A post-stack seismic inversion model construction, post-stack seismic inversion method and device

Abstract

The application discloses a post-stack seismic inversion model construction, a post-stack seismic inversion method and device. The method comprises the following steps: constructing a time domain relationship between a reflection coefficient sequence and wave impedance; constructing a frequency domain relationship between the reflection coefficient sequence and the wave impedance according to the time domain relationship and a frequency spectrum of a pulse signal in a frequency domain; constructing a spectrum relationship function of a seismic record spectrum, a reflection coefficient spectrum and a seismic wavelet spectrum according to an extracted seismic wavelet spectrum and the frequency domain relationship; and constructing a post-stack seismic inversion model in the frequency domain by performing symmetrization processing on the spectrum relationship function and increasing a total variation regularization constraint. The wave impedance is predicted in the frequency domain, the thin layer response of the seismic inversion is more obvious, more accurate thin layer wave impedance information can be obtained, and the inversion result is more accurate.

Description

Technical Field

[0001] This invention relates to the field of oil and gas field exploration technology, and in particular to a post-stack seismic inversion model construction, post-stack seismic inversion method and apparatus. Background Technology

[0002] Post-stack impedance inversion technology is an important method in the field of seismic exploration and is widely used in production applications. Impedance inversion technology is often used for lithology identification, fluid prediction, etc.

[0003] Post-stack impedance inversion techniques are often performed in the time domain. In areas with no or few wells, this method typically establishes the relationship between unknown variables (wave impedance) and seismic records through seismic wavelets. At the same time, it optimizes the inversion process using L1 and L2 regularization and model constraints to derive the post-stack impedance. Summary of the Invention

[0004] The inventors of this application have discovered that the L1 regularization constraint method in the prior art often produces sparse solutions, which can identify the interfaces of large strata, but at the same time, the calculation is prone to large errors when the seismic wavelet is not accurately obtained, making it difficult to meet the accuracy of thin-layer prediction; L2 regularization preserves the smooth characteristics of wave impedance and retains more information, but the interface is not clear, and the thin-layer information is submerged in the low-frequency shadow, making thin-layer prediction also difficult; it can be seen that the existing post-stack wave impedance inversion technology has relatively high prediction accuracy but poor prediction effect.

[0005] In view of the above problems, the present invention is proposed to provide a post-stack seismic inversion model construction, post-stack seismic inversion method and apparatus that overcomes or at least partially solves the above problems.

[0006] This invention provides a method for constructing a post-stack seismic inversion model, comprising:

[0007] Construct the time-domain relationship between the reflection coefficient sequence and wave impedance;

[0008] Based on the time-domain relationship and the frequency spectrum of the pulse signal, a frequency-domain relationship between the reflection coefficient sequence and the wave impedance is constructed.

[0009] Based on the extracted seismic wavelet spectrum and the frequency domain relationship, a spectral relationship function between the seismic record spectrum, the reflection coefficient spectrum, and the seismic wavelet spectrum is constructed.

[0010] The spectral relationship function is symmetricized and a total variational regularization constraint is added to construct a post-stack seismic inversion model in the frequency domain.

[0011] In some optional embodiments, constructing the time-domain relationship between the reflection coefficient sequence and the wave impedance includes:

[0012] Construct a nonlinear relationship between a single reflection coefficient and wave impedance;

[0013] Based on the aforementioned nonlinear relationship, a linear relationship between a single reflection coefficient and wave impedance is obtained through equivalent substitution.

[0014] Based on the linear relationship between a single reflection coefficient and wave impedance, a time-domain relationship between the reflection coefficient sequence and wave impedance is constructed.

[0015] In some optional embodiments, the construction of a nonlinear relationship between a single reflection coefficient and wave impedance; based on the nonlinear relationship, obtaining a linear relationship between a single reflection coefficient and wave impedance through equivalent substitution, includes:

[0016] Constructing a nonlinear relationship between a single reflection coefficient and wave impedance:

[0017] Where, r i Z is the reflection coefficient. i Acoustic impedance;

[0018] Using equivalent wave impedance x i For the wave impedance z in the nonlinear relationship i Perform equivalent substitution, where,

[0019] A linear relationship between a single reflection coefficient and wave impedance was established:

[0020] In some optional embodiments, based on the linear relationship between individual reflection coefficients and wave impedance, a time-domain relationship between the reflection coefficient sequence and wave impedance is constructed, including:

[0021] For the reflection coefficient sequence, construct the difference operator matrix:

[0022] Based on the difference operator matrix, the relationship between the reflection coefficient sequence and the wave impedance is constructed:

[0023] Among them, the reflection coefficient sequence wave impedance

[0024] The time-domain relationship between the reflection coefficient sequence and the wave impedance is obtained as: r = D·x.

