Well-seismic fusion full-frequency inversion method and apparatus

Abstract

Embodiments of the present application relate to the technical field of oil and gas geophysical exploration, and provide a well-seismic fusion full-frequency inversion method and apparatus. The method comprises: on the basis of second seismic data of a plurality of seismic traces to be inverted and second well logging data of a plurality of candidate sample wells, establishing a high-precision sequence stratigraphic framework of a preset work area; on the basis of the high-precision sequence stratigraphic framework, the second seismic data of said seismic traces, and second near-well trace seismic data of a first number of first sample wells, determining at least two frequency bands; on the basis of second well logging data and the second near-well trace seismic data of the first number of first sample wells, determining an initial full-frequency inversion result of said seismic traces; and on the basis of the high-precision sequence stratigraphic framework and the at least two frequency bands, correcting the initial full-frequency inversion result to obtain a target full-frequency inversion result of said seismic traces. In this way, seismic and well logging full-frequency band information can be comprehensively used to obtain a final inversion result, thereby improving the quality of the inversion result.

Description

Well-seismic fusion full-frequency inversion method and device

[0001] This application claims priority to Chinese Patent Application No. 202411585255.9, filed on November 7, 2024, entitled “Well-Seismic Fusion Full-Frequency Inversion Method and Apparatus”, the entire contents of which are incorporated herein by reference. Technical Field

[0002] This application relates to the field of oil and gas geophysical exploration technology, and in particular to a well-seismic fusion full-frequency inversion method and apparatus. Background Technology

[0003] The essence of seismic inversion is to combine three-dimensional seismic data and three-dimensional well logging data to predict the high-resolution reservoir structure of un-drilled areas within the entire three-dimensional seismic range.

[0004] Currently, commonly used seismic inversion methods include sparse pulse inversion and geostatistical inversion. Sparse pulse inversion is based on a seismic convolution model, obtaining seismic data by convolving P-wave impedance data from well logging data with seismic wavelets. During convolution, high-resolution P-wave impedance data is sampled to the same resolution as the seismic data, thus failing to yield high-resolution vertical inversion results. Geostatistical inversion, through stochastic simulation of well logging data, can obtain high-resolution vertical inversion results. However, the range in geostatistical inversion is typically a fixed constant (e.g., 1000 meters or 800 meters), and the generated subsurface attribute distribution may be insufficient to reflect small-scale lateral variations in the reservoir model structure. In other words, geostatistical inversion cannot obtain high-resolution lateral inversion results. Therefore, neither sparse pulse inversion nor geostatistical inversion yields high-quality inversion results.

[0005] Therefore, obtaining high-quality inversion results has become an urgent technical problem to be solved. Summary of the Invention

[0006] This application provides a well-seismic fusion full-frequency inversion method and apparatus to solve the technical problem of how to obtain high-quality inversion results and improve the quality of the obtained inversion results.

[0007] In a first aspect, embodiments of this application provide a well-seismic fusion full-frequency inversion method, including:

[0008] Acquire first seismic data of multiple seismic traces to be inverted, and first well logging data of multiple wells, wherein the multiple seismic traces to be inverted and the multiple wells are in a preset work area, and the first seismic data of the multiple seismic traces to be inverted includes the first well bypass seismic data of the multiple wells;

[0009] The first seismic data of the plurality of seismic traces to be inverted are filtered to obtain the second seismic data of the plurality of seismic traces to be inverted, wherein the second seismic data of the plurality of seismic traces to be inverted includes the second well bypass seismic data of the plurality of wells;

[0010] The first logging data of the multiple wells is cleaned to obtain the second logging data of the multiple wells;

[0011] Based on the second logging data of the plurality of wells, a plurality of candidate sample wells are determined from the plurality of wells;

[0012] Based on the second seismic data of the multiple seismic traces to be inverted and the second logging data of the multiple sample wells to be selected, a high-precision sequence stratigraphic framework for the preset work area is established.

[0013] For each seismic trace to be inverted, the characteristic distance between the second seismic data of the seismic trace to be inverted and the second well bypass seismic data of each candidate sample well is determined;

[0014] Among the plurality of candidate sample wells, determine the first number of first sample wells with the smallest feature distance;

[0015] Based on the high-precision sequence stratigraphic framework, the second seismic data of the seismic trace to be inverted, and the second well bypass seismic data of the first number of first sample wells, at least two frequency bands are determined.

[0016] Based on the second logging data and the second well bypass seismic data of the first sample wells, the full-frequency initial inversion result of the seismic trace to be inverted is determined;

[0017] Based on the high-precision sequence stratigraphic framework and the at least two frequency bands, the initial full-frequency inversion results are corrected to obtain the full-frequency target inversion results of the seismic trace to be inverted.

[0018] In some embodiments, determining the characteristic distance between the second seismic data of the seismic trace to be inverted and the second well bypass seismic data of each candidate sample well includes:

[0019] Based on the second seismic data of the seismic trace to be inverted, determine the characteristic parameters corresponding to the seismic trace to be inverted;

[0020] For each candidate sample well, the characteristic parameters corresponding to the candidate sample well are determined based on the second well bypass seismic data of the candidate sample well; the distance between the characteristic parameters corresponding to the seismic trace to be inverted and the characteristic parameters corresponding to the candidate sample well is determined as the characteristic distance between the second seismic data of the seismic trace to be inverted and the second well bypass seismic data of the candidate sample well.

[0021] In some embodiments, the characteristic parameters corresponding to the seismic trace to be inverted include at least two of the following: centroid parameter, mean parameter, variance parameter, and variable variance parameter.

[0022] In some embodiments, the centroid parameter satisfies the following formula: Wherein, C represents the centroid parameter, N represents the total number of sampling points in the second seismic data of the seismic trace to be inverted, s(n) represents the sampling point data with index n in the second seismic data of the seismic trace to be inverted, and ∑ represents the summation operation.

[0023] In some embodiments, the average value parameter satisfies the following formula: in, The average value parameter is denoted by s(n), which represents the sampling point data with index n in the second seismic data of the seismic trace to be inverted.

[0024] In some embodiments, the variance parameter satisfies the following formula: Wherein, σ represents the variance parameter.

[0025] In some embodiments, the variance parameter satisfies the following formula: Wherein, D represents the variable variance parameter, r(h) represents the mean of the sum of squares of the differences between the sample data s(n) with index n and the sample data s(n+h) with index (n+h) when there is an interval of (h-1) sample data points, and N1 represents the number of data points in the second seismic data of the seismic trace to be inverted that are spaced apart by (h-1) sample data points.

[0026] In some embodiments, the at least two frequency bands include: a low-frequency band, a seismic effective frequency band, a well logging effective frequency band, and a high-frequency band;

[0027] The determination of at least two frequency bands based on the high-precision sequence stratigraphic framework, the second seismic data of the seismic trace to be inverted, and the second well bypass seismic data of the first number of first sample wells includes:

[0028] Based on the high-precision sequence stratigraphic framework and the second seismic data of the seismic trace to be inverted, the effective low-frequency and effective high-frequency seismic data are determined.

[0029] Select the two target wells with the largest feature distance from the first number of first sample wells;

[0030] Based on the high-precision sequence stratigraphic framework and the second well bypass seismic data of the two target wells, the effective logging frequency is determined;

[0031] Based on the effective low frequency of the seismic earthquake, the effective high frequency of the seismic earthquake, and the effective frequency of the well logging, the first preset frequency band is divided into the low frequency band, the effective frequency band of the seismic earthquake, the effective frequency band of the well logging, and the high frequency band.

[0032] In some embodiments, determining the effective low-frequency and effective high-frequency seismic data based on the high-precision sequence stratigraphic framework and the second seismic data of the seismic trace to be inverted includes:

[0033] Using the high-precision sequence stratigraphic framework, the seismic data corresponding to the target layer is obtained from the second seismic data of the seismic trace to be inverted;

[0034] The seismic data corresponding to the target layer are subjected to spectral scanning to obtain amplitude and frequency information;

[0035] The first frequency corresponding to the preset amplitude in the amplitude frequency information is determined as the effective low frequency of the earthquake, and the second frequency corresponding to the preset amplitude is determined as the effective high frequency of the earthquake, wherein the first frequency is less than the second frequency.

[0036] In some embodiments, determining the effective logging frequency based on the high-precision sequence stratigraphic framework and the second well bypass seismic data from the two target wells includes:

[0037] For each target well, using the high-precision sequence stratigraphic framework, the second-well bypass seismic data corresponding to the target layer of the target well is obtained from the second-well bypass seismic data of the target well; the second-well bypass seismic data corresponding to the target layer is transformed from the time domain to the frequency domain to obtain the transformation result; the frequency domain data corresponding to the effective frequency range is determined from the transformation result.

[0038] Based on the frequency domain data corresponding to the effective frequency range of the two target wells, determine the differential amplitude correlation coefficients corresponding to multiple frequencies in the effective frequency range;

[0039] Based on the differential amplitude correlation coefficients corresponding to the plurality of frequencies, the frequencies that satisfy the first condition among the plurality of frequencies are determined as the effective logging frequencies.

[0040] In some embodiments, the two target wells include a first target well and a second target well, wherein the frequency domain data corresponding to the effective frequency range of the first target well includes a first amplitude corresponding to the plurality of frequencies, and the frequency domain data corresponding to the effective frequency range of the second target well includes a second amplitude corresponding to the plurality of frequencies;

[0041] Based on the frequency domain data corresponding to the effective frequency range of the two target wells, determine the differential amplitude correlation coefficients corresponding to multiple frequencies within the effective frequency range, including:

[0042] For each frequency, a correlation coefficient is determined between the first amplitude and the second amplitude corresponding to that frequency; the difference between the correlation coefficient between the first amplitude and the second amplitude corresponding to that frequency and the correlation coefficient between the first amplitude and the second amplitude corresponding to the next frequency is determined as the differential amplitude correlation coefficient corresponding to that frequency.

[0043] In some embodiments, the first condition is expressed as the following formula: Where w3 represents the effective logging frequency, argmax represents the maximum value operation, w represents the frequency within the effective frequency range, and dρ 12 (w) represents the differential amplitude correlation coefficient corresponding to w, || represents taking the absolute value, and a represents the first preset value.

[0044] In some embodiments, determining the initial full-frequency inversion result of the seismic trace to be inverted based on the second logging data and the second well bypass seismic data of the first sample wells includes:

[0045] Based on the second logging data and second well bypass seismic data of multiple candidate sample wells, the reservoir model structure variation function corresponding to the multiple candidate sample wells is determined;

[0046] Based on the reservoir model structure variation function corresponding to the first number of first sample wells among the plurality of candidate sample wells, the full-frequency initial inversion result of the seismic trace to be inverted is determined.

