A method, system, device, and readable storage medium for qualitative estimation of seismic attenuation of a sparse time-frequency spectrum

Abstract

This invention discloses a method, system, device, and readable storage medium for qualitative estimation of seismic attenuation using sparse time-spectrum data. The method includes calculating the sparse time-spectrum of seismic data, calculating the Fourier spectrum of the seismic data to obtain high-frequency and low-frequency frequencies, calculating an adjustment factor based on the high-frequency and low-frequency frequencies, applying the high-frequency and low-frequency frequencies to the sparse time-spectrum to obtain high-frequency and low-frequency seismic components, and subtracting the product of the high-frequency seismic component and the adjustment factor from the low-frequency seismic component to obtain the qualitative estimation result of seismic attenuation. Compared to the S-transform in existing technologies, sparse time-spectrum data can restore the dominant frequency of the signal's time-spectrum to its normal value, the accuracy of the extracted single-frequency components is higher, and the obtained qualitative estimation result of seismic attenuation can improve the readability and accuracy of the seismic attenuation profile.

Description

Technical Field

[0001] This invention relates to the field of seismic exploration technology, and in particular to a method, system, device, and readable storage medium for qualitative estimation of seismic attenuation in sparse time-frequency spectra. Background Technology

[0002] Currently, seismic wave analysis remains the primary method for oil and gas exploration. This method leverages the differences in elasticity and density of the subsurface medium, observing and analyzing the Earth's response to artificially generated seismic waves to infer the properties and morphology of subsurface rock strata. Artificially generated elastic waves are used to locate mineral deposits and obtain engineering geological information. Seismic wavelets attenuate during propagation due to the influence of the subsurface medium, manifesting as a decrease in amplitude and dominant frequency. High-frequency components attenuate rapidly, especially after passing through gas-bearing reservoirs. The difference between low-frequency and high-frequency components can be used to qualitatively estimate seismic attenuation, aiding in the accurate characterization of subsurface oil and gas resources. The high-frequency and low-frequency components of seismic data need to be extracted from its time-frequency spectrum.

[0003] Currently, time-frequency analysis tools such as the short-time Fourier transform, wavelet transform, and S-transform are widely used in seismic exploration. Among them, the S-transform has attracted much attention due to its ability to adaptively adjust the time-frequency resolution of low-frequency and high-frequency components of the signal. The Gaussian window applied during the calculation of the S-transform contains... This term, the ST wavelet normalization factor, is originally intended to ensure that the Gaussian window of the S-transform maintains a maximum value of 1 at different frequencies, thus guaranteeing energy conservation after the transform. However, |f| also acts as a filter, linearly related to amplitude and frequency, suppressing low frequencies and enhancing high frequencies, causing spectral distortion and a shift towards higher frequencies. This frequency shift caused by the S-transform can lead to discontinuities and spurious frequencies in the phase axis of the seismic attenuation profile, reducing the readability and accuracy of the seismic attenuation profile. Summary of the Invention

[0004] Therefore, it is necessary to provide a qualitative estimation method, system, device, and readable storage medium for seismic attenuation of sparse time spectrum, so as to solve the technical problem that the main frequency of the time spectrum after S-transformation of the signal will shift to a higher frequency in the prior art.

[0005] This invention provides a qualitative estimation method for seismic attenuation in sparse time-spectrum data, comprising the following steps:

[0006] Calculate the sparse temporal spectrum of seismic data;

[0007] Calculate the Fourier spectrum of the seismic data to obtain the high-frequency and low-frequency frequencies, and calculate the adjustment factor based on the high-frequency and low-frequency frequencies;

[0008] The high-frequency and low-frequency frequencies are respectively used in the sparse time spectrum to obtain high-frequency seismic components and low-frequency seismic components;

[0009] The qualitative estimation result of earthquake attenuation is obtained by subtracting the product of the high-frequency earthquake component and the adjustment factor from the low-frequency earthquake component.