[0025] In some optional embodiments, based on the time-domain relationship and the frequency-domain spectrum of the pulse signal, a frequency-domain relationship between the reflection coefficient sequence and the wave impedance is constructed, including:

[0026] Based on the spectrum F(δ(τ)) = 1 of the single-pulse signal, the spectrum F(δ(t-τ)) = e^(-τ) of the time-shifted single-pulse signal is obtained. -jωτ ;

[0027] Where F(·) is the Fourier transform, δ(τ) is the single pulse signal, and τ is the time shift;

[0028] Based on the spectrum of the time-shifted signal of the single-pulse signal, the spectrum R of the reflection coefficient sequence is obtained:

[0029]

[0030] Where, r i The reflection coefficients are in the reflection coefficient sequence;

[0031] Based on the time-domain relationship between the reflection coefficient sequence and the wave impedance, and the spectrum of the reflection coefficient sequence, the frequency-domain relationship between the reflection coefficient sequence and the wave impedance is obtained:

[0032]

[0033] Right now

[0034] In some optional embodiments, the step of constructing the spectral relationship function between the seismic record spectrum, the reflection coefficient spectrum, and the seismic wavelet spectrum based on the extracted seismic wavelet spectrum and the frequency domain relationship includes:

[0035] Based on the spectral characteristics of multiple seismic records, the median of the spectrum is extracted across the entire frequency band to obtain the seismic sub-wavelength spectrum;

[0036] Based on the frequency domain relationship between the seismic wavelet spectrum, the reflection coefficient sequence, and the wave impedance, a spectral relationship function between the seismic record spectrum, the reflection coefficient spectrum, and the seismic wavelet spectrum is constructed:

[0037]

[0038] Where S is the frequency spectrum of the earthquake record, and x is the wave impedance.

[0039] In some optional embodiments, the spectral relationship function is symmetrically processed, including...

[0040] The seismic record s is symmetrically processed to obtain the symmetrically processed seismic record s. * :

[0041]

[0042] Among them, s -1 This is the inversion of the earthquake record s;

[0043] Based on the symmetrically processed earthquake record s* Spectrum S * The spectral relationship function is transformed into:

[0044]

[0045] In some optional embodiments, the addition of total variational regularization constraints to construct a frequency-domain post-stack seismic inversion model includes:

[0046] Based on the difference operator matrix, the total variational regularization parameter is obtained:

[0047] By applying a total variational regularization parameter to the spectral relationship function, a post-stack seismic inversion model in the frequency domain is obtained:

[0048]

[0049] This invention also provides a post-stack seismic inversion method, comprising:

[0050] Based on the seismic sub-wave spectrum and seismic record spectrum, input the pre-constructed post-stack seismic inversion model to obtain wave impedance data that makes the post-stack seismic inversion model meet the preset optimization iteration conditions;

[0051] The post-stack seismic inversion model is constructed using the aforementioned post-stack seismic inversion model construction method.

[0052] This invention also provides a post-stack seismic inversion model construction device, comprising:

[0053] The first building module is used to construct the time-domain relationship between the reflection coefficient sequence and the wave impedance;

[0054] The second construction module is used to construct the frequency domain relationship between the reflection coefficient sequence and the wave impedance based on the time domain relationship and the frequency spectrum of the pulse signal.

[0055] The third construction module is used to construct a spectral relationship function between the seismic record spectrum, the reflection coefficient spectrum, and the seismic wavelet spectrum based on the extracted seismic wavelet spectrum and the frequency domain relationship.

[0056] The model building module is used to symmetrize the spectral relationship function, add total variational regularization constraints, and construct a post-stack seismic inversion model in the frequency domain.

[0057] This invention also provides a post-stack seismic inversion device, comprising:

[0058] The aforementioned post-stack seismic inversion model construction device is used to construct a post-stack seismic inversion model;

[0059] The inversion prediction module is used to input the constructed post-stack seismic inversion model based on the seismic sub-wave spectrum and seismic record spectrum, and obtain wave impedance data that makes the post-stack seismic inversion model meet the preset optimization iteration conditions.

[0060] This invention also provides a computer storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described post-stack seismic inversion model construction method and / or the above-described post-stack seismic inversion method.

[0061] This invention also provides a computer device, characterized in that it includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, it implements the above-described post-stack seismic inversion model construction method and / or the above-described post-stack seismic inversion method.

[0062] The beneficial effects of the above-described technical solutions provided in the embodiments of the present invention include at least the following:

[0063] The method described in this invention constructs a time-domain relationship between the reflection coefficient sequence and wave impedance, thereby obtaining a frequency-domain relationship between the reflection coefficient sequence and wave impedance. This leads to the construction of a spectral relationship function between the seismic record spectrum, the reflection coefficient spectrum, and the seismic sub-wave spectrum. The spectral relationship function is then symmetricized, and a total variational regularization constraint is added to construct a post-stack seismic inversion model in the frequency domain. This enables seismic inversion wave impedance prediction in the frequency domain. Compared to existing methods that predict in the time domain, frequency-domain prediction provides a clearer picture of the thin-layer response, resulting in higher accuracy and a prediction result closer to the true wave impedance, thus making the seismic inversion prediction more precise. To facilitate prediction in the frequency domain, symmetricization of the spectral relationship function and the addition of a total variational regularization constraint are used, simplifying the calculation process and reducing the difficulty of prediction.

[0064] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description, claims, and drawings.