[0047] In some embodiments, the reservoir model structure variation function corresponding to the Xth candidate sample well among the plurality of candidate sample wells satisfies the following formula: W1*F X (dS)=a X ×(dS) 2 +b X ×(dS)+b X Where W1 represents the seismic wavelet, F X The structure variation function of the reservoir model corresponding to the Xth candidate sample well is represented by *, where * denotes convolution operation, and a X ,b X ,c X Based on the coefficients in the structural variation function of the reservoir model corresponding to the Xth candidate sample well, dS represents the difference data between the second well bypass seismic data of the Xth candidate sample well and the second well bypass seismic data of the Yth candidate sample well among the plurality of candidate sample wells. X and Y are different, and × represents multiplication operation.

[0048] In some embodiments, the full-frequency initial inversion result of the seismic trace to be inverted is determined based on the reservoir model structure variation function corresponding to the first number of first sample wells among the plurality of candidate sample wells, including:

[0049] For each first sample well, the structure variation function of the reservoir model corresponding to the first sample well is solved by Markov chain Monte Carlo stochastic simulation to obtain multiple feasible solutions for the first sample well; the probability of each possible solution is determined by Bayesian analysis.

[0050] The Bayesian model average is performed on the second-highest number of feasible solutions with the highest probability to obtain the full-frequency initial inversion result of the seismic trace to be inverted.

[0051] In some embodiments, the at least two frequency bands include: a low-frequency band, a seismic effective frequency band, a well logging effective frequency band, and a high-frequency band;

[0052] Based on the high-precision sequence stratigraphic framework and the at least two frequency bands, the initial full-frequency inversion results are corrected to obtain the full-frequency target inversion results of the seismic trace to be inverted, including:

[0053] Obtain the first root mean square velocity corresponding to the seismic trace to be inverted;

[0054] Based on the first root mean square velocity corresponding to the seismic trace to be inverted and the second logging data of the multiple candidate sample wells, the first impedance data corresponding to the seismic trace to be inverted is determined.

[0055] Using the high-precision sequence stratigraphic framework, the data corresponding to the low-frequency band is determined in the full-frequency initial inversion result. Using the data corresponding to the low-frequency band in the first impedance data, the data corresponding to the low-frequency band in the full-frequency initial inversion result is corrected to obtain the first corrected full-frequency inversion data.

[0056] Based on the second seismic data of the seismic trace to be inverted, and the second logging data and second well bypass seismic data of the first sample wells of the first number of samples, the second impedance data corresponding to the seismic trace to be inverted is determined.

[0057] Using the high-precision sequence stratigraphic framework, the data corresponding to the effective seismic frequency band is determined in the first corrected full-frequency inversion data. Using the data corresponding to the effective seismic frequency band in the second impedance data, the data corresponding to the effective seismic frequency band in the first corrected full-frequency inversion data is corrected to obtain the second corrected full-frequency inversion data.

[0058] Based on the second seismic data of the plurality of seismic traces to be inverted, determine the third impedance data corresponding to the seismic traces to be inverted;

[0059] Using the high-precision sequence stratigraphic framework, the data corresponding to the effective logging frequency band is determined in the second corrected full-frequency inversion data. Using the data corresponding to the effective logging frequency band in the third impedance data, the data corresponding to the effective logging frequency band in the second corrected full-frequency inversion data is corrected to obtain the third corrected full-frequency inversion data.

[0060] Based on the second seismic data of the seismic trace to be inverted, and the second logging data and second well bypass seismic data of the first sample wells of the first number of samples, the fourth impedance data corresponding to the seismic trace to be inverted is determined.

[0061] Using the high-precision sequence stratigraphic framework, the data corresponding to the effective frequency band of the well logging is determined in the second corrected full-frequency inversion data. Using the data corresponding to the high-frequency band in the fourth impedance data, the data corresponding to the high-frequency band in the third corrected full-frequency inversion data is corrected to obtain the full-frequency target inversion result.

[0062] In some embodiments, determining the first impedance data corresponding to the seismic trace to be inverted based on the first root mean square velocity corresponding to the seismic trace to be inverted and the second logging data of the plurality of candidate sample wells includes:

[0063] For each candidate sample well, the DT in the second logging data of the candidate sample well is converted into the logging layer velocity corresponding to the candidate sample well;

[0064] The logging layer velocities corresponding to the multiple candidate sample wells are interpolated to the logging layer velocities corresponding to the seismic trace to be inverted.

[0065] The first root mean square velocity is subjected to full three-dimensional velocity enhancement processing to obtain the second root mean square velocity corresponding to the seismic trace to be inverted.

[0066] The second root mean square velocity is converted into the first layer velocity corresponding to the seismic trace to be inverted using the Dix formula.

[0067] Based on the logging layer velocity corresponding to the seismic trace to be inverted, the first layer velocity is corrected for the background trend of the well-controlled layer velocity to obtain the second layer velocity corresponding to the seismic trace to be inverted.

[0068] The second layer velocity is converted into the first impedance data corresponding to the seismic trace to be inverted.

[0069] In some embodiments, determining the second impedance data corresponding to the seismic trace to be inverted based on the second seismic data of the seismic trace to be inverted, and the second logging data and second well bypass seismic data of the first number of first sample wells, includes:

[0070] Based on the second seismic data of the seismic trace to be inverted, determine the maximum amplitude spectrum corresponding to the seismic trace to be inverted;

[0071] Based on the maximum amplitude spectrum corresponding to the seismic trace to be inverted, determine the first characteristic parameter corresponding to the seismic trace to be inverted;

[0072] For each first sample well, the maximum amplitude spectrum corresponding to the first sample well is determined based on the second well bypass seismic data of the first sample well, and the second characteristic parameter corresponding to the first sample well is determined based on the maximum amplitude spectrum of the first sample well.

[0073] Based on the first characteristic parameter corresponding to the seismic trace to be inverted and the second characteristic parameter corresponding to the first number of first sample wells, efficient simulated annealing clustering analysis is performed to obtain the second characteristic parameter with the largest characteristic distance from the first characteristic parameter.

[0074] The second logging data of the first sample well corresponding to the second feature parameter with the largest feature distance from the first feature parameter is used to determine the second impedance data corresponding to the seismic trace to be inverted.

[0075] In some embodiments, determining the maximum amplitude spectrum corresponding to the seismic trace to be inverted based on the second seismic data of the seismic trace to be inverted includes:

[0076] Based on the second seismic data of the seismic trace to be inverted, obtain multiple amplitude spectra corresponding to the seismic trace to be inverted;

[0077] Extract multiple maximum amplitudes and the frequencies corresponding to the multiple maximum amplitudes from the multiple amplitude spectra;

[0078] The maximum amplitude and the frequency corresponding to the maximum amplitude are combined in ascending order of frequency to form the maximum amplitude spectrum corresponding to the seismic trace to be inverted.

[0079] In some embodiments, determining the third impedance data corresponding to the seismic trace to be inverted based on the second seismic data of the seismic trace to be inverted includes:

[0080] Based on the second seismic data of the multiple seismic traces to be inverted, the seismic data difference space is determined;

[0081] Obtain the impedance difference space corresponding to the seismic data difference space;

[0082] Obtain the reservoir model structure variation function corresponding to each first sample well;

[0083] Based on the reservoir model structure variation function corresponding to each first sample well, an objective function is constructed.

[0084] Based on the elements in the impedance difference, the objective function is input, and the objective function is optimized to obtain the coefficients of the objective function and the minimum element in the impedance difference space.

[0085] Based on the coefficients of the objective function and the minimum element, the third impedance data corresponding to the seismic trace to be inverted is determined.

[0086] In some embodiments, the third impedance data satisfies the following formula: Where m3 represents the third impedance data, v takes the value of an integer between 1 and the first quantity, and L v M represents the v-th coefficient of the objective function. v F represents the second logging data of the vth first sample well. v Let dY represent the reservoir model structure variation function corresponding to the vth first sample well. x This represents the smallest element.

[0087] In some embodiments, the fourth impedance data corresponding to the seismic trace to be inverted is determined based on the second seismic data of the seismic trace to be inverted, and the second logging data and second well bypass seismic data of the first number of first sample wells.

[0088] For each first sample well, the fifth impedance data is determined based on the second logging data of the first sample well;

[0089] Statistical processing is performed on the fifth impedance data of the first number of first sample wells to obtain the cumulative probability distribution function;

[0090] For each first sample well, the fifth impedance data of the first sample well is subjected to a standard normal transformation according to the cumulative probability distribution function to obtain the first normal distribution impedance data;

[0091] The second well bypass seismic data of the first number of first sample wells and the second seismic data of the seismic trace to be inverted are processed by a collaborative filtering algorithm to determine a number of second sample wells in the first number of first sample wells.

[0092] For each second sample well, Kriging interpolation is performed on the first normal distribution impedance data corresponding to the second sample well to obtain the second normal distribution impedance data; a random sequence is generated through a random residual function; the random sequence is inserted into the second normal distribution impedance data to obtain random simulated impedance data;

[0093] Based on the random simulated impedance data of the multiple second sample wells, determine the random simulated impedance data of the seismic trace to be inverted;

[0094] The fourth impedance data is obtained by inverse transformation of the random simulated impedance data of the seismic trace to be inverted using the cumulative probability distribution function.

[0095] In some embodiments, a high-precision sequence stratigraphic framework for the preset work area is established based on the second seismic data of the plurality of seismic traces to be inverted and the second well logging data of the plurality of sample wells to be selected, including:

[0096] Based on the second seismic data of the multiple seismic traces to be inverted and the second logging data of the multiple sample wells to be selected, well-seismic calibration is performed to obtain well-seismic calibration results;

[0097] The second seismic data of the multiple seismic traces to be inverted are subjected to intelligent fault interpretation to obtain the fault cross-sections.

[0098] Based on the second seismic data of multiple seismic traces to be inverted and the well-seismic calibration results, intelligent interpretation of the seismic horizons is performed to obtain the seismic horizons;

[0099] For the seismic profile images corresponding to the second seismic data of multiple seismic traces to be inverted, perform intelligent identification of small and medium-level sequence layers to obtain the small and medium-level sequence interface information between the seismic layers.

[0100] The high-precision sequence stratigraphic framework is established based on the fault section, the seismic horizon, and the information on the small and medium-level sequence boundaries.

[0101] Secondly, embodiments of this application provide a well-seismic fusion full-frequency inversion device, comprising:

[0102] The acquisition module is used to acquire the first seismic data of multiple seismic traces to be inverted and the first logging data of multiple wells. The multiple seismic traces to be inverted and the multiple wells are in a preset work area. The first seismic data of the multiple seismic traces to be inverted includes the first well bypass seismic data of the multiple wells.

[0103] The filtering module is used to filter the first seismic data of the plurality of seismic traces to be inverted to obtain the second seismic data of the plurality of seismic traces to be inverted, wherein the second seismic data of the plurality of seismic traces to be inverted includes the second well bypass seismic data of the plurality of wells;

[0104] A cleaning module is used to clean the first logging data of the multiple wells to obtain the second logging data of the multiple wells.

[0105] The first determining module is used to determine multiple candidate sample wells from the multiple wells based on the second logging data of the multiple wells;

[0106] A module is established to build a high-precision sequence stratigraphic framework for the preset work area based on the second seismic data of the multiple seismic traces to be inverted and the second well logging data of the multiple sample wells to be selected.

