The Research And Application Of Time Frequency Analysis In Seismic Signal Processing

Time-frequency analysis is a powerful tool for non-stationary signals, and it becomes a hot spot in the modern signal processing study. In this paper, it introduces the basic concept of time-frequency analysis systematically, and gives the formulas of several commonly used methods and their nature in details, including short time Fourier transform, continuous wavelet transform, S transform and generalized S transform. Based on the generalized S transform developed by Mansinha, a new method is developed, and it's named as generalized S-transform with variable-factor. By processing the same non-stationary signal, I study the application effect of the several time-frequency analysis methods comparatively, and the results shows that, compared with other methods, the generalized S-transform with variable-factor can describe the local characteristics of non-stationary signals more detailedly, and it's much more objective and practical.Seismic signals are non-stationary signals, the time-frequency analysis method provides us joint distribution information in time-frequency domain of the signals, which describes the relationship clearly as the frequency of the seismic signals changing over time. In this paper, the time-frequency analysis methods are applied to seismic data processing. By processing the model data, I study the application effect of the short time Fourier transform, continuous wavelet transform and the generalized s transform with variable-factor comparatively in detecting the thin bed's boundary. At the same time, the generalized s transform with variable-factor is applied to the sub-frequency processing and the time-frequency filtering, and some desired results are achieved. The influences of the some factors on the peak frequency of the seismic wave are analyzed, such as the thickness, quality factor and wave velocity of the thin layer and the peak frequency of the source seismic wave. In the field seismic data processing, by combining the time-frequency analysis methods with spectral decomposition technique, I utilize the low frequency anomalies to detect the location of oil and gas reservoirs, and some desired results are also achieved.