[0010] Furthermore, the step of calculating the sparse temporal spectrum of the seismic data includes:

[0011] Read the two-dimensional stacked seismic data;

[0012] Select the reference and target horizons from the seismic profile data for seismic attenuation estimation;

[0013] The sparse temporal spectrum of the seismic data is calculated channel by channel.

[0014] The sparse time spectrum of a single channel is combined into a three-dimensional sparse time spectrum.

[0015] Furthermore, the step of calculating the sparse temporal spectrum of the seismic data includes:

[0016] Read the two-dimensional stacked seismic data d(x,t) and calculate the UGST time spectrum of d(x,t) for each trace;

[0017] Let the i-th seismic record be denoted as h. i (t), i = 1, 2, ... max(x), where the spectrum at UGST is represented as

[0018]

[0019] The original signal is obtained from its inverse transform.

[0020]

[0021] in It is a Gaussian window function, and the parameters k and p control the width of the Gaussian window function;

[0022] Equation (1) can be expressed in convolution form.

[0023]

[0024] According to the relationship between the time and frequency domains of the Fourier transform, we know that

[0025]

[0026] Transforming the UGST expression from the time domain to the frequency domain, and then discretizing equation (4), we can obtain...

[0027]

[0028] Where d and m represent the time and frequency sampling points, respectively; UGST i [d, m] rearranged into a vector by columns And establish a basis function matrix. Defined as

[0029]

[0030] Equation (4) is Or for

[0031] Solving the inverse problem The objective function is defined as follows:

[0032]

[0033] With the addition of l1 norm regularization terms, the objective function is:

[0034]

[0035] Where λ is the regularization parameter, the sparse solution x can be obtained by solving equation (8). i ;

[0036] Adding l2 norm regularization terms and combining them with l1 norm regularization terms constrains the objective function.

[0037]

[0038] Ψ decomposes into F H W, where W is the kernel function matrix of the inverse UGST transform, F H The inverse Fourier transform basis function matrix is ​​used to decompose Ψ and solve the equation using the Bregman iterative algorithm to obtain the sparse temporal spectrum of the UGST of the i-th seismic record, thus obtaining the sparse temporal spectrum TF. i .

[0039] Furthermore, the step of obtaining the sparse temporal spectrum of the UGST of the i-th seismic record further includes:

[0040] The sparse temporal spectrum of the UGST of the i-th seismic record is calculated for each track, and the sparse temporal spectra of the UGST of the seismic records are combined into a three-dimensional matrix d(x, t, f).

[0041] Furthermore, the step of calculating the sparse temporal spectrum of the seismic data also includes:

[0042] Based on the seismic data d(x,t), calculate the Fourier spectrum D(f) of the seismic data d(x,t);

[0043] Choose D(f) at a high frequency f H and low frequency f L ;

[0044] Based on the known reference layer H R (x), calculate its adjustment factor

[0045]

[0046] Furthermore, the step of calculating its adjustment factor also includes:

[0047] The time spectrum d(x, t, f) of two-dimensional seismic data calculated channel by channel in equation (1);

[0048] The extracted high frequency f H and low frequency f L Substituting d(x, t, f), we obtain the high-frequency single-frequency seismic component d(x, t, f). H ) and low-frequency single-frequency seismic components d(x, t, f) L ).

[0049] Furthermore, the high-frequency single-frequency seismic component d(x, t, f) is obtained. H ) and low-frequency single-frequency seismic components d(x, t, f) L The steps also include:

[0050] Subtracting the product of the high-frequency seismic component and the adjustment factor from the low-frequency seismic component yields a qualitative estimate of seismic attenuation.

[0051] Combining single-frequency seismic components and adjustment factors

[0052] QA(x,t)=d(x,t,f) L )-α(x)·d(x,t,f H (11)

[0053] Where QA(x,t) is the qualitative estimation result of the seismic data attenuation.