[0107] The second determining module is used to determine the characteristic distance between the second seismic data of the seismic trace to be inverted and the second well bypass seismic data of each candidate sample well for each seismic trace to be inverted;

[0108] The third determining module is used to determine the first number of first sample wells with the smallest feature distance among the plurality of candidate sample wells;

[0109] The fourth determining module is used to determine at least two frequency bands based on the high-precision sequence stratigraphic framework, the second seismic data of the seismic trace to be inverted, and the second well bypass seismic data of the first number of first sample wells;

[0110] The fifth determining module is used to determine the full-frequency initial inversion result of the seismic trace to be inverted based on the second logging data and the second well bypass seismic data of the first sample wells of the first number of sample wells;

[0111] The correction module is used to correct the initial full-frequency inversion results based on the high-precision sequence stratigraphic framework and the at least two frequency bands, so as to obtain the full-frequency target inversion results of the seismic trace to be inverted.

[0112] In some embodiments, the second determining module is specifically used for:

[0113] Based on the second seismic data of the seismic trace to be inverted, determine the characteristic parameters corresponding to the seismic trace to be inverted;

[0114] For each candidate sample well, the characteristic parameters corresponding to the candidate sample well are determined based on the second well bypass seismic data of the candidate sample well; the distance between the characteristic parameters corresponding to the seismic trace to be inverted and the characteristic parameters corresponding to the candidate sample well is determined as the characteristic distance between the second seismic data of the seismic trace to be inverted and the second well bypass seismic data of the candidate sample well.

[0115] In some embodiments, the characteristic parameters corresponding to the seismic trace to be inverted include at least two of the following: centroid parameter, mean parameter, variance parameter, and variable variance parameter.

[0116] In some embodiments, the centroid parameter satisfies the following formula: Wherein, C represents the centroid parameter, N represents the total number of sampling points in the second seismic data of the seismic trace to be inverted, s(n) represents the sampling point data with index n in the second seismic data of the seismic trace to be inverted, and ∑ represents the summation operation.

[0117] In some embodiments, the average value parameter satisfies the following formula: in, The average value parameter is denoted by s(n), which represents the sampling point data with index n in the second seismic data of the seismic trace to be inverted.

[0118] In some embodiments, the variance parameter satisfies the following formula: Wherein, σ represents the variance parameter.

[0119] In some embodiments, the variance parameter satisfies the following formula: Wherein, D represents the variable variance parameter, r(h) represents the mean of the sum of squares of the differences between the sample data s(n) with index n and the sample data s(n+h) with index (n+h) when there is an interval of (h-1) sample data points, and N1 represents the number of data points in the second seismic data of the seismic trace to be inverted that are spaced apart by (h-1) sample data points.

[0120] In some embodiments, the at least two frequency bands include: a low-frequency band, a seismic effective frequency band, a well logging effective frequency band, and a high-frequency band; the fourth determining module is specifically used for:

[0121] Based on the high-precision sequence stratigraphic framework and the second seismic data of the seismic trace to be inverted, the effective low-frequency and effective high-frequency seismic data are determined.

[0122] Select the two target wells with the largest feature distance from the first number of first sample wells;

[0123] Based on the high-precision sequence stratigraphic framework and the second well bypass seismic data of the two target wells, the effective logging frequency is determined;

[0124] Based on the effective low frequency of the seismic earthquake, the effective high frequency of the seismic earthquake, and the effective frequency of the well logging, the first preset frequency band is divided into the low frequency band, the effective frequency band of the seismic earthquake, the effective frequency band of the well logging, and the high frequency band.

[0125] In some embodiments, the fourth determining module is specifically used for:

[0126] Using the high-precision sequence stratigraphic framework, the seismic data corresponding to the target layer is obtained from the second seismic data of the seismic trace to be inverted;

[0127] The seismic data corresponding to the target layer are subjected to spectral scanning to obtain amplitude and frequency information;

[0128] The first frequency corresponding to the preset amplitude in the amplitude frequency information is determined as the effective low frequency of the earthquake, and the second frequency corresponding to the preset amplitude is determined as the effective high frequency of the earthquake, wherein the first frequency is less than the second frequency.

[0129] In some embodiments, the fourth determining module is specifically used for:

[0130] For each target well, using the high-precision sequence stratigraphic framework, the second-well bypass seismic data corresponding to the target layer of the target well is obtained from the second-well bypass seismic data of the target well; the second-well bypass seismic data corresponding to the target layer is transformed from the time domain to the frequency domain to obtain the transformation result; the frequency domain data corresponding to the effective frequency range is determined from the transformation result.

[0131] Based on the frequency domain data corresponding to the effective frequency range of the two target wells, determine the differential amplitude correlation coefficients corresponding to multiple frequencies in the effective frequency range;

[0132] Based on the differential amplitude correlation coefficients corresponding to the plurality of frequencies, the frequencies that satisfy the first condition among the plurality of frequencies are determined as the effective logging frequencies.

[0133] In some embodiments, the two target wells include a first target well and a second target well, wherein the frequency domain data corresponding to the effective frequency range of the first target well includes a first amplitude corresponding to the plurality of frequencies, and the frequency domain data corresponding to the effective frequency range of the second target well includes a second amplitude corresponding to the plurality of frequencies; the fourth determining module is specifically used for:

[0134] For each frequency, a correlation coefficient is determined between the first amplitude and the second amplitude corresponding to that frequency; the difference between the correlation coefficient between the first amplitude and the second amplitude corresponding to that frequency and the correlation coefficient between the first amplitude and the second amplitude corresponding to the next frequency is determined as the differential amplitude correlation coefficient corresponding to that frequency.

[0135] In some embodiments, the first condition is expressed as the following formula: Where w3 represents the effective logging frequency, argmax represents the maximum value operation, w represents the frequency within the effective frequency range, and dρ 12 (w) represents the differential amplitude correlation coefficient corresponding to w, || represents taking the absolute value, and a represents the first preset value.

[0136] In some embodiments, the fifth determining module is specifically used for:

[0137] Based on the second logging data and second well bypass seismic data of multiple candidate sample wells, the reservoir model structure variation function corresponding to the multiple candidate sample wells is determined;

[0138] Based on the reservoir model structure variation function corresponding to the first number of first sample wells among the plurality of candidate sample wells, the full-frequency initial inversion result of the seismic trace to be inverted is determined.

[0139] In some embodiments, the reservoir model structure variation function corresponding to the Xth candidate sample well among the plurality of candidate sample wells satisfies the following formula: W1*F X (dS)=a X ×(dS) 2 +b X ×(dS)+b X Where W1 represents the seismic wavelet, F X The structure variation function of the reservoir model corresponding to the Xth candidate sample well is represented by *, where * denotes convolution operation, and a X ,b X ,c X Based on the coefficients in the structural variation function of the reservoir model corresponding to the Xth candidate sample well, dS represents the difference data between the second well bypass seismic data of the Xth candidate sample well and the second well bypass seismic data of the Yth candidate sample well among the plurality of candidate sample wells. X and Y are different, and × represents multiplication operation.

[0140] In some embodiments, the fifth determining module is specifically used for:

[0141] For each first sample well, the structure variation function of the reservoir model corresponding to the first sample well is solved by Markov chain Monte Carlo stochastic simulation to obtain multiple feasible solutions for the first sample well; the probability of each possible solution is determined by Bayesian analysis.

[0142] The Bayesian model average is performed on the second-highest number of feasible solutions with the highest probability to obtain the full-frequency initial inversion result of the seismic trace to be inverted.

[0143] In some embodiments, the at least two frequency bands include: a low-frequency band, a seismic effective frequency band, a well logging effective frequency band, and a high-frequency band; the correction module is specifically used for:

[0144] Obtain the first root mean square velocity corresponding to the seismic trace to be inverted;

[0145] Based on the first root mean square velocity corresponding to the seismic trace to be inverted and the second logging data of the multiple candidate sample wells, the first impedance data corresponding to the seismic trace to be inverted is determined.

[0146] Using the high-precision sequence stratigraphic framework, the data corresponding to the low-frequency band is determined in the full-frequency initial inversion result. Using the data corresponding to the low-frequency band in the first impedance data, the data corresponding to the low-frequency band in the full-frequency initial inversion result is corrected to obtain the first corrected full-frequency inversion data.

[0147] Based on the second seismic data of the seismic trace to be inverted, and the second logging data and second well bypass seismic data of the first sample wells of the first number of samples, the second impedance data corresponding to the seismic trace to be inverted is determined.

[0148] Using the high-precision sequence stratigraphic framework, the data corresponding to the effective seismic frequency band is determined in the first corrected full-frequency inversion data. Using the data corresponding to the effective seismic frequency band in the second impedance data, the data corresponding to the effective seismic frequency band in the first corrected full-frequency inversion data is corrected to obtain the second corrected full-frequency inversion data.

[0149] Based on the second seismic data of the plurality of seismic traces to be inverted, determine the third impedance data corresponding to the seismic traces to be inverted;

[0150] Using the high-precision sequence stratigraphic framework, the data corresponding to the effective logging frequency band is determined in the second corrected full-frequency inversion data. Using the data corresponding to the effective logging frequency band in the third impedance data, the data corresponding to the effective logging frequency band in the second corrected full-frequency inversion data is corrected to obtain the third corrected full-frequency inversion data.

[0151] Based on the second seismic data of the seismic trace to be inverted, and the second logging data and second well bypass seismic data of the first sample wells of the first number of samples, the fourth impedance data corresponding to the seismic trace to be inverted is determined.

[0152] Using the high-precision sequence stratigraphic framework, the data corresponding to the effective frequency band of the well logging is determined in the second corrected full-frequency inversion data. Using the data corresponding to the high-frequency band in the fourth impedance data, the data corresponding to the high-frequency band in the third corrected full-frequency inversion data is corrected to obtain the full-frequency target inversion result.

[0153] In some embodiments, the correction module is specifically used for:

[0154] For each candidate sample well, the DT in the second logging data of the candidate sample well is converted into the logging layer velocity corresponding to the candidate sample well;

[0155] The logging layer velocities corresponding to the multiple candidate sample wells are interpolated to the logging layer velocities corresponding to the seismic trace to be inverted.

[0156] The first root mean square velocity is subjected to full three-dimensional velocity enhancement processing to obtain the second root mean square velocity corresponding to the seismic trace to be inverted.

[0157] The second root mean square velocity is converted into the first layer velocity corresponding to the seismic trace to be inverted using the Dix formula.

[0158] Based on the logging layer velocity corresponding to the seismic trace to be inverted, the first layer velocity is corrected for the background trend of the well-controlled layer velocity to obtain the second layer velocity corresponding to the seismic trace to be inverted.

[0159] The second layer velocity is converted into the first impedance data corresponding to the seismic trace to be inverted.

