Thin Bed Prediction Processing Technique Of Seismic Signal Based On S Transform

In oil exploration, it has become an urgent problem to use good underground geological structures to find underground oil and gas reservoirs. So geophysicists have proposed and developed many new methods and new technology. The spectral decomposition method based on the S transform time-frequency analysis method this paper used is one of them.First, this paper designs a few typical reflection coefficient series models of seismic thin layer, and analyzes the time domain and frequency domain characteristics of reflection coefficient theoretically, and then discusses influences of the waveform characteristics which is influenced by the thickness of seismic wave single layer, the size and polarity of multi-layer reflection coefficient to study the influence factors of characteristics of thin layers seismic wave. Secondly, for aiming at the application of the seismic data processing, several traditional time-frequency analysis methods are presented entirely. The original derivation, mathematical representation and basic features of S transform are introduced in detail.In order to enhance the resolving power of S transform on seismic signal thin layers, the adjustment factors are introduced into the variable Gaussian window function of S transform to obtain the generalized S transform whose time-frequency resolution is adjustable. It improves the utility and flexibility of S transform time-frequency analysis method in non-stationary signal processing. In order to compare the resolution of thin layer, we use S transform and generalized S transform to analyze the thin layer, and the results prove that generalized S transform is more effective than S transform.The simulations of thin layer models demonstrate that generalized S transform time-frequency method is able to exactly describe the localization characteristics of thin layer seismic signal. Utilizing the generalized S transform as a main tool for time-frequency spectral analysis of thin layer seismic signal, this paper studied the application of high-resolution processing of seismic thin layer based on the generalized S transform in the three aspects of the detecting of the arrival time of different signals, the thin layer identification and the making of high-resolution profile. A number of examples illustrate that generalized S transform potentially can be used to improve the resolution of thin layer seismic signal, and subtly describe the underground geological structure. They offer important basis for interpretation of seismic data.