[0054] Another embodiment of the present invention discloses a sparse time-spectrum seismic attenuation qualitative estimation system applied to the above-mentioned seismic attenuation qualitative estimation method, comprising:

[0055] The sparse temporal spectrum acquisition module is used to calculate the sparse temporal spectrum of seismic data;

[0056] The adjustment factor acquisition module is used to calculate the Fourier spectrum of seismic data and obtain high-frequency and low-frequency frequencies, and calculate the adjustment factor based on the high-frequency and low-frequency frequencies.

[0057] A high-frequency seismic component and low-frequency seismic component acquisition module is used to apply the high-frequency frequency and the low-frequency frequency to a sparse time spectrum to obtain high-frequency seismic components and low-frequency seismic components.

[0058] The earthquake attenuation qualitative estimation structure acquisition module is used to subtract the product of the high-frequency earthquake component and the adjustment factor from the low-frequency earthquake component to obtain the earthquake attenuation qualitative estimation result.

[0059] Another embodiment of the present invention discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, characterized in that the processor executes the computer program to implement the steps of the above-described qualitative estimation method for seismic attenuation based on a sparse time-frequency spectrum.

[0060] In another embodiment of the present invention, a computer-readable storage medium is disclosed, wherein the computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the above-described qualitative estimation method for seismic attenuation based on a sparse time-spectrum.

[0061] This invention provides a qualitative estimation method for seismic attenuation based on sparse time-spectrum data. It calculates the sparse time-spectrum of seismic data. Compared to the S-transform in existing technologies, this method removes |f| from the wavelet normalization factor in the S-transform, ensuring the dominant frequency of the transformed time-spectrum remains at the correct value. The sparse time-spectrum restores the dominant frequency of the signal's time-spectrum to its normal value. By applying high-frequency and low-frequency components to the sparse time-spectrum, high-frequency and low-frequency seismic components are obtained. The accuracy of the extracted single-frequency components is higher. Furthermore, the product of the high-frequency seismic component and the adjustment factor is subtracted from the low-frequency seismic component to obtain the qualitative estimation result of seismic attenuation. This qualitative estimation result improves the readability and accuracy of the seismic attenuation profile. Attached Figure Description

[0062] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0063] Figure 1 The dominant frequency values ​​extracted by different time-frequency transformation methods for different dominant frequency Ricker wavelets in the embodiments of the present invention are (ST-based: S-transform, UST-based: frequency-preserving S-transform, UGST-based: generalized frequency-preserving S-transform, SUGST-based: the present invention).

[0064] Figure 2The following are the main frequency values ​​extracted by different time-frequency transformation methods for the 30Hz main frequency Ricker wavelet with different noise levels in the embodiments of the present invention (ST-based: S-transform, UST-based: frequency-preserving S-transform, UGST-based: generalized frequency-preserving S-transform, SUGST-based: the present invention).

[0065] Figure 3 The synthesized seismic records in this embodiment of the invention are: (a) reflection coefficient, (b) noise-free synthesized seismic record, and (c) synthesized seismic record containing 5 dB additive white Gaussian noise.

[0066] Figure 4 The time-frequency spectra of noise-free synthetic seismic records in the embodiments of the present invention are: (a) ST, (b) UST, (c) UGST, (d) the present invention.

[0067] Figure 5 The time-frequency spectra of the synthetic seismic records containing 5dB additive white Gaussian noise in the embodiments of the present invention are: (a) ST, (b) UST, (c) UGST, (d) the present invention.

[0068] Figure 6 The two-dimensional synthetic seismic data models in this embodiment of the invention are: (a) a synthetic model of a tight sandstone gas reservoir and (b) a post-stack seismic profile based on the synthetic model in (a).

[0069] Figure 7 The following are low-frequency components of synthetic seismic data extracted at different times in the embodiments of the present invention: (a) ST (b) UST (c) UGST (d) the present invention.

[0070] Figure 8 The high-frequency components of the synthetic seismic data extracted at different times in the embodiments of the present invention are: (a) ST (b) UST (c) UGST (d) the present invention.

