CN-122017964-A - Multi-pole matching pursuit-based seismic spectrum decomposition method, device and equipment

CN122017964ACN 122017964 ACN122017964 ACN 122017964ACN-122017964-A

Abstract

The application provides a method, a device and equipment for decomposing a seismic spectrum based on multipole matching pursuit, which comprises the steps of inputting single-channel seismic data as an initial residual signal, conducting complex analysis to calculate an instantaneous envelope, determining delay time of matched wavelets, constructing a variable-phase wavelet library according to the delay time of the matched wavelets, carrying out convolution on each component and the variable-phase wavelet library after multiple parity decomposition to obtain a multipole wavelet library, matching an optimal wavelet, solving a corresponding sub-signal, subtracting the sub-signal from the current residual signal as a new residual signal, repeating the steps until iteration is finished, correspondingly multiplying the matched wavelets and amplitude coefficients to obtain a single-interface reflection seismic signal as a sub-signal set, carrying out Fourier transformation on the sub-signal, and obtaining an ideal time spectrum decomposition result on the delay time corresponding to an interface. The application reduces the redundancy of the overcomplete wavelet base, improves the efficiency of phase-change matching tracking, and realizes the seismic spectrum decomposition with high efficiency and high time-frequency resolution.

Inventors

WANG XINYU
LIU XIWU
LIU YUWEI
LIU JIONG

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 seismic spectrum decomposition method based on multipole matching pursuit is characterized by mainly comprising the following steps: step S1, inputting single-channel seismic data, and taking the single-channel seismic data as an initial residual signal; Step S2, carrying out complex analysis on the current residual signal to obtain Hilbert transformation of the current residual signal, calculating an instantaneous amplitude envelope of the complex analysis signal by utilizing the Hilbert transformation of the current residual signal, and taking the time corresponding to the maximum value of the instantaneous amplitude envelope as the delay time of the matched wavelet; S3, constructing a zero-phase Rake wavelet according to the delay time of the matched wavelet, scanning the main frequency and the phase of the zero-phase Rake wavelet, and constructing a variable-phase wavelet library; S4, performing n times of odd-even decomposition on the multi-layer reflection interface according to priori knowledge of the reflection interface to obtain 2 n pole sub-components, so that the reflection coefficient values in the pole sub-components are the same; S5, extracting a wavelet which is most matched with the current residual signal from the multipole wavelet library, and obtaining an amplitude coefficient corresponding to the most matched wavelet; subtracting the sub-signal obtained by decomposition from the current residual signal, and taking the obtained residual as a new residual signal; step S6, repeating the steps S2 to S5 until the iteration times reach a preset value or the residual energy is smaller than a certain threshold value; Step S7, after iteration is finished, obtaining a plurality of wavelets with the best matching residual signals and amplitude coefficients corresponding to the wavelets, and multiplying the wavelets with the best matching wavelets and the amplitude coefficients corresponding to the wavelets respectively to obtain single-interface reflection seismic signals as a sub-signal set; And S8, carrying out Fourier transform on each sub-signal in the sub-signal set, and corresponding the obtained Fourier transform spectrum to the delay time corresponding to the interface to obtain an ideal time spectrum decomposition result.
2. The method of claim 1, wherein the expression for performing complex analysis on the current residual signal is: S(t)=s(t)+is H (t) Where S (t) denotes a complex analysis signal, S (t) denotes a residual signal, S H (t) denotes a Hilbert transform of the residual signal, and i denotes an imaginary unit.
3. The method of claim 2, wherein the instantaneous amplitude envelope of the complex analysis signal is calculated by the formula: where a (t) represents the instantaneous amplitude envelope, s (t) represents the residual signal, and s H (t) represents the Hilbert transform of the residual signal.
4. The method of claim 1, wherein the zero-phase rake wavelet is calculated by the formula: w(t)={1-2[πf d (t-τ)] 2 }exp{-[πf d (t-τ)] 2 } where w (t) represents a zero-phase Rake wavelet, f d represents the dominant frequency of the Rake wavelet, t represents time, and τ represents the delay time of the matched wavelet.
5. The method of claim 4, wherein the phase-change wavelet library is calculated by the formula: w γ (t)＝w(t-τ,f d )cos(φ)-w H (t-τ,f d )sin(φ) Wherein w γ (t) represents a variable-phase wavelet base, phi represents phase, w (t- τ, f d ) represents zero-phase Rake wavelet with delay time τ and dominant frequency f d , and w H (t-τ,f d ) represents Hilbert transform of w (t- τ, f d ).
6. The method of claim 4, wherein extracting a wavelet from the multi-polar sub-wavelet base that best matches the current residual signal has the matching condition of: wherein, match index represents the catalog of the best matching wavelet in the phase-change wavelet base; representing the signal residual error of the previous iteration matching; Representing multipolar sub-components in the secondary parity decomposition, respectively representing odd component of odd component, even component of odd component, odd component of even component and even component of even component, respectively representing correspondent component and theta oo 、θ oe 、θ eo 、θ ee Vector included angle in multidimensional space.
7. The method of claim 1, wherein the fourier transform of each of the subset of signals is calculated as: Wherein S k (f) represents a fourier transform spectrum, S k (t) represents a sub-signal, t represents time, f represents frequency, and k represents the number of iterations; The calculation formula of the time spectrum decomposition result is as follows: Where IDFT s (t, f) represents the time spectrum of the residual signal s (t), K represents the total number of iterations, and δ (t- τ k ) represents the unit pulse function with a delay time τ k .
8. The utility model provides a seismic spectrum decomposition device based on multipole matching pursuit which characterized in that mainly includes following module: The data input module is used for inputting single-channel seismic data and taking the single-channel seismic data as an initial residual signal; The delay time acquisition module of the matching wavelet is used for carrying out complex analysis on the current residual signal to obtain a complex analysis signal, calculating the instantaneous amplitude envelope of the complex analysis signal, and taking the time corresponding to the maximum value of the instantaneous amplitude envelope as the delay time of the matching wavelet; The variable phase wavelet base construction module is used for constructing zero-phase Rake wavelets according to the delay time of the matched wavelets, scanning the main frequency and the phase of the zero-phase Rake wavelets and constructing a variable phase wavelet base; The multi-pole sub-wave library construction module is used for carrying out odd-even decomposition on the multi-layer reflection interface for n times according to priori knowledge of the reflection interface to obtain 2 n pole components, so that the reflection coefficient values in the pole components are the same; The new residual signal construction module is used for extracting a wavelet which is most matched with the current residual signal from the multipole sub-wavelet library, solving an amplitude coefficient corresponding to the most matched wavelet, multiplying the amplitude coefficient with the most matched wavelet to be used as a sub-signal obtained by decomposition; the cyclic iteration module is used for repeating the working processes of the delay time acquisition module, the phase-change wavelet database construction module, the multipole wavelet database construction module and the new residual signal construction module of the matched wavelet until the iteration times reach a preset value or the residual energy is smaller than a certain threshold value; the sub-signal set construction module is used for obtaining a plurality of wavelets which are the best match of residual signals and amplitude coefficients corresponding to the wavelets after iteration is finished, and multiplying the wavelets which are the best match with the amplitude coefficients corresponding to the wavelets respectively to obtain single-interface reflection seismic signals which are used as sub-signal sets; and the time-frequency spectrum decomposition result construction module is used for carrying out Fourier transform on each sub-signal in the sub-signal set, and corresponding the obtained Fourier transform spectrum to the delay time corresponding to the interface to obtain an ideal time-frequency spectrum decomposition result.
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