[0065] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0066] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0067] Figure 1 This is a spectral analysis diagram showing the similarities and differences between time-domain and frequency-domain deconvolution of thin-layer reflection in an embodiment of the present invention;

[0068] Figure 2 These are spectral feature analysis diagrams of asymmetric and symmetric signals in embodiments of the present invention;

[0069] Figure 3 This is a comparative analysis of the inversion effects of different methods for a simple seismic response model in this embodiment of the invention;

[0070] Figure 4 This is a comparative analysis of the results of different inversion methods for the seismic response model in this embodiment of the invention;

[0071] Figure 5 This is a flowchart of the post-stack seismic inversion model construction method in Embodiment 1 of the present invention;

[0072] Figure 6 This is a flowchart of the post-stack seismic inversion method in Embodiment 2 of the present invention;

[0073] Figure 7 This is a schematic diagram of the structure of the post-stack seismic inversion model construction device in an embodiment of the present invention.

[0074] Figure 8 This is a schematic diagram of the post-stack seismic inversion device in an embodiment of the present invention.

[0075] Figure 9 This is an example diagram of the inversion results of actual earthquake records in an embodiment of the present invention. Detailed Implementation

[0076] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0077] Post-stack seismic inversion is essentially a deconvolution problem, where reflection coefficients are derived from seismic records and extracted seismic wavelets. Currently, numerous methods exist for calculating reflection coefficients, but these coefficients need to be converted into wave impedance information to play a significant role in reservoir prediction. Current inversion methods for thin layers (also known as thin reservoirs) primarily rely on interpolation models based on well data, utilizing relatively little seismic information. This often leads to thin-layer predictions being more model-driven, yielding relatively reliable results in multi-well areas, but proving ineffective in areas with few or no wells.

[0078] Furthermore, since conventional wave impedance inversion techniques calculate wave impedance prediction in the time domain, it is difficult to identify the inversion results if the reflection interfaces are close together. This is also a challenge of thin-layer inversion. To address this, a frequency-domain thin-layer inversion method is proposed. Thin layers that are difficult to identify in the time domain do show significant changes in the frequency domain. By constructing an objective function in the frequency domain, the relationship between wave impedance and the seismic record spectrum is established. At the same time, total variational regularization is used to constrain the abrupt change of wave impedance, resulting in more accurate inversion results and more obvious thin-layer responses.

[0079] The inventors of this application have discovered that the resolution of thin-layer seismic response differs between the time and frequency domains. See also... Figure 1 As shown, by constructing a simple mathematical model, the differences and similarities in the spectral characteristics of thin-layer reflection in the time and frequency domains due to deconvolution can be observed.

[0080] Figure 1 Two reflection coefficient sequences, reflection coefficient 1 and reflection coefficient 2, are constructed. Reflection coefficient 1 consists of two closely spaced impulse responses, representing a thin layer, with a 1 / 2 relationship in magnitude and opposite polarities. Their combined seismic record is shown as composite record 1 in the figure. Reflection coefficient 2 is a single impulse response, representing the equivalent signal of reflection coefficient 1. Its combined seismic record is shown as composite record 2 in the figure. Comparing the composite seismic records of the two, it can be seen that the composite seismic records are almost equal, indicating that thin layers are difficult to predict in the time domain. However, in the frequency domain, a comparison of their spectra shows that the spectra of reflection coefficient 1 and reflection coefficient 2 (the amplitude spectra of reflection coefficient 1 and reflection coefficient 2 in the figure) are completely different and can be clearly distinguished. In other words, the seismic response of thin layers has different resolution capabilities in the time and frequency domains. This invention is based on this and conducts in-depth research.

[0081] Figure 2 For spectral feature analysis of asymmetric and symmetric signals, Figure 2 Signal 1 is an asymmetric signal, and signal 2 is a symmetric signal. Both signals have the same initial signal. Signal 1 has both non-zero real and imaginary parts of its spectrum, meaning it has a non-zero phase. Signal 2 has a non-zero real part and a zero imaginary part, meaning it has a zero-phase signal. Therefore, symmetrical processing of the signals can eliminate the imaginary part of the spectrum.

[0082] Figure 3 To analyze and compare the results processed using time-domain and frequency-domain inversion methods for thin-layer seismic responses, a reflection coefficient sequence was first constructed, including three impulse responses with thicknesses of 3 ms and 2 ms, and magnitudes of 1, 0.5, and 0.5 from top to bottom. (See [link to relevant documentation]). Figure 3 The graph corresponding to the reflection coefficients in the graph; see the graph for the actual wave impedance corresponding to this reflection coefficient sequence. Figure 3The graphs corresponding to the seismic wave impedance; the seismic record synthesized using a Ricker wavelet with a dominant frequency of 35Hz and the reflection coefficient; the wave impedance obtained by processing using the traditional time-domain inversion method, see [link to relevant documentation]. Figure 3 The graph corresponding to the traditional sparse inversion method shows that the intermediate thin layer could not be identified; the wave impedance obtained by processing using the frequency domain inversion method is shown in [reference]. Figure 3 The graphs corresponding to the sparse inversion in the mid-frequency domain show that the frequency domain inversion method can effectively identify the thin-layer response.