[0071] Figure 9 The qualitative attenuation estimation results of synthetic seismic data obtained from different time spectra in the embodiments of the present invention are: (a) ST (b) UST (c) UGST (d) the present invention.

[0072] Figure 10 This is a two-dimensional post-stack seismic profile of an oil and gas field in the Sichuan-Central Paleo-Uplift tectonic belt, as described in an embodiment of the present invention.

[0073] Figure 11 The qualitative attenuation estimation results of actual two-dimensional post-stack seismic data obtained from different time spectra in the embodiments of the present invention are: (a) ST (b) UST (c) UGST (d) the present invention.

[0074] Figure 12This is a schematic diagram of a sparse time-spectrum seismic attenuation qualitative estimation system according to an embodiment of the present invention.

[0075] Figure 13 This is a schematic diagram of a computer device in an embodiment of the present invention.

[0076] Main components:

[0077] 100. Computer equipment; 110. Memory; 120. Processor.

[0078] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0079] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0080] It should be noted that all directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of the present invention are only used to explain the relative positional relationship and movement of each component in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indication will also change accordingly.

[0081] Furthermore, the use of terms such as "first" and "second" in this invention is for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the term "and / or" throughout the text includes three solutions; taking A and / or B as an example, it includes technical solution A, technical solution B, and a technical solution that simultaneously satisfies A and B. Furthermore, the technical solutions of various embodiments can be combined with each other, but this must be based on the ability of a person skilled in the art to implement them. When the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by this invention.

[0082] In some embodiments, a qualitative estimation method for seismic attenuation in a sparse time spectrum includes the following steps:

[0083] S100: Calculate the sparse temporal spectrum of seismic data.

[0084] S200: Calculate the Fourier spectrum of the seismic data and obtain the high-frequency and low-frequency frequencies. Calculate the adjustment factor based on the high-frequency and low-frequency frequencies.

[0085] S300. High-frequency and low-frequency frequencies are used in the sparse time spectrum to obtain high-frequency seismic components and low-frequency seismic components, respectively.

[0086] The qualitative estimation result of seismic attenuation is obtained by subtracting the product of the high-frequency seismic component and the adjustment factor from the S400 and low-frequency seismic components.

[0087] This invention provides a qualitative estimation method for seismic attenuation based on sparse time-spectrum data. It calculates the sparse time-spectrum of seismic data. Compared to the S-transform in existing technologies, the sparse time-spectrum method can restore the dominant frequency of the signal's time-spectrum to its normal value. High-frequency and low-frequency components are used in the sparse time-spectrum to obtain high-frequency and low-frequency seismic components, respectively. The accuracy of the extracted single-frequency components is higher. Furthermore, the qualitative estimation result of seismic attenuation is obtained by subtracting the product of the high-frequency seismic component and the adjustment factor from the low-frequency seismic component. This qualitative estimation result improves the readability and accuracy of the seismic attenuation profile. Moreover, SUGST is used in S100 to calculate the time-spectrum of seismic data, and its time and frequency resolution are higher than those of ST, UST, and UGST. Therefore, the time resolution of the seismic attenuation profile calculated by S400 based on SUGST is also higher. In addition, since SUGST is a sparse time spectrum, it only retains the main time-frequency components of the time spectrum, thus suppressing noise to a certain extent. The noise components in the single-frequency components (high-frequency and low-frequency) extracted by S300 are suppressed, so the noise components in the seismic attenuation profile obtained by S400 are also suppressed, improving the noise resistance of seismic attenuation estimation.

[0088] Specifically, the steps for calculating the sparse temporal spectrum of seismic data include:

[0089] S110, Read the two-dimensional stacked seismic data.

[0090] S120. Select the reference and target layers from the seismic profile data for which seismic attenuation estimation will be performed.

[0091] S130, calculate the sparse temporal spectrum of seismic data channel by channel.

[0092] S140. Combine the sparse time spectrum of a single channel into a three-dimensional sparse time spectrum.

