using vibration signals. The proposal is validated and tested using synthetic signals and real measurements. In the latter, the fundamental frequency of a high-rise building, 1:20 scale model, is estimated before-, during-, and after-seismic excitation through the analysis of the generated vibrations. For comparison purposes, the short-time Fourier transform (STFT), a conventional time-frequency technique, is also used. The obtained results show a high accuracy and noise immunity in the analysis, which makes the proposal a suitable and reliable tool for this task.