CN-122017981-A - Gradual frequency expansion method and device, electronic equipment and storage medium

CN122017981ACN 122017981 ACN122017981 ACN 122017981ACN-122017981-A

Abstract

The application provides a gradual frequency expansion method, a device, electronic equipment and a storage medium, wherein the method comprises the steps of obtaining a pre-stack CRP gather; the method comprises the steps of performing interpretation CRP gather superposition on the prestack CRP gather to obtain superposition seismic data, performing scale domain spectrum equalization processing on the superposition seismic data to obtain a reconstructed seismic image, performing smoothing filtering processing on the reconstructed seismic image by adopting anisotropic diffusion filtering, selecting a scale range with high signal to noise ratio in the reconstructed seismic image after the smoothing filtering processing to extrapolate other scale segments to obtain seismic data of a plurality of scale segments, and performing signal reconstruction on the seismic data of all scale segments to obtain broadband seismic data. The method can remarkably widen the relative effective frequency band of the seismic data, better delineate stratum details, effectively improve the resolution of the seismic data, and has higher industrial practical value and application prospect in oil-gas seismic exploration.

Inventors

LIU LANFENG
LIU JUNZHOU
LIU XIWU

Assignees

中国石油化工股份有限公司
中国石油化工股份有限公司石油勘探开发研究院

Dates

Publication Date: 20260512
Application Date: 20241112

Claims (10)

