The Wavelet Analysis And Its Application In Deformation Prediction Research

Since the birth of wavelet theory, it has been used in the field of deformation data processing and analysis with its outstand time-frequency analysis ability and multi-resolution analysis ability widely, But in the current deformation analysis, there are two problems in wavelet analysis, which have not been solved completely, First, how to accurately evaluate the wavelet denoising, wavelet de-noising process,there are some factors to influence the wavelet denoising effect, which directly affects the effectiveness and practicability of deformation data after wavelet denoising, according to the current research status of wavelet denoising quality method is studied; Second, how to improve the advantages of wavelet multi-resolution analysis in extracting periodic terms, and effectively apply wavelet multi-resolution analysis to the analysis and application of deformation monitoring data. The main work of this paper is as follows:(1) This paper studies and analyzes the commonly used parameters which affect the wavelet denoising, The existing methods for determining the best decomposition scale of wavelet are studied, and their evaluation ability is obtained, Finally, the effectiveness of the method is verified by designing different types of experiments. Finally,combined with the actual project, the prediction model is established under the guidance of the new wavelet denoising quality evaluation method.(2) Wavelet multi-resolution analysis can be used to decompose the low frequency and high frequency, deformation data and long-term observation of obtaining implicit periodicity in the long-term trend,Wavelet multi-resolution analysis is proposed based on the advantages of a new regression analysis model, and introduces the main steps and methods to establish the model, finally with an engineering example to verify the validity of the model, and the deformation data is predicted and analyzed.