[0083] Figure 4 To analyze and compare the results processed using time-domain and frequency-domain inversion methods for seismic response, the left figure shows a random reflection coefficient sequence, which differs from... Figure 2 Here, the reflection coefficient sequence is random and there are many reflection coefficient sequences. The seismic record is synthesized using a Ricker wavelet with a dominant frequency of 35Hz and 20% Gaussian noise is added. The wave impedance is obtained by processing using time domain and frequency domain inversion methods. From the comparison of time domain inversion and true wave impedance, and the comparison of frequency domain inversion and true wave impedance, it can be seen that the frequency domain inversion result matches the true wave impedance better than the time domain inversion result, and the result is more accurate.

[0084] Based on the above analysis, this application proposes constructing an objective function in the frequency domain. Transforming the inversion of thin layers from the time domain to the frequency domain effectively improves the identification capability of thin layers. In conventional time-domain calculations, if reflection interfaces are close together, the inversion results are difficult to identify; this is the challenge of thin-layer inversion. However, in the frequency domain, the spectrum of a thin layer is completely different from that of a single layer, enabling effective identification. This maximizes the use of high-frequency seismic information to recover subsurface impedance information, making the inversion results more consistent with actual subsurface conditions.

[0085] Example 1

[0086] Embodiment 1 of the present invention provides a method for constructing a post-stack seismic inversion model, the process of which is as follows: Figure 5 As shown, it includes the following steps:

[0087] Step S101: Construct the time-domain relationship between the reflection coefficient sequence and the wave impedance.

[0088] Step S102: Based on the time-domain relationship between the reflection coefficient sequence and the wave impedance, and the frequency-domain spectrum of the pulse signal, construct the frequency-domain relationship between the reflection coefficient sequence and the wave impedance.

[0089] Step S103: Based on the extracted seismic wavelet spectrum and the frequency domain relationship between the constructed reflection coefficient sequence and wave impedance, construct the spectral relationship function between the seismic record spectrum, the reflection coefficient spectrum and the seismic wavelet spectrum.

[0090] Step S104: Perform symmetry processing on the spectral relationship function between the constructed seismic record spectrum, reflection coefficient spectrum and seismic wavelet spectrum.

[0091] Step S105: After symmetry processing, the spectral relationship function is subjected to total variational regularization constraints to construct a post-stack seismic inversion model in the frequency domain.

[0092] The process of constructing the time-domain relationship between the reflection coefficient sequence and the wave impedance in step 101 above includes: constructing a nonlinear relationship between a single reflection coefficient and the wave impedance; obtaining a linear relationship between a single reflection coefficient and the wave impedance through equivalent substitution based on the nonlinear relationship between the single reflection coefficient and the wave impedance; and constructing a time-domain relationship between the reflection coefficient sequence and the wave impedance based on the linear relationship between the single reflection coefficient and the wave impedance. Wherein:

[0093] Constructing a nonlinear relationship between a single reflection coefficient and wave impedance:

[0094]

[0095] Where, r i Z is the reflection coefficient. i This is the wave impedance.

[0096] To address the nonlinear relationship between the reflection coefficient and wave impedance, the equivalent wave impedance x is used. i For the wave impedance z in nonlinear relationships i Perform equivalent substitution, where:

[0097]

[0098] Where, x i For z i The equivalent wave impedance can be directly converted.

[0099] Substituting formula (2) into formula (1), we get:

[0100] r i =x i+1 -x i ,Right now

[0101] This establishes a linear relationship between a single reflection coefficient and wave impedance, facilitating the subsequent construction of the objective function.

[0102] The above establishes the relationship between a single reflection coefficient and wave impedance in the time domain, specifically a linear relationship in the time domain. Then, based on the linear relationship between a single reflection coefficient and wave impedance, a time domain relationship between the reflection coefficient sequence and wave impedance is constructed.

[0103] For a sequence of reflection coefficients consisting of multiple reflection coefficients, the difference operator matrix is ​​constructed as follows:

[0104]

[0105] Based on the difference operator matrix, the relationship between the reflection coefficient sequence and the wave impedance is constructed. This relationship can then be transformed into:

[0106]

[0107] in, Here, R indicates that D is a real matrix.

[0108] The time-domain relationship between the reflection coefficient sequence and the wave impedance can be simplified as: r = D·x, thus constructing the objective function for the linear relationship between the reflection coefficient sequence and the wave impedance.

[0109] This step mainly involves establishing the relationship between wave impedance and reflection coefficient. By using equivalent wave impedance substitution, the nonlinear relationship between wave impedance and reflection coefficient is transformed into a linear relationship.

[0110] The process of constructing the frequency domain relationship between the reflection coefficient sequence and the wave impedance in step S102 above includes:

[0111] This step establishes the relationship between the frequency domain reflection coefficient sequence and the reflection coefficient spectrum: based on the spectrum of the single-pulse signal F(δ(τ)) = 1, the spectrum of the time-shifted single-pulse signal F(δ(t-τ)) = e -jωτ Based on the spectrum of the time-shifted signal of the single-pulse signal, the spectrum R of the reflection coefficient sequence is obtained; based on the time-domain relationship between the reflection coefficient sequence and the wave impedance, and the spectrum of the reflection coefficient sequence, the frequency-domain relationship between the reflection coefficient sequence and the wave impedance is obtained. The construction process specifically includes:

[0112] The spectrum of a single-pulse signal is a single value across the entire frequency domain, i.e., F(δ(τ)) = 1.

