The invention discloses a time-varying mixed-phase seismic
wavelet extraction method based on time-
frequency spectrum simulation, which is characterized in that: on the basis of generalized S-transformation improved on non-stationary seismic records, a time-frequency filter is introduced for the first time to the seismic records The time-
frequency spectrum is filtered, and then the spectrum
simulation method is used to estimate the
wavelet amplitude spectrum at each moment. Under the assumption that the
wavelet phase is time-invariant, the high-order cumulant bispectral method is used to remove the time-varying wavelet amplitude spectrum. The
mixed phase spectrum of the wavelet is reconstructed from the seismic
record, and finally the amplitude spectrum and the
phase spectrum are combined to realize the extraction of the time-varying
mixed phase wavelet. Compared with the conventional segmentation extraction method, the time-varying wavelet extraction method proposed by the present invention does not require the assumption that the wavelet is segmentally stable, and can accurately estimate the time-varying wavelet when there is a large change in the wavelet properties of adjacent
layers , and the extraction of
mixed phase of wavelet is more in line with the actual situation, it can analyze and process seismic data more finely and accurately, and has good application value in improving the resolution of seismic data.