[0093] Furthermore, the steps for calculating the sparse temporal spectrum of seismic data include:

[0094] Read the two-dimensional stacked seismic data d(x,t) and calculate the UGST time spectrum of d(x,t) for each trace;

[0095] Let the i-th seismic record be denoted as h. i(t), i = 1, 2, ... max(x), where the spectrum at UGST is represented as

[0096]

[0097] The original signal is obtained from its inverse transform.

[0098]

[0099] in The Gaussian window function is used, with parameters k and p controlling its width. Since the Gaussian window function of the S-transform cannot be manually adjusted, but only its width adaptively changes according to frequency, its low-frequency resolution is not adjustable, and similarly, its high-frequency resolution is also not adjustable. This invention's qualitative seismic attenuation estimation method introduces k and p parameters to construct a generalized frequency-preserving S-transform (UGST) to adjust its low-frequency time resolution and provides its sparse time-frequency spectrum, SUGST, to further improve the time-frequency resolution. Combining SUGST with the qualitative seismic attenuation estimation method improves the reliability of the qualitative seismic attenuation estimation results.

[0100] Equation (1) can be expressed in convolution form.

[0101]

[0102] According to the relationship between the time and frequency domains of the Fourier transform, we know that

[0103]

[0104] Transforming the UGST expression from the time domain to the frequency domain, and then discretizing equation (4), we can obtain...

[0105]

[0106] Where d and m represent the time and frequency sampling points, respectively; UGST i [d, m] rearranged into a vector by columns And establish a basis function matrix. Defined as

[0107]

[0108] Equation (4) can be rewritten as Or for (because Ψ) H Ψ = I). To obtain the sparse solution x i This requires solving the inverse problem. The objective function is defined as follows:

[0109]

[0110] The equation is clearly an underdetermined problem with multiple solutions. To obtain a unique solution, a regularization term constraint is needed. Since the objective is to find a sparse solution, we first consider the regularization term to be of the l0 norm. However, the optimization problem with the l0 norm regularization term is non-convex, making it inconvenient to solve. Therefore, we can use its equivalent to replace the l1 regularization term. The objective function is...

[0111]

[0112] Where λ is the regularization parameter, the sparse solution x can be obtained by solving equation (8). i To improve the continuity of the sparse solution and reduce its excessive sparsity, an l2 norm regularization term is added, which, together with the l1 norm regularization term, constrains the objective function.

[0113]

[0114] Ψ decomposes into F H W, where W is the kernel function matrix of the inverse UGST transform, F H Ψ is the basis function matrix of the inverse Fourier transform. The equation is solved by decomposing Ψ and using the Bregman iterative algorithm. The Bregman iterative algorithm divides the optimization problem (9) into two sub-optimization problems, as shown in equation (12):

[0115]

[0116] Then, the sparse temporal spectrum of the UGST for the i-th seismic record can be obtained using the alternating iterative algorithm, i.e., the SUGST temporal spectrum TF. i After calculating each channel, these time-frequency spectra are combined into a three-dimensional matrix d(x, t, f).

[0117] Specifically, after obtaining the sparse temporal spectrum of the UGST for the i-th seismic record, the following steps are also included:

[0118] The sparse temporal spectrum of the UGST of the i-th seismic record is calculated for each track, and the sparse temporal spectra of the UGST of the seismic records are combined into a three-dimensional matrix d(x, t, f).

[0119] More specifically, the steps following the calculation of the sparse temporal spectrum of seismic data also include:

[0120] Based on the seismic data d(x,t), calculate the Fourier spectrum D(f) of the seismic data d(x,t), and select the high-frequency f of D(f). H and low frequency f L When making a selection, follow these principles:

[0121] ①f H and f L It must be within the effective frequency band, and a 3dB bandwidth is generally selected as the effective frequency band.

[0122] ②D(f H ) and D(f L They should be roughly equal.