1. The gradual frequency expansion method is characterized by mainly comprising the following steps of: Acquiring a pre-stack CRP gather; Performing explanatory CRP gather superposition on the pre-stack CRP gathers to obtain superposition seismic data; performing spectrum equalization processing on the overlapped seismic data in a scale domain to obtain a reconstructed seismic image; Performing smoothing filtering treatment on the reconstructed seismic image by adopting anisotropic diffusion filtering; Selecting a scale range with high signal-to-noise ratio in the reconstructed seismic image after smoothing filtering processing to extrapolate other scale sections to obtain seismic data of a plurality of scale sections; And reconstructing the signals of the seismic data of all the scale sections to obtain broadband seismic data.
2. The gradual frequency development method according to claim 1, wherein the performing of the interpreted CRP gather superposition on the pre-stack CRP gather is specifically: And selecting waveform characteristics on the pre-stack CRP gather to be stable, wherein the amplitude and the phase of a target layer are reliable, and offset distance segments with good signal noise are overlapped.
3. The progressive frequency extension method according to claim 1, wherein the performing scale-domain spectrum equalization on the superimposed seismic data to obtain a reconstructed seismic image comprises: Performing wavelet transformation on the superimposed seismic data to obtain a scale domain signal; calculating an amplitude envelope curve for each scale domain signal, smoothing the amplitude envelope curve, and generating a gain curve under a corresponding scale; Dividing each scale domain signal by a gain curve under a corresponding scale to obtain an equalized scale signal; and carrying out wavelet inverse transformation reconstruction on all the balanced scale signals to generate a reconstructed seismic image.
4. The gradual frequency development method according to claim 1, wherein the anisotropic diffusion filtering is calculated as follows: Wherein u (x, y, 0) represents a reconstructed seismic image, div represents a divergence operator, D represents a diffusion tensor, S ρ represents a structure tensor; T represents time, u represents seismic data, and u 0 (x, y) represents initial seismic data; The calculation formula of the structure tensor S ρ is as follows: Wherein s 11 、s 12 、s 13 、s 22 、s 23 、s 33 represents each element of the structure tensor, G ρ represents a gaussian function, σ represents a gaussian function parameter, and x, y and z are three directions in space; The formula for the diffusion tensor D is as follows: Wherein, the Wherein v 1 、v 2 and v 3 respectively represent eigenvectors of the structure tensor matrix, μ 1 、μ 2 and μ 3 respectively represent diffusion coefficients, d 11 、d 22 、d 33 、d 12 、d 13 、d 23 respectively represent each element of the diffusion tensor, α∈ (0, 1), λ 1 、λ 2 、λ 3 respectively represent eigenvalues of the structure tensor matrix; The expression for each element d 11 、d 22 、d 33 、d 12 、d 13 、d 23 of the diffusion tensor is as follows: Wherein v 11 、v 21 、v 31 represents three elements of a first eigenvector of the structure matrix, v 12 、v 22 、v 32 represents three elements of a second eigenvector of the structure matrix, and v 13 、v 23 、v 33 represents three elements of a third eigenvector of the structure matrix, respectively; discretizing the formula (1) by using the formulas (2) - (5) to obtain an anisotropic diffusion filtering iterative formula as follows: Wherein u k and u k+1 respectively represent the filtering results of the reconstructed seismic image at kΔt and (k+1) Δt, Δt being the diffusion time of one iteration; the conversion of equation (6) into discrete form is as follows:
5. The gradual frequency expansion method according to claim 1, wherein extrapolation of other scale segments is performed in a scale range with high signal-to-noise ratio in the reconstructed seismic image after the smoothing filtering processing is selected to obtain seismic data of a plurality of scale segments, specifically: selecting a scale range with high signal-to-noise ratio in the reconstructed seismic image after smoothing filtering processing to obtain an dominant scale signal; extrapolating the selected dominant scale signal for other scale segments to obtain an extrapolated scale signal; and carrying out inverse wavelet transformation reconstruction on the dominant scale signal and the extrapolated scale signal together to a time domain, and generating seismic data of a plurality of scale segments.
6. The gradual frequency development method according to claim 5, wherein the extrapolating the selected dominant scale signal by other scale segments to obtain an extrapolated scale signal, specifically: wavelet transformation is carried out on the dominant scale signals to obtain dominant scale domain seismic signals, and the calculation formula is as follows: Wherein Wf (a, b) represents the dominant scale domain seismic signal, a represents the scale, b represents the time shift time, f (t) represents the dominant scale signal, Representing the wavelet basis function, representing the complex conjugate, t representing time, Performing Fourier transform on the dominant scale domain seismic signals, performing scale extrapolation by utilizing the frequency shift property of the Fourier transform to obtain extrapolated scale signals, wherein the calculation formula is as follows: where Wf (a ', b) represents the extrapolated scale signal, a' represents the extrapolated scale, ω 0 represents the amount of frequency shift, j represents the imaginary number, and b represents time.
7. The gradual frequency expansion method according to claim 1, wherein the calculation formula of the signal reconstruction is: Where f' (t) represents reconstructed broadband seismic data, Representing a wavelet dependent constant, wf (a, b) representing a dominant scale domain seismic signal, a representing a scale, b representing time-shifted time, To reconstruct the wavelet.
8. The structure block diagram of the gradual frequency expanding device is characterized by mainly comprising the following modules: the data acquisition module is used for acquiring a pre-stack CRP gather; the interpretation CRP superposition module is used for performing interpretation CRP gather superposition on the pre-stack CRP gathers to obtain superposition seismic data; the scale domain spectrum equalization module is used for performing scale domain spectrum equalization processing on the superimposed seismic data to obtain a reconstructed seismic image; The guiding filter construction module is used for carrying out smooth filtering treatment on the reconstructed seismic image by adopting anisotropic diffusion filtering; the dominant scale extrapolation module is used for selecting a scale range with high signal-to-noise ratio in the reconstructed seismic image after the smoothing filter processing to extrapolate other scale segments to obtain seismic data of a plurality of scale segments; and the reconstruction module is used for reconstructing the signals of the seismic data of all the scale sections to obtain broadband seismic data.
9. An electronic device, comprising: A processor; a memory; and a computer program, wherein the computer program is stored in the memory, the computer program comprising instructions that, when executed by the processor, cause the electronic device to perform the method of any one of claims 1 to 7.
10. A computer readable storage medium, characterized in that the computer readable storage medium comprises a stored program, wherein the program, when run, controls a device in which the computer readable storage medium is located to perform the method of any one of claims 1 to 7.

