Application of Wavelet Transform in Stochastic Loading Characteristics Analysis

Abstract

Understanding the characteristics of stochastic loading is the premise and foundation for structure analysis. Traditional amplitude spectrum and power spectral density (PSD) method based on Fourier transform (FT) has limitation, that is, it loses any information with time and can not reflect the non-stationary characteristics of stochastic loading. Wavelet transform possesses assembling ability in both time and frequency domain,and it possess stronger ability of analyzing non-stationary signal. In this paper, the method of estimating PSD and time-dependent PSD of stochastic loading using wavelet transform is researched. The calculating process is programmed and the method is validated by a numerical example. The analysis result of example indicates that the method presented in this paper is feasible.

Bayesian dynamic linear model framework for structural health monitoring data forecasting and missing data imputation during typhoon events

“…Wang et al 23 utilized VHGP to model and forecast the deformation monitoring data of high-speed rail track slab, and obtained robust forecasting results. Regarding Spectro-Temporal methods, Bai et al 24 used the wavelet transform method to estimate the power spectral density (PSD) of random load data. Compared with the traditional Fourier transform, non-stationary characteristics of random load can be reflected by this method, which overcomes the defect of Fourier transform lacking time information.…”