Study Of Application For Structural Nondestructive Testing Based On Wavelet Analysis

The Wavelet Analysis is a newly developed theory which overcomes the shortage of the traditional Fourier Analysis. There are very good localization characteristics for wavelet both in the time domain and in the frequency domain. It has been used in many fields such as signal processing, image processing and sound analyzing. This thesis is focused on the theory and its application for structural non-destructive testing based on wavelet transform in view of complexity and easily effect of civil engineering structures detection. Two kinds of methods about the concrete structure inner defects identification are present and testified.For the concrete structure inner defects identification, the non-destructive testing method based on wavelet analysis singularity's detection is firstly investigated. The singular points of signal often provide some important information, which is one of the main characters of a singular signal. The theory of singularity's detection based on wavelet transform is explained. The simulated signal shows that singularity's detection based on wavelet transform can effectively detect the different condition singular points of signal. The experiment confirmed that the inner defects of the concrete structure is identified. The different position defects and small defects can't be identified by Fourier transform. The common filter wave is also effectively detected by this method. The final analysis of the data shows this method is convenient and accurate to identify if there are the inner defects and the location of the inner defects of the concrete structure on processing signal.The detecting signal is often disturbed by various noises in the process of the application of structural non-destructive testing. The signal that is disturbed by noises may be processed for the accurate results. In the thesis, the fundamental theory and method of the signal denoise based on wavelet transform is explained. It is testified by the analysis of the simulated signal and experimental signal, the signal's noise is effectively eliminated by wavele(?) denoise technology and the signal's original information including the signal's singular points is retained.In the thesis, for the concrete structure inner defects identification , another non-destructive testing methodbased on wavelet packet analysis is researched .Contrast with wavelet analysis, the advantage of wavelet packets analysis lies in analyzing the high-frequency part of received waves, as well as the low-frequency part to improve the resolving power in time domain. In the ultrasonic detecting experiment of the concrete structure's inner quality, the received ultrasonic signal was decomposed and reconstructed by wavelet packets, then each frequency band energy was acquired. The energy of each frequency band was constructed as the characteristic vector of nondestructive testing to effectively diangose inner quality of concrete. The research of Mahalanobis distance's identification methods shows that the concrete structure inner defects identification based on wavelet packet analysis can have a quantitative results and the defects' size can be also estimated by this methods.