Description

Gradual frequency expansion method and device, electronic equipment and storage medium Technical Field The application relates to the field of geophysical exploration, in particular to a gradual frequency expansion method, a gradual frequency expansion device, electronic equipment and a storage medium. Background With the increasing development degree of oil and gas field exploration, thin interbed and small target body become important components of upstream storage enhancement, and high-quality seismic data are inexhaustible in the process of finding out the hidden oil and gas reservoirs, so that the improvement of the resolution of the seismic data is always a focus of research of students at home and abroad. The technology commonly used at present is mainly deconvolution technology, and the deconvolution technology achieves the purpose of improving the time resolution of seismic data by compressing seismic wavelets. The Massachusetts institute of technology geophysical analysis group proposed deconvolution in 60 s, and based on the Robinson model, the inverse operator was first solved, and the inverse operator was used to further solve the reflection coefficient sequence. Based on this, various deconvolution methods have been proposed according to different needs through continuous efforts and attempts by those skilled in the art. In 1969, predictive deconvolution was initiated. In 1975, the Berger deconvolution was proposed, which is based on the maximum entropy analysis method. Homomorphic deconvolution avoids the assumption of minimum phase and reflection coefficient white noise of the seismic wavelet, and can extract the seismic wavelet and the reflection coefficient at the same time. The minimum entropy deconvolution can effectively strengthen the characteristic of sharp pulses and enhance the identification capability of bright spots of a section. Maximum likelihood deconvolution appears in the 80 s of the 20 th century, and the method effectively solves the defect that signal-to-noise separation cannot be realized by the traditional deconvolution. Blind deconvolution then occurs, which has a better practical value without any hypothetical conditions than conventional deconvolution assuming many ideal conditions. At the end of the 90 th century, with the further research and application of wavelet analysis, deconvolution with resolution varying with wavelet scale is performed on each scale in the wavelet transform domain of seismic signals, and the high-resolution deconvolution result is approximated from the low-resolution deconvolution result by utilizing the correlation between deconvolution results with different resolutions and measuring the attenuation characteristic of noise with scale. Gabor deconvolution is based on an unsteady seismic trace deconvolution model, deconvolution processing is carried out in a time domain, and a traditional time domain or frequency domain deconvolution mode is broken through. The non-white noise characteristic of the actual reflection coefficient sequence is considered in the colored deconvolution, the signal purity spectrum is used as the expected amplitude spectrum output by the deconvolution, and the colored compensation is carried out, so that the resolution of the seismic data is further improved. The deconvolution technology based on the white noise reflection coefficient and the known wavelet phase assumption condition is used for time domain wavelet compression, high-frequency information is usually lifted, but the high-frequency information of an actual seismic signal is seriously polluted by noise, and noise is also improved when a high-frequency effective signal is lifted, so that the signal-to-noise ratio of a processed seismic section is low, and a transverse phase axis is discontinuous. On the other hand, the low-frequency information of the seismic signals is not expanded, so that the relative amplitude relation in the longitudinal direction is kept poor, and the relative bandwidth expansion of the processed seismic signals is limited. After seismic data acquisition and conventional seismic data processing, the seismic interpretation stage is entered, and the stepwise widening method for the seismic frequency bands is provided, which is an explanatory widening method. Based on the existing manpower and resources, the seismic broadening frequency band flow and method for the interpretation stage can not be performed under the condition that conventional repeated seismic processing is impossible. The method for widening the seismic frequency band step by step from pre-stack CRP gather to post-stack on the basis of pre-conventional seismic processing aims to extrapolate low-frequency and high-frequency information of signals on the basis of keeping the effective frequency band and the original signal-to-noise ratio of original seismic signals, thereby improving the seismic resolution and meeting the recognition of thin reservoirs and small targets