Time-frequency Analysis And Its Implication In Seismic Signal

Seismic signal is one of very complicated non-stationary signals, and time-frequency analysis is an important method in non-stationary signal analysis. So this thesis presents the study of applying time-frequency analysis to seismic signal processing works.There are four chapters in the thesis. First: the study background. Second: the basis of time-frequency analysis, introduces the basic theory and deduces some theorems and formulas in detail. Third: general time-frequency analysis methods, introduces some general time-frequency analysis in seismic methods such as short-time Fourier transformation, wavelet transformation, quadratic time-frequency analysis (includes Wigner-Ville transformation and Affine transformation); and then improves Wigner-Ville transformation by smoothing and windowing based on the study of improvement of quadratic time-frequency analysis. Finally, the application of time-frequency analysis in seismic signal processing works, respectively.The main achievements of the thesis are as follows:(1) Improves the Wigner-Ville transformation by smoothing and windowing on the basis of the study of improving quadratic time-frequency analysis.(2) By comparing these methods gets the result that Wigner-Ville transformation has best time-frequency resolution and smoothing and windowing improvement can eliminate the disturb terms in Wigner-Ville transformation.(3) The Improved Wigner-Ville transformation is applied to seismic processing, such as: displaying seismic profile by single frequency mode, detecting the low frequency shadow related to gas concentration and so on. The result shows that the time-frequency analysis can be used to noise suppression directly indicate hydrocarbon and so on, all of which would be very important for seismic processing works.