The Study Of Sedimentary Cycle Analysis Method Based On Generalized S Transform

As a new method of sedimentary cycle research, time-frequency analysis technology is widely used in the study of sedimentary cycle. Traditional time-frequency analysis methods mainly include, short time Fourier transform（STFT）,Hilbert transform（HT）,Wavelet transform（CWT）, Wigner-Ville distribution （WVD）. S transformation（ST）.etc. However, the windows size of STFT is fixed, and the effect of low and high frequency analysis is also poor. HT can be well used to analysis signal, but it is difficult to analyze multi-frequency signal. Because the wavelet of CWT is uncertain, CWT is more results. Moreover, there is no relationship between scale factor and frequency of CWT, so HT is not a true time-frequency analysis. ST is combined with the advantages of STFT and CWT, and with good time-frequency analysis feature. However, the basic wavelet of ST is fixed, so there is much limitation in practical application. In a word, due to their shortages, the time-frequency resolution of traditional time-frequency analysis methods are not very good, and limited in the actual application, so they are very difficult to meet the high accuracy time-frequency analysis nowadays.Because of introducing two parameters （A., p） to change the window function of ST, the time-frequency resolution of generalized S transform （GST） is improved greatly. The principle and window function of GST are deeply studied. Based on this, the time-frequency analysis algorithm of GST is designed. The software of GST time-frequency analysis is written by language of C++in the programming environment of Qt3.3.2, at the same time, the software is debugged and analyzed. In order to verify the high time-frequency resolution of GST, the contrast experiments of ST and GST are made by the stationary signals and non-stationary signals.In order to explore the effect of GST in sedimentary cycle study, according to the rules of actual stratum sedimentary, four typical cycle models,including normal cycle model, inverse cycle model,and complex cycle model, are structured and simulated by Tessral₂D wave forward modeling software. These forward synthetic seismograms are analyzed by the time-frequency analysis software of GST. It turned out that the method of using GST to study sedimentary cycle is right and superior. Finally, the method is applied to study the sedimentary cycle by actual seismic data, and achieving a good effect. It is proved that sedimentary cycle study by GST is effective and correct once again.