Application Of Wavelet Analysis On Elastic Wave Signal Process Detecting Thickness Of Top Coal

This thesis mainly studies the applications of the wavelet transform and the time-frequency showing on the detection of the arriving points of the elastic wave transient signal in detecting thickness of the top coal. The main contents can be summarized as follows:The thesis studies the principle, physical model and method of detecting thickness of the top coat by means of elastic wave way and analyzes the existing problems and corresponding solutions .The thesis describes the characteristics of wavelet bases, and main reference criterions of choosing the wavelet base are proposed. Proper wavelet base is put forward to detect the thickness of the top coal and studies the problem of continuous wavelet transform to measure arriving point of transient state signal. For transient state characteristics of the signals, time-frequency characteristics and shows which are hold in continuous wavelet transform are studied and applied to detect the thickness of the top coal, which proves the exactness about choosing Db6 wavelet in detecting the catastrophe point. At the same time, it is showed that suitable wavelet base is favorable for display transient state signal's catastrophe point and feasible to detect transient signals. Besides, the energy show's theory of Wigner in the time-frequency field is studied. There are some questions are existed in the method in the course of detecting. A nuclear function of Choi-Williams is put forward and used to dispel influences of cross items in Wigner contribution and to process signal in detecting the thickness of top coal. The result indicates that the method can efficiently dispel influences of cross items and can improve the ability in discerning the signal in detecting the thickness of the top coal. Multi-scales analysis of which can measure the catastrophe point of signal is utilized to analyze the signal by different form of expression of signal and noise in different scales. On this basis, the wavelet mould maximum principle is proposed to eliminate noise of signal, which utilize the frequency range and signal characteristics to reconstruct the signal and eliminate the noise of signal by selecting the adaptive threshold value. That is to say the : signal reconstruction is executed based choose special frequency to detect arriving point of transient signal through measureyardstick peak value and specific frequency sections of signal. At the same time, the method of detecting arriving point of transient signal based adaptive threshold value of wavelet, which by the peak value and phase characteristic in different scales automatically calculate adaptive value, produce the value about the signal's time to eliminate the noise and reconstruct the signal which has eliminated the noise. Besides the theories of wavelet are applied to signal's procession of detecting the thickness of the top coal by elastic wave way. The several methods are showed to be feasible through the result of detecting the arriving point of transient state's signals and can judge the arriving point of transient state's signals. By applying of wavelet transform in detection of elastic transient signal of the thickness of the top coal. Theoretical analysis and the results of experiments both indicate that this method can improve the precision of measuring top seam and is simple, reliable and have wide development prospects.Matlab tool software of Mamworks Company is used in the article, which have strong and convenient data procession, the functions of wavelet toolbox and time-frequently toolbox. Thus it is proved in theory that wavelet and time-frequency theories show more merits than traditional FFT methods in processing with steady and non-steady signals. And time-frequency characteristics are narrated and form frequency spectrum when the theories of wavelet and time-frequency show are applied. While the signal is processed, the method has strong resolution capability especially when dealing with transient signal in detecting the arriving point of processing signals. Through detecting the thickne...