[0160] In some embodiments, the correction module is specifically used for:

[0161] Based on the second seismic data of the seismic trace to be inverted, determine the maximum amplitude spectrum corresponding to the seismic trace to be inverted;

[0162] Based on the maximum amplitude spectrum corresponding to the seismic trace to be inverted, determine the first characteristic parameter corresponding to the seismic trace to be inverted;

[0163] For each first sample well, the maximum amplitude spectrum corresponding to the first sample well is determined based on the second well bypass seismic data of the first sample well, and the second characteristic parameter corresponding to the first sample well is determined based on the maximum amplitude spectrum of the first sample well.

[0164] Based on the first characteristic parameter corresponding to the seismic trace to be inverted and the second characteristic parameter corresponding to the first number of first sample wells, efficient simulated annealing clustering analysis is performed to obtain the second characteristic parameter with the largest characteristic distance from the first characteristic parameter.

[0165] The second logging data of the first sample well corresponding to the second feature parameter with the largest feature distance from the first feature parameter is used to determine the second impedance data corresponding to the seismic trace to be inverted.

[0166] In some embodiments, the correction module is specifically used for:

[0167] Based on the second seismic data of the seismic trace to be inverted, obtain multiple amplitude spectra corresponding to the seismic trace to be inverted;

[0168] Extract multiple maximum amplitudes and the frequencies corresponding to the multiple maximum amplitudes from the multiple amplitude spectra;

[0169] The maximum amplitude and the frequency corresponding to the maximum amplitude are combined in ascending order of frequency to form the maximum amplitude spectrum corresponding to the seismic trace to be inverted.

[0170] In some embodiments, the correction module is specifically used for:

[0171] Based on the second seismic data of the multiple seismic traces to be inverted, the seismic data difference space is determined;

[0172] Obtain the impedance difference space corresponding to the seismic data difference space;

[0173] Obtain the reservoir model structure variation function corresponding to each first sample well;

[0174] Based on the reservoir model structure variation function corresponding to each first sample well, an objective function is constructed.

[0175] Based on the elements in the impedance difference, the objective function is input, and the objective function is optimized to obtain the coefficients of the objective function and the minimum element in the impedance difference space.

[0176] Based on the coefficients of the objective function and the minimum element, the third impedance data corresponding to the seismic trace to be inverted is determined.

[0177] In some embodiments, the third impedance data satisfies the following formula: Where m3 represents the third impedance data, v takes the value of an integer between 1 and the first quantity, and L v M represents the v-th coefficient of the objective function. v F represents the second logging data of the vth first sample well. v Let dY represent the reservoir model structure variation function corresponding to the vth first sample well. x This represents the smallest element.

[0178] In some embodiments, the correction module is specifically used for:

[0179] For each first sample well, the fifth impedance data is determined based on the second logging data of the first sample well;

[0180] Statistical processing is performed on the fifth impedance data of the first number of first sample wells to obtain the cumulative probability distribution function;

[0181] For each first sample well, the fifth impedance data of the first sample well is subjected to a standard normal transformation according to the cumulative probability distribution function to obtain the first normal distribution impedance data;

[0182] The second well bypass seismic data of the first number of first sample wells and the second seismic data of the seismic trace to be inverted are processed by a collaborative filtering algorithm to determine a number of second sample wells in the first number of first sample wells.

[0183] For each second sample well, Kriging interpolation is performed on the first normal distribution impedance data corresponding to the second sample well to obtain the second normal distribution impedance data; a random sequence is generated through a random residual function; the random sequence is inserted into the second normal distribution impedance data to obtain random simulated impedance data;

[0184] Based on the random simulated impedance data of the multiple second sample wells, determine the random simulated impedance data of the seismic trace to be inverted;

[0185] The fourth impedance data is obtained by inverse transformation of the random simulated impedance data of the seismic trace to be inverted using the cumulative probability distribution function.

[0186] In some embodiments, the establishment module is specifically used for:

[0187] Based on the second seismic data of the multiple seismic traces to be inverted and the second logging data of the multiple sample wells to be selected, well-seismic calibration is performed to obtain well-seismic calibration results;

[0188] The second seismic data of the multiple seismic traces to be inverted are subjected to intelligent fault interpretation to obtain the fault cross-sections.

[0189] Based on the second seismic data of multiple seismic traces to be inverted and the well-seismic calibration results, intelligent interpretation of the seismic horizons is performed to obtain the seismic horizons;

[0190] For the seismic profile images corresponding to the second seismic data of multiple seismic traces to be inverted, perform intelligent identification of small and medium-level sequence layers to obtain the small and medium-level sequence interface information between the seismic layers.

[0191] The high-precision sequence stratigraphic framework is established based on the fault section, the seismic horizon, and the information on the small and medium-level sequence boundaries.

[0192] Thirdly, embodiments of this application provide an electronic device, including: a memory and a processor;

[0193] The memory stores the instructions that the computer executes;

[0194] The processor executes computer execution instructions stored in memory, causing any of the methods provided in the first aspect to be performed.

[0195] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement any of the methods provided in the first aspect.

[0196] Fifthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the method provided in any of the first aspects.

[0197] This application provides a well-seismic fusion full-frequency inversion method and apparatus. The method includes: acquiring first seismic data of multiple seismic traces to be inverted and first logging data of multiple wells, wherein the multiple seismic traces to be inverted and the multiple wells are located in a preset work area, and the first seismic data of the multiple seismic traces to be inverted includes first well-side seismic data of the multiple wells; filtering the first seismic data of the multiple seismic traces to be inverted to obtain second seismic data of the multiple seismic traces to be inverted, wherein the second seismic data of the multiple seismic traces to be inverted includes second well-side seismic data of the multiple wells; cleaning the first logging data of the multiple wells to obtain second logging data of the multiple wells; determining multiple candidate sample wells from the multiple wells based on the second logging data of the multiple wells; establishing a high-precision sequence stratigraphic framework of the preset work area based on the second seismic data of the multiple seismic traces to be inverted and the second logging data of the multiple candidate sample wells; and performing a full-frequency inversion for each seismic trace to be inverted. The method involves determining the characteristic distance between the second seismic data of the seismic trace to be inverted and the second well-side seismic data of each candidate sample well; identifying the first number of first sample wells with the smallest characteristic distance among the candidate sample wells; determining at least two frequency bands based on the high-precision sequence stratigraphic framework, the second seismic data of the seismic trace to be inverted, and the second well-side seismic data of the first number of first sample wells; determining the initial full-frequency inversion result of the seismic trace to be inverted based on the second logging data and the second well-side seismic data of the first number of first sample wells; and correcting the initial full-frequency inversion result based on the high-precision sequence stratigraphic framework and the at least two frequency bands to obtain the full-frequency target inversion result of the seismic trace to be inverted. This method of frequency band correction of the initial full-frequency inversion model systematically integrates and optimizes the seismic data of each frequency band, thereby improving the accuracy of the final full-frequency target inversion result and enabling a more comprehensive and accurate reflection of the underground geological structure. Attached Figure Description

[0198] To more clearly illustrate the technical solutions in this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0199] Figure 1 is one of the flowcharts of the well-seismic fusion full-frequency inversion method provided in the embodiments of this application;

[0200] Figure 2 is a schematic diagram showing the distribution of multiple seismic traces to be inverted and multiple wells in the preset work area provided in the embodiment of this application;

[0201] Figure 3 shows the seismic data profiles before and after global noise attenuation provided in the embodiments of this application.

[0202] Figure 4 shows the seismic data profiles before and after local strong noise attenuation provided in the embodiments of this application.

[0203] Figure 5 shows the acoustic curves before and after compaction trend correction provided in the embodiments of this application;

[0204] Figure 6 shows the density curves before and after compaction trend correction provided in the embodiments of this application;

[0205] Figure 7 is a histogram of the unnormalized natural gamma data provided in the embodiments of this application;

[0206] Figure 8 is a histogram of normalized natural gamma data provided in the embodiments of this application;

[0207] Figure 9 is a cross-plot of volumetric density logging and P-wave velocity curves provided in an embodiment of this application;

[0208] Figure 10 shows the logging data before and after rock physical environment correction provided in the embodiments of this application.

[0209] Figure 11 is a flowchart of the method for establishing the high-precision sequence stratigraphic framework provided in an embodiment of this application;

[0210] Figure 12 is one of the schematic diagrams of well vibration calibration effect provided in an embodiment of this application;

[0211] Figure 13 is a flowchart of intelligent tomography interpretation provided in an embodiment of this application;

[0212] Figure 14 is a cross-sectional tomographic interpretation diagram provided in an embodiment of this application;

[0213] Figure 15 is a three-dimensional spatial display effect diagram of the fault provided in the embodiment of this application;

[0214] Figure 16 is a schematic diagram of the layer intelligent interpretation process provided in the embodiment of this application;

[0215] Figure 17 is a cross-sectional view illustrating the layer interpretation provided in an embodiment of this application;

[0216] Figure 18 is a three-dimensional stereoscopic display effect diagram of the layer interpretation provided in the embodiment of this application;

[0217] Figure 19 is a seismic phase body profile provided in an embodiment of this application;

[0218] Figure 20 is a cross-section of an inclined body provided in an embodiment of this application;

[0219] Figure 21 is a schematic diagram of the high-precision sequence stratigraphic framework provided in the embodiments of this application;

[0220] Figure 22 is a schematic flowchart of the method for determining feature distance provided in an embodiment of this application;

[0221] Figure 23 is a schematic diagram of obtaining the correlation coefficient based on Euclidean distance according to an embodiment of this application;

[0222] Figure 24 is a schematic diagram of obtaining the correlation coefficient based on the dynamic time planning algorithm provided in the embodiment of this application;

[0223] Figure 25 is a statistical chart of the correlation coefficients between the channel to be inverted and 39 candidate sample wells provided in the embodiment of this application;

[0224] Figure 26 is a schematic flowchart of a method for determining at least two frequency bands provided in an embodiment of this application;

[0225] Figure 27 is an amplitude-frequency information diagram obtained by performing a spectral scan on the seismic data corresponding to the target layer according to an embodiment of this application.

[0226] Figure 28 is a schematic diagram of the four frequency bands provided in the embodiments of this application;

[0227] Figure 29 is a schematic flowchart of the method for determining the initial inversion result of the full frequency according to an embodiment of this application;

[0228] Figure 30 is a schematic diagram of the seismic waveform structure and reservoir model structure of well 1 and well 2 provided in the embodiments of this application;

[0229] Figure 31 is a schematic diagram of the well logging data of the final seismic trace to be inverted obtained by averaging the probabilities of the six feasible solutions using a Bayesian model according to an embodiment of this application.