[0113] Where F(·) is the Fourier transform and δ(τ) is the single pulse signal.

[0114] The time shift of a single-pulse signal varies, and the time-domain signal shift is expressed in the frequency domain as: F(δ(t-τ))=e -jωτ

[0115] Where τ is the time shift.

[0116] The reflection coefficient sequence can be considered as a combination of multiple sparse impulse responses, that is, it includes multiple pulse signals:

[0117]

[0118] Where, τi For layer thickness, r i The reflection coefficients are for different layers.

[0119] Let R represent the spectrum of the reflection coefficient sequence. Based on the spectrum of the time-shifted signal of the single-pulse signal, the reflection coefficient sequence is represented in the frequency domain as follows:

[0120]

[0121] Where, r i Let ω be the reflection coefficient in the reflection coefficient sequence, where ω = 2πf, f is the frequency. The reflection coefficient can have pulse values ​​of different magnitudes, thus constructing the basic framework of the frequency domain reflection coefficient.

[0122] Substituting the derived relationship r = D·x between the reflection coefficient sequence and the wave impedance, we get:

[0123]

[0124] Simplifying the above formula:

[0125]

[0126] This allows us to obtain the frequency domain relationship between the reflection coefficient sequence and the wave impedance based on the time domain relationship between the reflection coefficient sequence and the wave impedance, as well as the spectrum of the reflection coefficient sequence.

[0127] In this step, based on the linear relationship between the reflection coefficient and its spectrum, and considering that the reflection coefficient is composed of multiple pulse signals, each reflection coefficient point can be treated as a time-shifted pulse signal, thus establishing a linear relationship between the reflection coefficient and its spectrum. Substituting the wave impedance into the linear relationship between the reflection coefficient and its spectrum, a linear relationship between the wave impedance and the reflection coefficient spectrum is constructed.

[0128] The process of constructing the spectral relationship function between the seismic record spectrum, the reflection coefficient spectrum, and the seismic wavelet spectrum in step S103 above includes:

[0129] Based on the spectral characteristics of multiple seismic records, the median of the spectrum is extracted across the entire frequency band to obtain the seismic wavelet spectrum. The spectrum of multiple seismic records can be fitted, and the median of each frequency band is extracted to fit the statistical wavelet spectrum W, i.e., the seismic wavelet spectrum W.

[0130] Then, based on the frequency domain relationship between the seismic wavelet spectrum, reflection coefficient sequence, and wave impedance, a spectral relationship function between the seismic record spectrum, reflection coefficient spectrum, and seismic wavelet spectrum is constructed.

[0131]

[0132] Where S is the frequency spectrum of the earthquake record, and x is the wave impedance.

[0133] The construction process of the above objective function can be as follows:

[0134] Since S = W·R, meaning the seismic record spectrum is the product of the reflection coefficient sequence spectrum and the seismic wavelet spectrum, therefore, Since r = D·x, we can further obtain:

[0135]

[0136] According to (7), set τ i Set the value to 1, and construct the spectrum relationship function:

[0137]

[0138] This step extracts the seismic sub-wave spectrum, establishes a bridge between the reflection coefficient sequence and the seismic record, and incorporates it into the wave impedance and reflection coefficient spectrum based on the idea of ​​the convolution model, thus constructing a linear relationship between wave impedance and the seismic record spectrum.

[0139] The process of symmetrizing the spectral relationship function between the constructed seismic record spectrum, reflection coefficient spectrum, and seismic wavelet spectrum in step S104 above includes:

[0140] The aforementioned function relating wave impedance and seismic record spectrum contains complex numbers. To facilitate optimization, this equation is transformed into a workable real number form, as shown in the objective function. Since the data is in complex form, it can be converted into a workable real form by inverting the signal. The processing includes: symmetricizing the seismic record s to obtain the symmetricized seismic record s. * Based on the symmetrically processed seismic record s * Spectrum S * The spectral relationship function is symmetrically processed.

[0141] Transform the seismic record. In signal analysis, the spectrum of a symmetrical signal is a real number. Based on this, the seismic record is optimized as follows: Where s is the earthquake record, s -1 For the inversion of earthquake records, s * For the new earthquake record formed by the two, s * Since it is a symmetrical signal, and the spectrum of a symmetrical signal is a real number, therefore... The imaginary part is zero, and its spectrum S * If are real numbers, then the formula on the right side of the equation is also real numbers. Here, the seismic statistical wavelet is a zero-phase wavelet, so its spectrum W is a real variable, and D and x are both real matrices. The virtual variables in the equation can be omitted, and it is known that: Therefore, among them If omitted, the objective function becomes:

[0142]

[0143] In formula (9), x is the wave impedance. This allows us to achieve the desired result based on the symmetrically processed seismic record s. * Spectrum S * The spectral relationship function is transformed into formula (9). This spectral relationship function is in the real number domain, which facilitates optimization and solution.

[0144] This step addresses the problem that the spectral relationship function cannot be solved due to its complexity by symmetric processing of the seismic signal, eliminating the influence of imaginary numbers and simplifying the solution process.

