Wavelet Transform With Applications In Data Processing Of Bridge's Deformation Observation

The deformation of a bridge likes a signal which is changing with time and space, so the deformation analysis is a signal analysis. Wavelet analysis theory was born in 1980s, one of the latest localization analysis of time-frequency,which was considered to be the breakthrough since the Fourier analysis method. Wavelet analysis overcomes the shortcomings of Fourier analysis that can not combine the time-domain with the frequency domain of a signal together to describe the observation, which can combine the time domain with frequency domain together to analysis the deformation signal. Wavelet Transform can not only preprocess the deformation data, but also can effectively eliminate measurement error and extract deformation features so as to analyze the deformation rules. Multi-scale analysis of wavelet transform reveal the nature of things and make it clear, concise, computational complexity low. Using wavelet multi-resolution can decompose bridge deformation monitoring signal at different scales,so we can find more information in different resolutions and reduce the uncertainty and complexity of problem. The method of wavelet filtering denoising in non-stationary signal has unparalleled advantages than other methods. Wavelet theory will play an important role in the data processing of dynamic deformation monitoring.In the paper, wavelet analysis was used as a tool to deal with the obtained deformation observation data of a bridge engineering practice.To full use of the advantages of frequency analysis, singularity detection, extraction deformation trends and gross error detection of wavelet analysis to deal with the bridge observation data signals. Study on the theory of wavelet denoising processing applications in the bridge observation data mainly. Research on the parameters selection of wavelet decomposition and reconstruction denoising method and nonlinear wavelet transform thresholding denoising. The deformation monitoring data was decomposed by multi-scale using wavelet filtering of signals, including a variety of signal frequency components to be broken down, removing the observation noise sequence,and extracting deformation information. Studies show that we can largely eliminate the high frequency noise and improve the accuracy of observation data, by choosing the appropriate wavelet function, decomposition layers, threshold and denoising method.Finally,we applied the wavelet analysis to the bridge vertical deformation monitoring project engineering, including process the monitoring data and analysis the result of monitoring data's wavelet transform and achieved a good result. And researched the relationship between the swing deformation speed of the bridge and speed of wind and proved that only if the frequency of wind and speed of the wind under certain scope could guarantee the safety of the bridge swing.