[0123] Based on the known reference layer G R (x), calculate its adjustment factor

[0124]

[0125] Furthermore, the step following the calculation of its adjustment factor also includes:

[0126] The time spectrum d(x, t, f) of two-dimensional seismic data calculated channel by channel in equation (1);

[0127] The extracted high frequency f H and low frequency f L Substituting d(x, t, f), we obtain the high-frequency single-frequency seismic component d(x, t, f). H ) and low-frequency single-frequency seismic components d(x, t, f) L ).

[0128] The high-frequency single-frequency seismic component d(x, t, f) is obtained. H ) and low-frequency single-frequency seismic components d(x, t, f) L The steps also include:

[0129] Subtracting the product of the high-frequency seismic component and the adjustment factor from the low-frequency seismic component yields a qualitative estimate of seismic attenuation.

[0130] Combining single-frequency seismic components and adjustment factors

[0131] QA(x,t)=d(x,t,f) L )-α(x)·d(x,t,f H (11)

[0132] Here, QA(x,t) is the qualitative estimation result of the attenuation of seismic data. The larger the QA(x,t) value, the stronger the regional attenuation.

[0133] like Figure 12 As shown, in another embodiment, a sparse time-spectrum seismic attenuation qualitative estimation system is applied to a seismic attenuation qualitative estimation method, characterized by comprising:

[0134] The sparse temporal spectrum acquisition module is used to calculate the sparse temporal spectrum of seismic data;

[0135] The adjustment factor acquisition module is used to calculate the Fourier spectrum of seismic data and obtain high-frequency and low-frequency frequencies, and calculate the adjustment factor based on the high-frequency and low-frequency frequencies.

[0136] The high-frequency and low-frequency seismic component acquisition module is used to apply the high-frequency and low-frequency frequencies to the sparse time spectrum to obtain the high-frequency and low-frequency seismic components.

[0137] The seismic attenuation qualitative estimation structure acquisition module is used to subtract the product of the high-frequency seismic component and the adjustment factor from the low-frequency seismic component to obtain the seismic attenuation qualitative estimation result.

[0138] like Figure 13 As shown, in another embodiment, a computer device 100 includes a memory 110 and a processor 120. The memory stores a computer program, and the processor executes the computer program to implement the steps of a qualitative estimation method for seismic attenuation based on a sparse time-spectrum.

[0139] In another embodiment, a computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of a qualitative estimation method for seismic attenuation based on a sparse temporal spectrum.

[0140] like Figure 1 As shown, under noise-free conditions and at different clock frequencies, the clock frequency value extracted based on ST is always greater than the actual clock frequency value, while the clock frequency value extracted based on SUGST is basically consistent with the actual clock frequency value. This demonstrates that SUGST's effect on clock frequency recovery is effective under different clock frequency conditions.

[0141] like Figure 2 As shown, at a main frequency of 30Hz and with different signal-to-noise ratios, the main frequency value extracted based on ST is always greater than the true main frequency value, while the main frequency value extracted based on SUGST is basically consistent with the true main frequency value. This indicates that the main frequency recovery effect of SUGST is basically unaffected by noise.

[0142] like Figure 3 As shown, (a) presents three sets of reflection coefficients with intervals of 20ms, 30ms and 40ms respectively; (b) presents the synthetic seismic record obtained by convolving the Ricker wavelet with the reflection coefficient sequence in (a) at a dominant frequency of 30Hz; and (c) presents the noise-added result of the synthetic seismic record obtained in (b).

[0143] like Figure 4As shown, the time spectrum presented in (d) has significantly higher time and frequency resolution than the results in (a) to (c). ST, UST, and UGST all struggle to distinguish the three sets of reflection coefficients, exhibiting varying degrees of coupling. In (c), the UGST result barely distinguishes the positions of the reflection coefficients, but only at high frequencies, and its frequency resolution is severely reduced. In (d), except for the 20ms set which shows slight coupling, the SUGST result clearly identifies the positions of all reflection coefficients.