[0230] Figure 32 shows the well logging data of the seismic trace to be inverted provided in an embodiment of this application;

[0231] Figure 33 shows the seismic waveform of the seismic trace to be inverted provided in an embodiment of this application;

[0232] Figure 34 is a cross-sectional view of the full-frequency initial inversion model of the preset work area provided in the embodiment of this application;

[0233] Figure 35 is a schematic flowchart of the method for correcting the initial inversion result of the full frequency provided in an embodiment of this application;

[0234] Figure 36 is a resolution diagram of the first root mean square velocity provided in an embodiment of this application;

[0235] Figure 37 is a resolution diagram of the second root mean square velocity provided in an embodiment of this application;

[0236] Figure 38 is a cross-sectional view of the first root mean square velocity provided in an embodiment of this application;

[0237] Figure 39 is a cross-sectional view of the second root mean square velocity provided in an embodiment of this application;

[0238] Figure 40 is a plan view of the first root mean square velocity provided in an embodiment of this application;

[0239] Figure 41 is a plan view of the second root mean square velocity provided in an embodiment of this application;

[0240] Figure 42 is a comparison diagram of the logging layer velocity corresponding to the seismic trace to be inverted and the first layer velocity provided in the embodiment of this application;

[0241] Figure 43 is a comparison diagram of the logging layer velocity and the second layer velocity corresponding to the seismic trace to be inverted provided in the embodiment of this application;

[0242] Figure 44 is a cross-sectional comparison diagram of the logging layer velocity and the first layer velocity provided in the embodiment of this application;

[0243] Figure 45 is a cross-sectional comparison diagram of the logging layer velocity and the second layer velocity provided in the embodiment of this application;

[0244] Figure 46 is a schematic diagram of the low-frequency band data in the first impedance data provided in the embodiment of this application;

[0245] Figure 47 is a schematic diagram of the first corrected full-frequency inversion data provided in an embodiment of this application;

[0246] Figure 48 is a schematic diagram of the data corresponding to the effective seismic frequency band in the second impedance data provided in the embodiments of this application;

[0247] Figure 49 is a schematic diagram of the second corrected full-frequency inversion data provided in an embodiment of this application;

[0248] Figure 50 is a schematic diagram of the data corresponding to the effective frequency band of well logging in the third impedance data provided in the embodiments of this application;

[0249] Figure 51 is a schematic diagram of the third corrected full-frequency inversion data provided in an embodiment of this application;

[0250] Figure 52 is a schematic diagram of the high-frequency band data in the fourth impedance data provided in the embodiments of this application;

[0251] Figure 53 is a schematic diagram of the fourth frequency inversion data provided in the embodiment of this application;

[0252] Figure 54 is a schematic diagram of the well-seismic fusion full-frequency inversion device provided in the embodiment of this application;

[0253] Figure 55 is a schematic diagram of the structure of the electronic device provided in the embodiment of this application.

[0254] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0255] In this application, the term "comprising" and its variations may refer to a non-limiting inclusion. The terms "first," "second," etc., are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. The term "at least one" refers to one or more. "More than one" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist; for example, A and / or B can represent: A alone, A and B simultaneously, or B alone. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. Words such as "exemplary" or "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as "exemplary" or "for example" in this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0256] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0257] Figure 1 is a schematic flowchart of one of the well-seismic fusion full-frequency inversion methods provided in this application. As shown in Figure 1, the method includes:

[0258] S101. Obtain the first seismic data of multiple seismic traces to be inverted, and the first logging data of multiple wells. The multiple seismic traces to be inverted and the multiple wells are in a preset work area. The first seismic data of the multiple seismic traces to be inverted includes the first well-side seismic data of the multiple wells.

[0259] Well logging data typically includes one or more of the following: sonic transit time (DT), resistivity (RT), density compensated (DEN), and natural gamma ray (GR). Because logging tools directly contact the wellbore, they can capture very fine formation variations, thus providing rock-physical properties with higher vertical resolution than seismic data. Well logging data can provide high-resolution reservoir model structure and geological information at the well location. It includes various rock-physical parameters that accurately describe reservoir characteristics such as lithology, porosity, and fluid content.

[0260] Seismic data consists of reflected waveforms of seismic signals, possessing specific frequencies and coherence related to geological interfaces.

[0261] Seismic data can be obtained through 3D seismic exploration. Seismic data is three-dimensional and can provide subsurface information covering a wide area. When seismic waves propagate underground, each reflection event represents the average response of a segment of strata, rather than the response of a single interface. Therefore, its vertical resolution is lower than that of well logging data acquired through well logging tools that directly contact the strata. This information has higher lateral resolution but lower vertical resolution.

[0262] Seismic data can reveal geological structures and continuity over large areas, but it usually cannot directly provide detailed information on reservoir structure and rock physical properties.

[0263] In this embodiment of the application, high-resolution well logging data in the vertical direction is used to guide and correct low-resolution seismic data in the vertical direction in order to construct more accurate inversion results.

[0264] The distribution of multiple seismic traces to be inverted and multiple wells in the preset work area is described below with reference to Figure 2. Figure 2 is a schematic diagram of the distribution of multiple seismic traces to be inverted and multiple wells in the preset work area provided in the embodiment of this application. As shown in Figure 2, the multiple seismic traces to be inverted are densely distributed in the preset work area, and the interval between the seismic traces to be inverted can be, for example, 25 meters. The multiple wells are sparsely distributed in the preset work area, and the interval between the wells can be, for example, several hundred meters or several kilometers.

[0265] S102. Filter the first seismic data of multiple seismic traces to be inverted to obtain the second seismic data of multiple seismic traces to be inverted. The second seismic data of multiple seismic traces to be inverted includes the second well bypass seismic data of multiple wells.

[0266] Optionally, methods 1 to 4 can be used to obtain second seismic data for multiple seismic traces to be inverted.

[0267] Method 1 involves using a global noise attenuation filter to perform global noise attenuation on the first seismic data of multiple seismic traces to be inverted, thereby obtaining the second seismic data of multiple seismic traces to be inverted. Figure 3 shows the seismic data profiles before and after global noise attenuation provided in an embodiment of this application.

[0268] The goal of global noise attenuation is to reduce the overall noise data (or background noise data) in the first seismic data. Noise data includes one or more of the following: random noise, environmental noise, equipment noise, etc., in order to improve the signal-to-noise ratio of the seismic data, making the interpretation of the second seismic data more reliable, so as to more clearly identify and interpret geological structures.

[0269] Optionally, the global noise attenuation filter can be designed based on one or more of the following information: data difference information, preset adaptive algorithm, and filter parameters.

[0270] Optionally, the method for obtaining data difference information includes: for at least one seismic trace to be inverted among multiple seismic traces, the first seismic data of the seismic trace to be inverted is transformed from the time domain to the frequency domain by Discrete Fourier Transform (DFT) or Fast Fourier Transform (FFT) to obtain frequency domain seismic data.

[0271] Determine the power spectral density (PSD) based on frequency domain seismic data;

[0272] Spectral analysis of the power spectral density yields valid and noise data;

[0273] Determine the differences between valid and noisy data;

[0274] The data difference information includes the difference between valid data and noisy data in the first seismic data of at least one seismic trace to be inverted.

[0275] Optionally, in the process of designing global noise attenuation, different types of global noise attenuation can be designed and applied based on data difference information to achieve global noise attenuation of the first seismic data of multiple seismic traces to be inverted, so as to retain as much effective data as possible in the first seismic data.

[0276] Global noise attenuation can be achieved using a Wiener filter.

[0277] Optionally, to achieve more refined global noise attenuation, the first seismic data can be decomposed into multiple frequency range segments (sub-bands), and global noise attenuation can be performed on the seismic data within each sub-band to optimize the overall signal processing effect. Alternatively, a filter bank can be used to decompose the first seismic data into multiple sub-bands. Each sub-band can be processed independently based on its signal-to-noise ratio and other characteristics. This allows for different processing strategies to be applied to sub-bands with different frequency contents and noise characteristics; within each sub-band, more refined filtering, such as adaptive filtering, can be applied to more effectively remove noise.

[0278] Pre-defined adaptive algorithms include, for example, the Least Mean Square (LMS) algorithm and the Recursive Least Squares (RLS) algorithm. The LMS algorithm can automatically adapt to different statistical characteristics of the first seismic data, thereby effectively reducing noise in the first seismic data.

[0279] Filter parameters may include one or more of the following: cutoff frequency, filter bandwidth, transition band (slope), stopband and passband attenuation.

[0280] Optionally, a denoising evaluation metric corresponding to global noise attenuation can be established. This denoising evaluation metric may include, for example, signal-to-noise ratio (SNR), coherence, waveform fidelity, and objective quality evaluation parameters.

[0281] Signal-to-noise ratio (SNR) represents the ratio of the power of valid data to the power of noisy data in the second seismic dataset. A high SNR indicates that the valid data is clearer and the noisy data is less.

[0282] Coherence characterizes the similarity between the second seismic data of two adjacent seismic traces to be inverted. High coherence indicates better consistency between the second seismic data of two adjacent seismic traces to be inverted, which is an important goal in seismic data processing.

[0283] Waveform fidelity characterizes the similarity between the first and second seismic data of the seismic trace to be inverted. It is used to ensure that no distortion or loss of important geological information is introduced during the processing.

[0284] Objective quality evaluation parameters may include one or more of the following: mean-square error (MSE) or structural similarity (SSIM).

[0285] Optionally, the MSE of the first seismic data and the MSE of the second seismic data of the seismic trace to be inverted can be determined. When the difference between the MSE of the first seismic data and the MSE of the second seismic data is less than or equal to the MSE threshold, the global noise attenuation effect is considered to be good.

[0286] Optionally, the SSIM of the first seismic data and the SSIM of the second seismic data to be inverted can be determined. When the difference between the SSIM of the first seismic data and the SSIM of the second seismic data is less than or equal to the SSIM threshold, the global noise attenuation effect is considered to be good.

[0287] Method 2 involves using a matched filter to perform local noise attenuation on the first seismic data of multiple seismic traces to be inverted, resulting in the second seismic data of the same seismic traces. Figure 4 shows the seismic data profiles before and after local strong noise attenuation, as provided in an embodiment of this application.

[0288] Local noise attenuation refers to the process where, in certain areas, the initial seismic data may appear weak due to seismic wave attenuation or the complexity of geological conditions. Performing local noise attenuation can enhance the initial seismic data in these areas, aiding in the identification of geological features.

[0289] Optionally, the coefficients of the matched filter can be calculated by designing a geological model framework that matches the desired signal.

[0290] Within the geological model framework, a detailed geological model is created based on geological information (including but not limited to lithology, stratigraphic interfaces, fault distribution, etc.) and geophysical data (including but not limited to the first seismic data of multiple seismic traces to be inverted and the first well logging data of multiple wells). The geological model includes the spatial distribution of parameters such as velocity and density.

[0291] Seismic forward modeling technology is used to synthesize seismic data based on geological models. The synthesized seismic data is then processed (e.g., including but not limited to phase correction and / or amplitude adjustment) to obtain a reference waveform template.

[0292] The coefficients of the matched filter are obtained by convolving the reference waveform template with the autocorrelation function of the desired signal.

[0293] A proper template can significantly improve the performance of matched filters, ensuring that the characteristics of valid data are highlighted while noise and unwanted data components are suppressed. In cases where valid data is weak and the signal-to-noise ratio is low in the initial seismic data, it is difficult to extract a valid signal "template" from the actual seismic signal. In such situations, a synthetic theoretical waveform template can be used as a reference waveform template.