[0145] The process of adding total variational regularization constraints to the spectral relation function and constructing the post-stack seismic inversion model in the frequency domain in step S105 above includes: obtaining the total variational regularization parameters based on the difference operator matrix: By applying total variational regularization parameters to the spectral relation function, a post-stack seismic inversion model in the frequency domain is obtained.

[0146] The purpose of adding a total variational regularization constraint is to prevent overfitting of the objective function. The total variational regularization constraint is applied by multiplying the variable (wave impedance) by a difference operator matrix before solving, i.e.:

[0147]

[0148] Adding (10) to the transformed spectral relation function yields the following objective function:

[0149]

[0150] Where λ is the equilibrium parameter.

[0151] The objective function can be solved using the Split Bregman optimization algorithm. The optimization iteratively calculates x, which represents the wave impedance, for example, the wave impedance of a thin layer. This iterative optimization algorithm is computationally efficient and accurate, making it suitable for large-scale data computation. This method demonstrates high accuracy in predicting the wave impedance of thin layers.

[0152] The method described in this embodiment enables seismic inversion impedance prediction in the frequency domain. Compared to existing methods that predict in the time domain, frequency domain prediction provides a clearer picture of the thin-layer response, resulting in higher accuracy and a prediction result closer to the true impedance, thus making the seismic inversion prediction more precise. To facilitate prediction in the frequency domain, symmetry processing of the spectral relationship function and the addition of total variational regularization constraints are employed, thereby simplifying the calculation process and reducing the difficulty of prediction.

[0153] Example 2

[0154] Embodiment 2 of the present invention provides a post-stack seismic inversion method, the process of which is as follows: Figure 6 As shown, it includes the following steps:

[0155] Step S201: Construct a post-stack seismic inversion model using the above-described post-stack seismic inversion model construction method.

[0156] Step S202: Based on the seismic sub-wave spectrum and seismic record spectrum, input the pre-constructed post-stack seismic inversion model to obtain wave impedance data that makes the post-stack seismic inversion model meet the preset optimization iteration conditions.

[0157] Based on the statistical analysis of seismic wavelet spectra and the frequency spectrum of seismic records, the wave impedance of different reservoirs can be derived, and even for thin layers, more accurate prediction data of the wave impedance can be obtained.

[0158] The method described in this embodiment, which performs seismic inversion wave impedance prediction in the frequency domain, can effectively identify thin-layer characteristics.

[0159] Based on the same inventive concept, embodiments of the present invention also provide a post-stack seismic inversion model construction device, which can be installed in a computer device, and the structure of the device is as follows. Figure 7 As shown, it includes:

[0160] The first construction module 11 is used to construct the time-domain relationship between the reflection coefficient sequence and the wave impedance;

[0161] The second construction module 12 is used to construct the frequency domain relationship between the reflection coefficient sequence and the wave impedance based on the time domain relationship between the constructed reflection coefficient sequence and the wave impedance and the frequency domain spectrum of the pulse signal.

[0162] The third construction module 13 is used to construct the frequency domain relationship function between the seismic record spectrum, the reflection coefficient spectrum and the seismic sub-wave spectrum based on the extracted seismic sub-wave spectrum and the constructed reflection coefficient sequence and wave impedance.

[0163] Model building module 14 is used to symmetrize the spectral relationship function, add total variational regularization constraints, and construct a post-stack seismic inversion model in the frequency domain.

[0164] Based on the same inventive concept, embodiments of the present invention also provide a post-stack seismic inversion device, which can be installed in a computer device, and the structure of the device is as follows. Figure 8 As shown, it includes:

[0165] The aforementioned post-stack seismic inversion model construction device 1 is used to construct a post-stack seismic inversion model;

[0166] The inversion prediction module 2 is used to input the constructed post-stack seismic inversion model based on the seismic sub-wave spectrum and seismic record spectrum, and obtain wave impedance data that makes the post-stack seismic inversion model meet the preset optimization iteration conditions.

[0167] This invention also provides a computer storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described post-stack seismic inversion model construction method and / or the above-described post-stack seismic inversion method.

[0168] This invention also provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the above-described post-stack seismic inversion model construction method and / or the above-described post-stack seismic inversion method.

[0169] Regarding the apparatus in the above embodiments, the specific manner in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0170] The method and apparatus described in this invention construct an equation relating the reflection coefficient to its spectrum. Simultaneously, based on the relationship between wave impedance and the reflection coefficient, a new equation relating wave impedance to the reflection coefficient spectrum is established. This avoids complex optimization in the frequency domain. Based on the principle that the spectrum of a symmetrical seismic signal is real, the target constraint is reconstructed by inverting the seismic signal and the seismic signal itself. Then, the seismic sub-wave spectrum is extracted, a difference operator matrix is ​​established, and the seismic sub-wave spectrum is substituted into the equation relating wave impedance to the reflection coefficient spectrum to establish a target function relating wave impedance to the seismic record spectrum. To avoid overfitting the target function, a total variational regularization constraint is introduced. The calculated wave impedance is multiplied by the difference operator matrix as a regularization constraint term, ensuring a stepped distribution of the wave impedance and suppressing noise. This method can effectively identify thin layers in the frequency domain, and the total variational regularization constraint prevents overfitting of the wave impedance, resulting in high accuracy of the calculated wave impedance.