[0144] like Figure 5 As shown, the same Figure 4 Among the noise-free results, the SUGST result in (d) has higher time-frequency resolution. Only the reflection coefficients in the 20ms group are affected by noise to some extent, causing coupling and making it difficult to distinguish the position of the reflection coefficients. The remaining reflection coefficients are clearly visible. In addition, the time-frequency background noise in (a) to (c) is very obvious, while the time-frequency spectrum of SUGST in (d) only retains the main time-frequency components, and the background noise is basically suppressed.

[0145] like Figure 6 As shown, (a) presents a synthetic model of a tight sandstone gas reservoir with dimensions of [0, 5505 m] × [0, 3505 m], a gas reservoir thickness of 100 m, and a width of 600 m; (b) presents a post-stack seismic profile obtained by forward modeling based on the synthetic model in (a).

[0146] like Figure 7 As shown, the low-frequency seismic profiles (a) to (c) have very low time resolution, making it impossible to distinguish between the two layers. (d) shows a low-frequency seismic profile based on SUGST, where the layer distribution can be identified.

[0147] like Figure 8 As shown, the high-frequency seismic profiles presented in (a) to (c) have very low temporal resolution, making it almost impossible to distinguish between the two layers. (d) shows a high-frequency seismic profile obtained based on SUGST, which can identify the layer distribution, proving that the temporal resolution of the single-frequency seismic profile obtained based on SUGST is superior to the other three methods.

[0148] like Figure 9 As shown, the seismic attenuation profiles obtained by the four methods can all show the distribution of the strongest attenuation layers, the difference being the resolution. The seismic attenuation profiles obtained in (a) to (c) have low resolution, severe interference between upper and lower layers, and the strongest attenuation layer is not prominent enough. In (d), a "bright spot" can be clearly seen at the location indicated by the arrow, which is the layer with the strongest attenuation, i.e., the gas reservoir.

[0149] like Figure 10As shown, Well A is highly productive in layer H2 and Well C is highly productive in layer H1, while the other layers and Well B are low productive in both H1 and H2.

[0150] like Figure 11 As shown, the seismic attenuation profiles obtained by the four methods generally match the well logging results for Well A and Well B, with the main difference being in the Well C results. Only the SUGST-based results in (d) detected strong attenuation at layer H1, as indicated by the arrow. Furthermore, their qualitative attenuation profiles also differ in phase axis continuity and temporal resolution. (a) and (b) exhibit poor phase axis continuity and relatively low temporal resolution. (c) shows improved temporal resolution; for example, a clear phase axis can be seen above H1, but the continuity of the phase axis is not good enough. The qualitative attenuation profile in (d) achieves a balance between temporal resolution and phase axis continuity.

[0151] The above description is merely a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural transformations made using the contents of the present invention's specification and drawings under the inventive concept of the present invention, or direct / indirect applications in other related technical fields, are included within the patent protection scope of the present invention.

Claims

1. A method of qualitative estimation of seismic attenuation of a sparsely sampled time-frequency spectrum, characterized by, The method comprises the following steps: calculating a sparse time-frequency spectrum of seismic data; calculating a Fourier spectrum of the seismic data and obtaining a high frequency and a low frequency, and calculating an adjustment factor according to the high frequency and the low frequency; using the high frequency and the low frequency respectively for the sparse time-frequency spectrum to obtain a high frequency seismic component and a low frequency seismic component; subtracting the product of the high frequency seismic component and the adjustment factor from the low frequency seismic component to obtain a qualitative estimation result of seismic attenuation; The step of calculating a sparse time-frequency spectrum of seismic data comprises: Reading two-dimensional stacked seismic data and computing the UGST time-frequency spectrum of each trace The first The seismic record is denoted by where UGST is the frequency spectrum representation of （1） obtaining the original signal from its inverse transform （2） wherein is a Gaussian window function, the parameter controls the width of the Gaussian window function; expressing formula (1) in a convolution form (3) According to the time domain and frequency domain relationship of Fourier transform (4) The UGST expression is converted from the time domain to the frequency domain, and formula (4) is discretized to obtain (5) where and denote time and frequency sample points, respectively; the is rearranged as a vector by column and a basis function matrix is defined as (6) Formula (5) is or is ; Solving inverse problems the objective function is defined as (7) addition norm regularizer, the objective function is (8) wherein is a regularization parameter, and the sparse solution is obtained by solving equation (8) ; addition norm regularizer, joint norm regularizer constraint objective function (9) decomposed into where is the kernel function matrix of the UGST inverse transform, is the inverse Fourier transform basis function matrix, by decomposing , the Bregman iterative algorithm is used to solve equation (9) to obtain the first sparse time-frequency spectrum of the UGST of the local seismic record, and the sparse time-frequency spectrum is obtained.