[0294] Optionally, after local noise attenuation, the improvement in signal-to-noise ratio and / or whether there is excessive signal distortion can be evaluated.

[0295] Optionally, the matched filter can be iteratively optimized based on the evaluation results after local noise attenuation, for example, by adjusting the coefficients, time window, or application range of the matched filter.

[0296] Method 3 involves sequentially performing global noise attenuation and local noise attenuation on the first seismic data of multiple seismic traces to be inverted, thereby obtaining the second seismic data of multiple seismic traces to be inverted.

[0297] Method 4 involves sequentially performing local and global noise attenuation on the first seismic data of multiple seismic traces to be inverted, thereby obtaining the second seismic data of multiple seismic traces to be inverted.

[0298] S103. Clean the first logging data of multiple wells to obtain the second logging data of multiple wells.

[0299] Well logging data is collected during geophysical exploration to describe the physical and chemical properties of rocks downhole. These curves are typically obtained by inserting different types of logging tools into the borehole, which can measure a variety of parameters related to rock and fluid properties.

[0300] Common problems in well logging data cleaning include: missing data, instrument effects, environmental influences, and data noise.

[0301] Data missing: Data in certain intervals may be missing due to measurement or recording issues, requiring appropriate estimation or imputation.

[0302] Instrument effect: Measurement results from different instruments may have systematic biases, which need to be processed according to standards.

[0303] Environmental impacts: Changes in environmental factors such as temperature and pressure may affect the readings of well logging data and require correction.

[0304] Measurement error: Improper operation or equipment failure may lead to erroneous data readings, which need to be identified and corrected by comparing multiple measurement results.

[0305] Data noise: The heterogeneity of the formation itself or external factors such as drilling fluid intrusion can introduce noise, which needs to be reduced through filtering or other methods.

[0306] Optionally, S103 includes: S1031 to S1033.

[0307] S1031 integrates the first logging data from multiple wells.

[0308] Optionally, S1031 includes: reviewing the data range, confirming whether the collected data covers the first logging data of all wells in the predetermined target area, and aggregating data from different sources (such as different wells, different regions, and different time periods);

[0309] If the file format of the first logging data is not a common file format, convert the file format of the first logging data to a common file format to ensure that the first logging data of multiple wells adopts a unified format.

[0310] Depth calibration is performed on the first logging data of multiple wells to ensure that the first logging data of multiple wells correspond at the same depth point.

[0311] Common file formats include, for example, LAS format.

[0312] S1032 standardizes the logging data from multiple integrated wells.

[0313] Seismic reservoir inversion is a complex task, where the quality of well logging and core data directly impacts the reliability of the evaluation results. Therefore, quality control and preprocessing steps are crucial for improving data quality, ensuring data consistency and comparability, and ultimately establishing accurate well logging interpretation models. Standardization of well logging data is essential, especially when data comes from different years, instruments, and operators. The purpose of standardization is to ensure fair comparison between data, thereby enabling accurate interpretation of reservoir geological characteristics. Data correction and standardization typically consider two approaches: depth-dependent compaction trend correction methods and depth-independent histogram methods.

[0314] Optionally, for the method of correcting compaction trend with depth, S1032 includes:

[0315] Based on the integrated logging data from multiple wells, logging curves for multiple wells are plotted on the same depth scale.

[0316] Extract the logging curves corresponding to the standard layer from the logging curves of multiple wells;

[0317] For well logging curves that do not meet the pre-defined change conditions in the standard layer, compaction trend correction is performed.

[0318] The selection of standard layers is crucial. Standard layers should be selected from strata that are prevalent throughout the region, have stable lithology, and are relatively thick. Their electrical characteristics should be obvious and easily identifiable in different wells.

[0319] Standard layers include, for example, pure mudstone layers, dense limestone layers, gypsum-salt rock layers, or coal seams.

[0320] Figure 5 shows the acoustic curves before and after compaction trend correction provided in the embodiments of this application.

[0321] Figure 6 shows the density curves before and after compaction trend correction provided in the embodiments of this application.

[0322] It should be noted that when the first logging data is sonic data, the sonic curve shown in Figure 5 can be obtained. Similarly, when the first logging data is density data, the density curve shown in Figure 6 can be obtained. Figures 5 and 6 present depth cross-plots of the sonic and density curves before and after compaction trend correction, which is particularly important when the formation has a large range of burial depth variations. Compaction alters the porosity and elastic properties of the rock, which in turn affects the sonic and density measurements.

[0323] It should be noted that Figures 5 and 6 are illustrated using multiple wells with well numbers including WY-1, WY-4, WY-5, WY-6, WY-8, WZ11-1-2, WZ11-2-2, WZ11-1-5, WZ11-6-1, WZ11-6-1Sa, WZ11-6-2, WZ11-6-4, WAN2, WAN4, WAN9, WZ10-3-1, WZ10-3-3, WZ10-3-15, WZ10-6S-1, WZ11-6-7d, WZ11-6-7da, and WZ11-6-3d as examples.

[0324] By correcting the compaction trend of logging data from different wells, more consistent data can be obtained. For example, the DT correction is between -5 and 5 (µs / ft), and the DEN correction is between -0.02 and 0.03 (g / cm³). 3 Compaction trend correction helps to more accurately assess reservoir characteristics.

[0325] By performing statistical analysis on the standardized correction values ​​for each well in Table 1, we can understand the general trend of the correction process and any possible anomalies.

[0326] Table 1. Statistics of Well Logging Data Correction Quantities

[0327] Optionally, for the histogram method that does not change with depth, when the logging data is natural gamma data, S1032 includes: determining a standard well among multiple wells; and normalizing the logging data of other wells among the multiple wells, excluding the standard well, to between 0 and 200 API based on the mean and variance of the standard well, where API represents the unit of natural gamma data.

[0328] Figure 7 is a histogram of the natural gamma data before normalization provided in the embodiments of this application.

[0329] Figure 8 is a histogram of normalized natural gamma data provided in the embodiments of this application.

[0330] S1033, rock physical environment correction is performed on the logging data of multiple standardized wells to obtain the second logging data of multiple wells.

[0331] Figure 9 is a cross-plot of the volumetric density logging and P-wave velocity curves provided in the embodiments of this application. As shown in Figure 9, the blue curve represents the upper limit of the volumetric density logging and P-wave velocity curves of sandstone in this region, and the red curve represents the lower limit of the volumetric density logging and P-wave velocity curves of sandstone in this region. Normal data points should be distributed between these two lines. By using the elastic parameter limits predicted by the model, abnormal data points that exceed the normal range can be identified.

[0332] Optionally, when the logging data of multiple wells after standardization is DT, S1033 includes: for each well, determining the environmental correction amount DT_faust for the well; and correcting the DT in the enlargement portion of the DT of the standardized well based on DT_faust.

[0333] DT_faust satisfies the formula: Where VP_f is the P-wave velocity curve, K is the formation parameter, Depth is the depth, and RT is the resistivity data. The value of K is between 2000 and 2500.

[0334] Optionally, when the logging data of multiple wells after standardization is DEN, S1033 includes: for each well, determining the environmental correction amount DEN_gardner for the well; and correcting the DEN in the enlarged portion of the DEN of the standardized well based on DEN_gardner.

[0335] DEN_gardner satisfies the formula: Where R1 = -0.0261, R2 = 0.373, R3 = 1.4 to 1.6; or R1 = -0.0115, R2 = 0.261, R3 = 1.5 to 1.6.

[0336] After correction by S1033, outliers in the logging data were eliminated, making the logging data more reasonable, ensuring that the logging data is more in line with the laws of rock physics, and providing real and reliable logging data for inversion.

[0337] Figure 10 shows the logging data before and after rock physical environment correction provided in the embodiments of this application.

[0338] S104. Based on the second logging data of multiple wells, identify multiple candidate sample wells among the multiple wells.

[0339] Optionally, sample wells can provide accurate references to the subsurface rock physical properties of seismic data. The selection criteria for sample wells include: representativeness, well type and spatial distribution, and the completeness of the logging data.

[0340] Among them, representativeness means that the sample well can cover the main geological features and rock types in the preset work area, ensuring that the logging data obtained through the sample well can represent the geological changes of the entire preset work area.

[0341] Different types of wells reveal different reservoir and fluid characteristics. For example, exploratory wells may reveal unknown stratigraphic sequences, while production wells provide data on reservoir quality and production performance. Selecting sample wells based on their type and spatial distribution, ensuring that the spatial distribution of sample wells adequately captures the complexity of geological structures, can improve the accuracy of seismic data interpretation.

[0342] The integrity of well logging data is achieved by providing a complete well logging dataset, including sonic transit time, density, gamma rays, resistivity, etc., to ensure the establishment and calibration of a more accurate rock physics model.

[0343] S105. Based on the second seismic data of multiple seismic traces to be inverted and the second logging data of multiple sample wells to be selected, establish a high-precision sequence stratigraphic framework for the preset work area.

[0344] For a detailed description of S105, please refer to the embodiment in Figure 11, which will not be repeated here.

[0345] S106. For each seismic trace to be inverted, determine the characteristic distance between the second seismic data of the seismic trace to be inverted and the second well bypass seismic data of each candidate sample well.

[0346] The characteristic distance between the second seismic data of the seismic trace to be inverted and the second well bypass seismic data of each candidate sample well can be, for example, Euclidean distance.

[0347] A larger feature distance indicates a smaller similarity (greater difference) between the second seismic data and the second well bypass seismic data, while a smaller feature distance indicates a greater similarity (smaller difference) between the second seismic data and the second well bypass seismic data.

[0348] For a detailed description of S106, please refer to the embodiment shown in Figure 22, which will not be repeated here.

[0349] S107. Determine the first number of first sample wells with the smallest feature distance among multiple candidate sample wells.

[0350] The first quantity is, for example, 6, 7, 10, etc.

[0351] S108. Based on the high-precision sequence stratigraphic framework, the second seismic data of the seismic trace to be inverted, and the second well bypass seismic data of the first number of first sample wells, determine at least two frequency bands.

[0352] At least two frequency bands, for example, include: low frequency band, seismic effective frequency band, well logging effective frequency band, and high frequency band.

[0353] For a detailed description of S108, please refer to the embodiment shown in Figure 26, which will not be repeated here.

[0354] S109. Based on the second logging data of the first sample well and the seismic data of the second well, determine the initial full-frequency inversion result of the seismic trace to be inverted.

[0355] For a detailed description of S109, please refer to the embodiment in Figure 29, which will not be repeated here.

[0356] S110. Based on the high-precision sequence stratigraphic framework and at least two frequency bands, the initial full-frequency inversion results are corrected to obtain the full-frequency target inversion results of the seismic trace to be inverted.

[0357] For a detailed description of S110, please refer to the embodiment shown in Figure 35, which will not be repeated here.

[0358] Figure 11 is a flowchart of a method for establishing a high-precision sequence stratigraphic framework provided in an embodiment of this application. As shown in Figure 11, the method includes:

[0359] S1101. Based on the second seismic data of multiple seismic traces to be inverted and the second logging data and second well bypass seismic data of multiple candidate sample wells, well-seismic calibration is performed to obtain well-seismic calibration results.