Multi-pole matching pursuit-based seismic spectrum decomposition method, device and equipment Technical Field The application relates to the field of geophysical exploration, in particular to a method, a device and equipment for decomposing a seismic spectrum based on multipole matching pursuit. Background The seismic spectrum decomposition method based on the Matching Pursuit (MP) is an adaptive time-frequency analysis method, and the main flow is that firstly, an overcomplete wavelet base is constructed, then sparse decomposition of signals is realized based on the matching pursuit, finally, the time-frequency distribution of each sub-component signal after decomposition is obtained by utilizing WVD, and the summation is overlapped. The seismic spectrum decomposition result based on the matching pursuit has very high time-frequency resolution, and avoids the influence of cross terms, so that the method is widely applied to the fields of seismic data interpretation and inversion. The common phase-changing Rake wavelet matching tracking algorithm generally exhausts the main frequency and phase of the seismic wavelet and the reflectance value and position of a reflection interface to construct an overcomplete wavelet base, so that the redundancy of the wavelet base is extremely high and the matching efficiency is low. In 2015, zhang Fanchang and the like dynamically generate wavelet libraries by adopting a method of constructing an adaptive filter by adopting the characteristics of the instantaneous amplitude, the instantaneous frequency and the instantaneous phase of the seismic data, but the method cannot guarantee global optimal matching based on the linear assumption of the seismic signals and noise. In 2019, the dynamic wavelet base matching tracking method based on deep learning is realized by using Chun-yan and the like, but the method faces the problems of data deviation and over-fitting, and meanwhile, a large amount of annotation data training work possibly causes no efficiency increase and no efficiency decrease. Therefore, it is highly desirable to provide a method for decomposing a seismic spectrum that can achieve high efficiency and high time-frequency resolution. Disclosure of Invention In view of the above, the application provides a method, a device and equipment for decomposing a seismic spectrum based on multipole matching pursuit, which aim to solve the problem of low matching efficiency caused by high redundancy of an overcomplete wavelet base. High-efficiency and high-time-frequency resolution seismic spectrum decomposition is realized. In a first aspect, an embodiment of the present application provides a method for decomposing a seismic spectrum based on multipole matching pursuit, which mainly includes the following steps: step S1, inputting single-channel seismic data, and taking the single-channel seismic data as an initial residual signal; Step S2, carrying out complex analysis on the current residual signal to obtain Hilbert transformation of the current residual signal, calculating an instantaneous amplitude envelope of the complex analysis signal by utilizing the Hilbert transformation of the current residual signal, and taking the time corresponding to the maximum value of the instantaneous amplitude envelope as the delay time of the matched wavelet; S3, constructing a zero-phase Rake wavelet according to the delay time of the matched wavelet, scanning the main frequency and the phase of the zero-phase Rake wavelet, and constructing a variable-phase wavelet library; S4, performing n times of odd-even decomposition on the multi-layer reflection interface according to priori knowledge of the reflection interface to obtain 2 n pole sub-components, so that the reflection coefficient values in the pole sub-components are the same; S5, extracting a wavelet which is most matched with the current residual signal from the multipole wavelet library, and obtaining an amplitude coefficient corresponding to the most matched wavelet; subtracting the sub-signal obtained by decomposition from the current residual signal, and taking the obtained residual as a new residual signal; step S6, repeating the steps S2 to S5 until the iteration times reach a preset value or the residual energy is smaller than a certain threshold value; Step S7, after iteration is finished, obtaining a plurality of wavelets with the best matching residual signals and amplitude coefficients corresponding to the wavelets, and multiplying the wavelets with the best matching wavelets and the amplitude coefficients corresponding to the wavelets respectively to obtain single-interface reflection seismic signals as a sub-signal set; And S8, carrying out Fourier transform on each sub-signal in the sub-signal set, and corresponding the obtained Fourier transform spectrum to the delay time corresponding to the interface to obtain an ideal time spectrum decomposition result. In one possible implementation, the expression for per