[0171] Figure 9To perform frequency domain inversion on actual seismic data, 3D seismic data from a work area in Sichuan Province was used. The upper figure shows the pre-stack time migration profile, and the lower figure shows the post-stack inversion profile processed using the method of this invention. After processing using the method of this invention, a low-resistivity thin layer response appeared at the location marked with reservoir response in Well 1, which is consistent with the well logging interpretation results and can effectively guide the thin layer prediction work. From Figure 9 It can be seen that the method of the present invention can effectively reflect the thin-layer response results, making the thin-layer prediction more accurate.

[0172] Unless otherwise specifically stated, terms such as processing, calculation, operation, determination, display, etc., may refer to the actions and / or processes of one or more processing or computing systems or similar devices that represent the manipulation and conversion of data representing physical (e.g., electronic) quantities within the registers or memory of the processing system into other data similarly representing physical quantities within the memory, registers, or other such information storage, transmission, or display devices of the processing system. Information and signals can be represented using any of a variety of different techniques and methods. For example, data, instructions, commands, information, signals, bits, symbols, and chips mentioned throughout the above description can be represented by voltage, current, electromagnetic waves, magnetic fields or particles, light fields or particles, or any combination thereof.

[0173] It should be understood that the specific order or hierarchy of steps in the disclosed process is an example of an exemplary method. Based on design preferences, it should be understood that the specific order or hierarchy of steps in the process may be rearranged without departing from the scope of this disclosure. The appended method claims provide elements of various steps in an exemplary order and are not intended to limit the scope to the specific order or hierarchy described.

[0174] In the detailed description above, various features are combined together in a single embodiment to simplify this disclosure. This approach to disclosure should not be construed as reflecting an intention that embodiments of the claimed subject matter require more features than are explicitly stated in each claim. Rather, as reflected in the appended claims, the invention is presented with fewer features than all of the features in a single disclosed embodiment. Therefore, the appended claims are hereby explicitly incorporated into the detailed description, wherein each claim stands alone as a preferred embodiment of the invention.

[0175] Those skilled in the art will also understand that the various illustrative logic blocks, modules, circuits, and algorithm steps described in conjunction with the embodiments herein can be implemented as electronic hardware, computer software, or a combination thereof. To clearly illustrate the interchangeability between hardware and software, the various illustrative components, blocks, modules, circuits, and steps described above are generally described in terms of their functionality. Whether such functionality is implemented as hardware or software depends on the specific application and the design constraints imposed on the overall system. Those skilled in the art can implement the described functionality in alternative ways for each specific application; however, such implementation decisions should not be construed as departing from the scope of this disclosure.

[0176] The steps of the methods or algorithms described in conjunction with the embodiments herein can be directly embodied in hardware, software modules executed by a processor, or a combination thereof. The software modules can reside in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disks, removable disks, CD-ROMs, or any other form of storage medium well known in the art. An exemplary storage medium is connected to the processor, enabling the processor to read information from and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and storage medium can reside in an ASIC. The ASIC can reside in a user terminal. Alternatively, the processor and storage medium can exist as discrete components in the user terminal.

[0177] For software implementation, the techniques described in this application can be implemented using modules (e.g., procedures, functions, etc.) that perform the functions described in this application. This software code can be stored in memory units and executed by a processor. The memory units can be implemented within the processor or outside the processor; in the latter case, they are communicatively coupled to the processor via various means, as is well known in the art.

[0178] The foregoing description includes examples of one or more embodiments. It is certainly impossible to describe all possible combinations of components or methods in order to describe the above embodiments, but those skilled in the art will recognize that further combinations and arrangements of the various embodiments are possible. Therefore, the embodiments described herein are intended to cover all such changes, modifications, and variations that fall within the scope of the appended claims. Furthermore, the term "comprising" as used in the specification or claims is interpreted in a manner similar to the term "including," as interpreted when used as a conjunction in the claims. Additionally, the use of any term "or" in the specification of the claims is intended to mean "non-exclusive or."

Claims

1. A method for constructing a post-stack seismic inversion model, characterized in that, include: Based on the pre-constructed difference operator matrix D, the time-domain relationship between the reflection coefficient sequence r and the wave impedance x is constructed: ; Based on the time-domain relationship and the frequency spectrum of the pulse signal, the frequency-domain relationship between the reflection coefficient sequence and the wave impedance is constructed: Where R is the spectrum of the reflection coefficient sequence, This represents the time shift of a single-pulse signal. , For frequency; Based on the extracted seismic wavelet spectrum and the aforementioned frequency domain relationship, a spectral relationship function between the seismic record spectrum, the reflection coefficient spectrum, and the seismic wavelet spectrum is constructed: ;in, This refers to the spectrum of earthquake records. Wave impedance; The spectral relationship function is symmetricized, and a total variational regularization constraint is added to construct a post-stack seismic inversion model in the frequency domain, including: based on the symmetricized seismic record s. Spectrum The spectral relationship function is transformed into: ； Based on the difference operator matrix, the total variational regularization parameter is obtained: ; By applying a total variational regularization parameter to the spectral relationship function, a post-stack seismic inversion model in the frequency domain is obtained: ； in, These are the balancing parameters.