[0360] Specifically, S1101 includes S11011 to S11015.

[0361] S11011. For each candidate sample well, select the second well bypass seismic data of that candidate sample well from the second well bypass seismic data of multiple candidate sample wells.

[0362] S11012. Extract acoustic transit time data and density data from the second logging data of the candidate sample wells to determine impedance data. Use the impedance data to convolve with the seismic wavelet to obtain a synthetic seismic record.

[0363] S11013. Identify the in-phase axis with high signal intensity in the selected well-side seismic data, align it with the in-phase axis with the highest signal intensity in the synthetic seismic record, determine the correspondence between the depth of the candidate sample well and the time of the well-side seismic data, and obtain the time-depth conversion relationship.

[0364] S11014. Select strong and relatively continuous reflection events near the well as calibration points, fine-tune the time-depth conversion relationship, and obtain the fine-tuned time-depth conversion relationship;

[0365] S11015. Based on the fine-tuned time-depth conversion relationship, complete the well-seismic calibration work and obtain the well-seismic calibration results.

[0366] Well-seismic calibration can transform seismic data into a more accurate geological timeframe, thus providing a basis for more precise stratigraphic stratification and geological model building.

[0367] Figure 12 is one of the schematic diagrams of well vibration calibration effect provided in the embodiments of this application.

[0368] S1102. Perform intelligent fault interpretation on the second seismic data of multiple seismic traces to be inverted to obtain the fault cross sections.

[0369] The core of intelligent fault interpretation technology is to utilize automated and intelligent algorithms to transform the traditional manual fault interpretation workflow into a fast, efficient, and accurate interpretation process. This technology, by integrating planar and cross-sectional views of seismic data with relevant geological knowledge, can better address the problems of high difficulty and low efficiency in manually interpreting fault closure in traditional fault interpretation. Figure 13 is a flowchart of intelligent fault interpretation provided in an embodiment of this application.

[0370] Optionally, S1102 includes: S11021 to S11024.

[0371] S11021. The second seismic data of multiple seismic traces to be inverted are processed by the edge detection algorithm to obtain the strike and dip of the fault; the cross section of the fault is determined based on the dip angle and strike.

[0372] S11022. The fault features are extracted from the fault surface using a texture analysis algorithm to construct an initial fault surface model.

[0373] S11023. Perform automatic intelligent interpretation of the initial cross-section model, and finely adjust and control the cross-section model during the intelligent interpretation process.

[0374] S11024. The process of continuously iterating and finely adjusting and quality-controlling the cross-section model outputs the final closed cross-section composed of fault bars, thus completing the intelligent interpretation of the fault.

[0375] Near a fault, due to the contrast change caused by the displacement of the strata, the fault characteristics will show significant changes in gradient values.

[0376] Figure 14 is a cross-sectional fault interpretation diagram provided in an embodiment of this application. Figure 15 is a three-dimensional spatial display diagram of the fault provided in an embodiment of this application.

[0377] S1103. Based on the second seismic data and well-seismic calibration results of multiple seismic traces to be inverted, perform intelligent interpretation of the stratigraphic horizons to obtain the seismic horizons.

[0378] The layered intelligent interpretation employs a machine learning and gridded tracking interpretation model. Through "micro-patch tracking," it shifts the interpretation perspective from traditional two-dimensional to three-dimensional. Computer tracking improves interpretation efficiency and reduces interpretation intensity. Furthermore, through real-time manual monitoring and interactive intelligent methods such as adjusting the closure of main survey lines and connecting lines, it solves the controllability problem. Therefore, it is applicable to the rapid interpretation of various structurally complex areas, significantly improving interpretation efficiency and reducing labor intensity.

[0379] Figure 16 is a schematic diagram of the layer intelligent interpretation process provided in the embodiment of this application.

[0380] Optionally, S1103 includes:

[0381] Based on the well seismic calibration results and the second seismic data of multiple seismic traces to be inverted, the phase axis for determining the seismic horizon was determined;

[0382] By setting up interpretation grids, interpretation sample lines, automatic layer tracking, monitoring and adjustment, and iterative looping, the seismic horizon is obtained by tracking the phase axis.

[0383] Setting up the interpretation grid involves several steps. Before using the FastGrid bedding system, you first need to set an appropriate grid size (e.g., 4x4). This will determine the tracing resolution and computational load during the tracing process. The grid size directly affects the level of detail in the traced bedding layers. Smaller grid sizes can provide higher resolution but also increase computational complexity and time. The quality of the seismic data is also a factor to consider when setting up the interpretation grid. If the seismic data has strong phase axis continuity and small dip variations, a smaller grid size can be used to capture more geological details. If the seismic data has poor phase axis continuity and large dip variations, a larger grid size may be needed to ensure the reliability of the results.

[0384] Explain the sample line, including:

[0385] Seismic lines with strong typicality and good geological structure representation are selected as sample lines to provide more effective training data. Interpreters first select the first line on the interpretation grid as the sample line for manual interpretation. This step serves as a training process, providing a reference model for machine learning algorithms. Through manual interpretation, seismic interpreters can define criteria for horizon tracking and mark geological boundaries, which will guide the subsequent automated tracking process.

[0386] Automatic layer tracking, including:

[0387] Automatic tracking is a core feature of the FastGrid stratigraphic tracking system. It allows the system to automatically identify and track similar stratigraphic features in seismic data after the user has interpreted sample lines. First, stratigraphic features are extracted from the manually interpreted sample lines. These features may include reflection intensity, continuity, frequency characteristics, amplitude characteristics, and phase variations. The extracted features are used to train a machine learning model. During training, the algorithm adjusts its internal parameters to better identify well-defined stratigraphic horizons and geological interfaces. The trained model is then used to analyze uninterpreted seismic lines, searching for patterns that match the stratigraphic features learned from the sample lines. The machine learning model predicts unknown horizons that are similar to known stratigraphic features and extends these interpretations to the next interpretation grid.

[0388] Monitoring and adjustments, including:

[0389] Machine learning models use automatic tracking to quickly identify strata and interpret geological interfaces on large amounts of seismic data. After rapid tracking, the system provides a preliminary interpretation result, but this result may not be the final version and requires manual monitoring and verification. During manual monitoring, interpreters will look for potential problems, such as interruptions in stratigraphic tracking or incorrect stratigraphic identification. If errors are found in the automatic tracking results, interpreters will make necessary manual adjustments. This may involve redefining stratigraphic boundaries, correcting the tracking path, or relabeling stratigraphic boundaries.

[0390] Loop iteration, including:

[0391] Interpreters manually adjust the automated tracking results after monitoring and analyzing them. These adjustments incorporate expert insights and a deep understanding of geological features. The adjusted tracking results are essentially manually annotated data, labeled with the correct stratigraphic and structural information. This adjusted data is used to update and train the machine learning model. The model learns from this new annotated data, adjusting its parameters to account for these manually adjusted conditions in future predictions. As more data is manually corrected and fed into the system, the machine learning model can gradually improve its recognition and prediction capabilities. This reinforcement learning process allows the model to become increasingly accurate, and through continuous feedback loops, the system's performance gradually improves, enhancing the accuracy of subsequent automated tracking steps.

[0392] The intelligent interpretation of the stratigraphy of the entire 3D work area is completed through automatic stratigraphic tracking, manual monitoring and adjustment, and continuous iterative processes, ensuring that the interpretation results meet the accuracy and reliability requirements of geological interpretation.

[0393] Figure 17 is a cross-sectional view of the layer interpretation provided in an embodiment of this application. Figure 18 is a three-dimensional stereoscopic view of the layer interpretation provided in an embodiment of this application.

[0394] S1104. Perform intelligent identification of small and medium-level sequence layers on the seismic profile images corresponding to the second seismic data of multiple seismic traces to be inverted, and obtain the small and medium-level sequence layer interface information between seismic layers.

[0395] In seismic sequence stratigraphy, seismic horizons (including first-order and second-order sequence boundary information) can typically be determined through intelligent interpretation of the horizons. These boundary information are often associated with significant geological events, such as widespread unconformities caused by sea-level changes or long-period sedimentary features, and therefore appear as strong reflection events on seismic profiles.

[0396] For small- to medium-level sequence boundary information, such as third- and fourth-order sequence boundaries, their seismic reflection signals may not be strong or continuous, as they may only represent sedimentary changes over a short period. Therefore, identifying them using traditional seismic interpretation methods can be quite difficult. Furthermore, because these boundaries are denser and more subtle, traditional stratigraphic interpretation would require a tremendous amount of work and very high interpretation accuracy.

[0397] Optionally, S1104 includes S11041 to S11043.

[0398] S11041. Based on the seismic profile images corresponding to the second seismic data of multiple seismic traces to be inverted, the seismic phase volume is calculated to obtain the seismic phase volume.

[0399] A phase volume reflects the phase changes of seismic waves propagating in the subsurface medium, revealing the continuity and discontinuities of geological structures. The Hilbert transform is used to process and analyze seismic data. A significant feature of the Hilbert transform is its ability to introduce a 90-degree phase shift into the original signal, thus constructing an analytic signal. The analytic signal consists of the original signal and the signal obtained from its Hilbert transform, presented in complex form, where the real part represents the original signal and the imaginary part represents the signal obtained from the Hilbert transform. For each time point of the analytic signal, its complex phase angle is calculated, which is the instantaneous phase. By repeating the above process at every point in the entire seismic volume, a phase volume can be constructed. This phase volume can reveal the continuity and discontinuities of seismic events, as well as other useful geological features. Figure 19 shows a seismic phase volume profile provided in an embodiment of this application.

[0400] S11042. Determine the dip angle body based on the seismic phase body.

[0401] Among them, determining the dip body based on the seismic phase body includes: determining the derivative (or gradient) of the seismic phase body in three spatial dimensions (usually a spatial dimension composed of X, Y, and Z). Figure 20 is a dip body profile provided in an embodiment of this application.

[0402] S11043. Automatic tracking is performed on the dip body to obtain information on the small and medium-level sequence interfaces between seismic horizons.

[0403] S1105. Based on the fault cross sections, seismic horizons, and information on small and medium-sized sequence boundaries, establish a high-precision sequence stratigraphic framework.

[0404] Optionally, seismic horizons obtained from intelligent interpretation of horizons can be used as top and bottom and inter-layer constraints. Fault sections can be used for cutting, and seismic phase bodies and dip bodies can be established as constraints. More refined lateral interpolation tracking can be performed on the labeled small and medium-level sequence interface information to better understand the stratigraphic continuity and main geological trends in seismic data. These trends are key guides for sequence layering and horizon tracking.

[0405] In this way, by combining the interpolated small- and medium-level sequence boundary information with the seismic tectonic framework model, a high-precision sequence stratigraphic framework is generated. The high-precision sequence stratigraphic framework integrates detailed tectonic information and sequence framework information at different levels, enabling a clear representation of the spatial distribution and historical evolution of sedimentary units.