2. The method of qualitative estimation of seismic attenuation according to claim 1, characterized in that, The step of calculating a sparse time-frequency spectrum of seismic data comprises: reading two-dimensional stacked seismic data; selecting a reference horizon and a target horizon in seismic profile data to be subjected to seismic attenuation estimation; calculating a sparse time-frequency spectrum of the seismic data channel by channel; combining the sparse time-frequency spectrum of a single channel into a three-dimensional sparse time-frequency spectrum.

3. The method of qualitative estimation of seismic attenuation according to claim 1, characterized in that, The obtained first The step of obtaining the sparse time-frequency spectrum of the seismic record UGST further comprises: calculating the UGST of the seismic record for each trace combining the UGST of the seismic record for each trace into a three-dimensional matrix .

4. The method of qualitative estimation of seismic attenuation according to claim 1, characterized in that, The step of calculating a sparse time-frequency spectrum of seismic data further comprises: Based on seismic data , calculating seismic data Fourier spectrum ; selecting high frequency and low frequency ; based on a known reference horizon , and calculating an adjustment factor thereof （10）。 5. The method of qualitative estimation of seismic attenuation according to claim 4, characterized in that, The step of calculating an adjustment factor further comprises: time-frequency spectrum of two-dimensional seismic data based on the calculation of each trace in equation (1) ; The extracted high frequency frequency and low frequency frequency are substituted into to obtain a high frequency single frequency seismic component and a low frequency single frequency seismic component .

6. The method of qualitative estimation of seismic attenuation according to claim 5, characterized in that, the step of obtaining a high frequency single frequency seismic component and a low frequency single frequency seismic component further comprises subtracting the product of the high frequency seismic component and the adjustment factor from the low frequency seismic component to obtain a qualitative estimation result of seismic attenuation: combining the results obtained in claim 4 and claim 5 （11） wherein is a qualitative estimate of the attenuation of the seismic data.

7. A system for qualitative estimation of seismic attenuation of a sparse time-frequency spectrum, applied to the method for qualitative estimation of seismic attenuation according to any one of claims 1 to 6, characterized by, comprise: a sparse time-frequency spectrum acquisition module for calculating a sparse time-frequency spectrum of seismic data; an adjustment factor acquisition module for calculating a Fourier spectrum of the seismic data and obtaining a high frequency and a low frequency, and calculating an adjustment factor according to the high frequency and the low frequency; a high frequency seismic component and a low frequency seismic component acquisition module for using the high frequency and the low frequency respectively for the sparse time-frequency spectrum to obtain a high frequency seismic component and a low frequency seismic component; a seismic attenuation qualitative estimation structure acquisition module for subtracting the product of the high frequency seismic component and the adjustment factor from the low frequency seismic component to obtain a qualitative estimation result of seismic attenuation.

8. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that, The processor executes the computer program to realize the steps of the sparse time-frequency spectrum seismic attenuation qualitative estimation method of any one of claims 1 to 6.

9. A computer readable storage medium storing a computer program, wherein the computer program is executed by a processor to realize the steps of the sparse time-frequency spectrum seismic attenuation qualitative estimation method of any one of claims 1 to 6.

Images