2. The method as described in claim 1, characterized in that, The construction of the time-domain relationship between the reflection coefficient sequence and the wave impedance includes: Construct a nonlinear relationship between a single reflection coefficient and wave impedance; Based on the aforementioned nonlinear relationship, a linear relationship between a single reflection coefficient and wave impedance is obtained through equivalent substitution. Based on the linear relationship between a single reflection coefficient and wave impedance, a time-domain relationship between the reflection coefficient sequence and wave impedance is constructed.

3. The method as described in claim 2, characterized in that, The process involves establishing a nonlinear relationship between a single reflection coefficient and wave impedance; based on this nonlinear relationship, a linear relationship between a single reflection coefficient and wave impedance is obtained through equivalent substitution, including: Constructing a nonlinear relationship between a single reflection coefficient and wave impedance: ; in, The reflection coefficient, Acoustic impedance; Using equivalent wave impedance For the wave impedance in the nonlinear relationship Perform equivalent substitution, where, ; A linear relationship between a single reflection coefficient and wave impedance was established: .

4. The method as described in claim 2, characterized in that, Based on the linear relationship between a single reflection coefficient and wave impedance, a time-domain relationship between the reflection coefficient sequence and wave impedance is constructed, including: For the reflection coefficient sequence, construct the difference operator matrix: ; Based on the difference operator matrix, the relationship between the reflection coefficient sequence and the wave impedance is constructed: Among them, the reflection coefficient sequence wave impedance ; The time-domain relationship between the reflection coefficient sequence and the wave impedance is obtained: .

5. The method as described in claim 1, characterized in that, Based on the time-domain relationship and the frequency spectrum of the pulse signal, a frequency-domain relationship between the reflection coefficient sequence and the wave impedance is constructed, including: Based on the spectrum of a single pulse signal The spectrum of the time-shifted signal of the single-pulse signal is obtained. ; in, For Fourier transform, It is a single pulse signal. It is a time shift quantity; Based on the spectrum of the time-shifted signal of the single-pulse signal, the spectrum R of the reflection coefficient sequence is obtained: ； in, The reflection coefficients are in the reflection coefficient sequence; Based on the time-domain relationship between the reflection coefficient sequence and the wave impedance, and the spectrum of the reflection coefficient sequence, the frequency-domain relationship between the reflection coefficient sequence and the wave impedance is obtained: ； Right now .

6. The method as described in claim 1, characterized in that, Extracting the seismic wavelet spectrum, including: Based on the spectral characteristics of multiple earthquake records, the median of the spectrum is extracted across the entire frequency band to obtain the seismic sub-wavelength.

7. The method as described in claim 1, characterized in that, The seismic record s is symmetrically processed to obtain the symmetrically processed seismic record. : ； in, This is the inversion of the earthquake record s.

8. A post-stack seismic inversion method, characterized in that, include: Based on the seismic sub-wave spectrum and seismic record spectrum, input the pre-constructed post-stack seismic inversion model to obtain wave impedance data that makes the post-stack seismic inversion model meet the preset optimization iteration conditions; The post-stack seismic inversion model is constructed using the post-stack seismic inversion model construction method as described in any one of claims 1-7.

9. A device for constructing a post-stack seismic inversion model, characterized in that, include: The first building module is used to construct the time-domain relationship between the reflection coefficient sequence and the wave impedance; The second construction module is used to construct the frequency domain relationship between the reflection coefficient sequence and the wave impedance based on the time domain relationship and the frequency spectrum of the pulse signal: Where R is the spectrum of the reflection coefficient sequence, This represents the time shift of a single-pulse signal. , For frequency; The third construction module is used to construct a spectral relationship function between the seismic record spectrum, the reflection coefficient spectrum, and the seismic wavelet spectrum based on the extracted seismic wavelet spectrum and the frequency domain relationship: ;in, This refers to the spectrum of earthquake records. Wave impedance; The model building module is used to symmetrize the spectral relationship function, add total variational regularization constraints, and construct a post-stack seismic inversion model in the frequency domain, including: Based on the symmetric processing of earthquake record s Spectrum The spectral relationship function is transformed into: ； Based on the difference operator matrix, the total variational regularization parameter is obtained: ; By applying a total variational regularization parameter to the spectral relationship function, a post-stack seismic inversion model in the frequency domain is obtained: ； in, These are the balancing parameters.

10. A post-stack seismic inversion device, characterized in that, include: The post-stack seismic inversion model construction device as described in claim 9 is used to construct a post-stack seismic inversion model; The inversion prediction module is used to input the constructed post-stack seismic inversion model based on the seismic sub-wave spectrum and seismic record spectrum, and obtain wave impedance data that makes the post-stack seismic inversion model meet the preset optimization iteration conditions.

11. A computer storage medium, characterized in that, The computer storage medium stores computer-executable instructions, which, when executed by a processor, implement the post-stack seismic inversion model construction method according to any one of claims 1-7 and / or the post-stack seismic inversion method according to claim 8.

12. A computer device, characterized in that, include: The memory, the processor, and the computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the post-stack seismic inversion model construction method according to any one of claims 1-7 and / or the post-stack seismic inversion method according to claim 8.

Images