TIME-VARYING SPECTRUM ESTIMATION OF EARTHQUAKE GROUND MOTION VIA MODERN TIME-FREQUENCY ANALYSIS
Liang Hu
Abstract
Liang Hu
Abstract
The Problems of estimating the time-varying spectrum of earthquake ground motions estimation using STFT, Wigner-Ville distribution, continuous wavelet transform, S-transform and improved S-transform with a complex window are addressed. Mathematical expression of wavelet energy spectrum based on continuous wavelet transform was derived. Numerical results show that estimating the time-varying spectrum of separable non-stationary ground motion via STFT is effective provided the width of window is chosen properly, however,wavelet transform,S-transform and improved S-transform will lead to energy overlapping in high frequency range. The improved S-transform is found to be most effective to estimate the time-varying spectrum of ground motion with high non-stationariness in frequency by adjusting the parameter of time window.The continuous wavelet transform has effective time-domain resolution even though the energy overlapping in high frequency range.
OpenAlex reports 1 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
The Problems of estimating the time-varying spectrum of earthquake ground motions estimation using STFT, Wigner-Ville distribution, continuous wavelet transform, S-transform and improved S-transform with a complex window are addressed. Mathematical expression of wavelet energy spectrum based on continuous wavelet transform was derived. Numerical results show that estimating the time-varying spectrum of separable non-stationary ground motion via STFT is effective provided the width of window is chosen properly, however,wavelet transform,S-transform and improved S-transform will lead to energy overlapping in high frequency range. The improved S-transform is found to be most effective to estimate the time-varying spectrum of ground motion with high non-stationariness in frequency by adjusting the parameter of time window.The continuous wavelet transform has effective time-domain resolution even though the energy overlapping in high frequency range.
Key concepts: Short-time Fourier transform, Wavelet transform, Harmonic wavelet transform, S transform, Continuous wavelet transform, Wavelet, Discrete wavelet transform, Constant Q transform