[0406] Figure 21 is a schematic diagram of the high-precision sequence stratigraphic grid provided in the embodiment of this application.

[0407] Figure 22 is a schematic flowchart of the method for determining feature distance provided in an embodiment of this application. As shown in Figure 22, the method includes:

[0408] S2201. Based on the second seismic data of the seismic trace to be inverted, determine the characteristic parameters corresponding to the seismic trace to be inverted.

[0409] In some embodiments, the characteristic parameters corresponding to the seismic trace to be inverted include at least two of the following: centroid parameter, mean parameter, variance parameter, and variable variance parameter.

[0410] In some embodiments, the centroid parameter satisfies Formula 1:

[0411] Where C represents the centroid parameter, N represents the total number of sampling points in the second seismic data of the seismic trace to be inverted, s(n) represents the sampling point data with index n in the second seismic data of the seismic trace to be inverted, and ∑ represents the summation operation.

[0412] For example, if the second seismic data of the seismic trace to be inverted is represented as S, then S = {s(1), s(2), ..., s(N)}.

[0413] In some embodiments, the average value parameter satisfies Formula 2:

[0414] in, This represents the average value parameter of the second seismic data of the seismic trace to be inverted.

[0415] In some embodiments, the variance parameter satisfies Formula 3:

[0416] Where σ represents the variance parameter.

[0417] In some embodiments, the variable variance parameter satisfies Formula 4:

[0418] Where D represents the variable variance parameter, r(h) represents the mean of the sum of squares of the differences between the sample data s(n) with index n and the sample data s(n+h) with index (n+h) when there are (h-1) sample data points, and N1 represents the number of data points in the second seismic data of the seismic trace to be inverted that are spaced (h-1) sample data points apart.

[0419] S2202. For each candidate sample well, determine the characteristic parameters corresponding to the candidate sample well based on the seismic data of the second well bypass of the candidate sample well.

[0420] Optionally, the specific steps for "determining the characteristic parameters corresponding to each candidate sample well based on the second well bypass seismic data of the sample well" are similar to those in S2201, and will not be repeated here.

[0421] Optionally, the characteristic parameters corresponding to the sample well are the same as the characteristic parameters corresponding to the seismic data from the second well bypass.

[0422] S2203. The distance between the characteristic parameters corresponding to the seismic trace to be inverted and the characteristic parameters corresponding to the sample well to be selected is determined as the characteristic distance between the second seismic data of the seismic trace to be inverted and the second well bypass seismic data of the sample well.

[0423] Optionally, S2203 includes:

[0424] The eigenvectors corresponding to the seismic traces to be inverted are determined by the characteristic parameters corresponding to the seismic traces to be inverted.

[0425] The feature vector corresponding to the sample well is determined by the feature parameters corresponding to the sample well.

[0426] The distance between the eigenvector corresponding to the seismic trace to be inverted and the eigenvector corresponding to the sample well is determined as the characteristic distance between the second seismic data of the seismic trace to be inverted and the second well bypass seismic data of the sample well.

[0427] Taking a seismic trace to be inverted and a sample well to be selected as an example, the eigenvector corresponding to the seismic trace to be inverted satisfies the formula: Where A1 represents the eigenvector corresponding to the seismic trace to be inverted, S1 represents the second seismic data of the seismic trace to be inverted, and C1 represents the centroid parameter corresponding to the seismic trace to be inverted. σ1 represents the average value parameter of the second seismic data of the seismic trace to be inverted, D1 represents the variance parameter of the seismic trace to be inverted, and D1 represents the variable variance parameter of the seismic trace to be inverted.

[0428] The eigenvectors corresponding to the sample wells satisfy the formula: Where A2 represents the eigenvector corresponding to the sample well, S2 represents the second-well bypass seismic data of the sample well, and C2 represents the centroid parameter corresponding to the sample well. σ2 represents the average value parameter of the second well bypass seismic data of the sample well, D2 represents the variance parameter of the sample well, and D2 represents the variable variance parameter of the sample well.

[0429] C1 and C2 are both calculated according to Formula 1. All are calculated according to Formula 2, σ1 and σ2 are calculated according to Formula 3, and D1 and D2 are calculated according to Formula 4.

[0430] The characteristic distance, the eigenvector corresponding to the seismic trace to be inverted, and the eigenvector corresponding to the sample well satisfy the formula: Where L(A1,A2) represents the feature distance.

[0431] In practical 3D seismic data processing, due to real-time variations in formation thickness, the seismic data of the trace to be inverted and the seismic data of the sample wells are not equal in length. The Euclidean distance algorithm requires the curves being compared to be of equal length, therefore it is not suitable for similarity measurement of seismic data of uniform length. In contrast, the dynamic time programming algorithm performs excellently when processing seismic data with time stretching or compression, and it is more accurate than the Euclidean distance algorithm in determining the characteristic distance. Therefore, the dynamic time programming algorithm can also be used to obtain the characteristic distance between the second seismic data of the trace to be inverted and the second well bypass seismic data of the sample wells.

[0432] The dynamic time programming algorithm can be expressed as: Where l represents S int The corresponding eigenvector A(S) int ) and S i The corresponding eigenvector A(S) i The characteristic distance between S) int For the second seismic data of the seismic trace to be inverted, S iLet K1 represent the i-th candidate well among multiple candidate wells, and let dis and S represent the number of candidate wells. int Index of candidate sample wells with the highest similarity (smallest feature distance).

[0433] Optionally, the feature distance can be represented by the correlation coefficient, which ranges from [0, 1]. A correlation coefficient close to 1 means that the feature distance between the feature vectors corresponding to the two seismic data is close, that is, the two seismic data have a high positive correlation. A correlation coefficient close to 0 means that the feature distance between the feature vectors corresponding to the two seismic data is far, that is, the two seismic data have no linear correlation.

[0434] Figure 23 is a schematic diagram of obtaining the correlation coefficient based on Euclidean distance according to an embodiment of this application. As shown in Figure 23, based on Euclidean distance, the correlation coefficient between the second seismic data of the seismic trace to be inverted and the second well bypass seismic data of a candidate sample well is 0.79.

[0435] Figure 24 is a schematic diagram of the correlation coefficient obtained based on the dynamic time programming algorithm provided in the embodiment of this application. As shown in Figure 24, the correlation coefficient between the second seismic data of the seismic trace to be inverted and the second well bypass seismic data of a candidate sample well is determined to be 0.88 through the dynamic time programming algorithm. The feature distance result determined by the dynamic time programming algorithm is closer to 1, that is, it is more accurate.

[0436] Figure 25 is a statistical chart of the correlation coefficients between the channel to be inverted and 39 candidate sample wells provided in the embodiment of this application. As shown in Figure 25, the first 6 correlation coefficients are the highest, and the correlation coefficients of the 6 correlation coefficients are between 0.79 and 0.94. The 6 candidate sample wells corresponding to the first 6 correlation coefficients can be determined as the first number of first sample wells. At this time, the first number is 6.

[0437] Figure 26 is a schematic flowchart of a method for determining at least two frequency bands provided in an embodiment of this application. As shown in Figure 26, the method includes:

[0438] S2601. Based on the high-precision sequence stratigraphic framework and the second seismic data of the seismic traces to be inverted, determine the effective low-frequency and effective high-frequency seismic data.

[0439] In some embodiments, S2601 includes: obtaining seismic data corresponding to the target layer from the second seismic data of the seismic trace to be inverted using a high-precision sequence stratigraphic framework;

[0440] Spectral scanning is performed on the seismic data corresponding to the target layer to obtain amplitude and frequency information;

[0441] The first frequency corresponding to the preset amplitude in the amplitude frequency information is determined as the effective low frequency of the earthquake, and the second frequency corresponding to the preset amplitude is determined as the effective high frequency of the earthquake. The first frequency is less than the second frequency.

[0442] Optionally, when performing spectral scanning on the seismic data corresponding to the target layer, the top and bottom boundary layers of the target layer are determined, and the spectral scanning segment is not less than 2 wavelet lengths, i.e., not less than 200ms.

[0443] Optionally, when the target layer thickness is small, the scanning range can be extended vertically and horizontally towards the top layer to obtain more reliable spectral information.

[0444] Figure 27 is an amplitude-frequency information diagram obtained by spectral scanning of seismic data corresponding to the target layer according to an embodiment of this application. As shown in Figure 27, the horizontal axis represents frequency in Hz. The vertical axis represents the normalized seismic amplitude-frequency information, with a value range of 0-1. The preset amplitude is 0.25, and the first frequency (i.e., the effective low frequency of the seismic wave) corresponding to 0.25 is F1, and the second frequency (i.e., the effective high frequency of the seismic wave) is F2.

[0445] S2602. Select the two target wells with the largest feature distance from the first number of first sample wells.

[0446] S2603. Based on the high-precision sequence stratigraphic framework and the second well bypass seismic data of the two target wells, determine the effective logging frequency.

[0447] In some embodiments, S2603 includes S26031 to S26033.

[0448] S26031. For each of the two target wells, using a high-precision sequence stratigraphic framework, obtain the second well bypass seismic data corresponding to the target layer of the target well from the second well bypass seismic data of the target well; transform the second well bypass seismic data corresponding to the target layer from the time domain to the frequency domain to obtain the transformation result; determine the frequency domain data corresponding to the effective frequency range in the transformation result.

[0449] For example, two target wells include the first target well C1 and the second target well C2.

[0450] Optionally, for C1, the frequency domain data corresponding to the effective frequency range satisfies the following formula: Among them, M C1 (t,w) represents the frequency domain data corresponding to the effective frequency range of C1, where t represents time, w represents frequency, ψ1(jw) represents the transformation result corresponding to C1, and -W to W represents the effective frequency range. Here, j is the imaginary unit.

[0451] Optionally, for C2, the frequency domain data corresponding to the effective frequency range satisfies the following formula: Among them, M C2(t,w) represents the frequency domain data corresponding to the effective frequency range of C2, where t represents time, w represents frequency, and ψ2(jw) represents the transformation result corresponding to C2.

[0452] S26032. Based on the frequency domain data corresponding to the effective frequency range of the two target wells, determine the differential amplitude correlation coefficients corresponding to multiple frequencies in the effective frequency range.

[0453] The effective frequency range of the first target well includes multiple frequencies corresponding to the first amplitude in the frequency domain data. The effective frequency range of the second target well includes multiple frequencies corresponding to the second amplitude in the frequency domain data.

[0454] In some embodiments, S26032 includes: for each frequency, determining a correlation coefficient between a first amplitude and a second amplitude corresponding to the frequency;

[0455] The difference between the correlation coefficient between the first and second amplitudes at the corresponding frequency and the correlation coefficient between the first and second amplitudes at the next frequency is determined as the differential amplitude correlation coefficient at the corresponding frequency.

[0456] The correlation coefficient between the first and second amplitudes satisfies the formula: Where, ρ 12 (w) represents the first amplitude M corresponding to frequency w. C1 (t,w) and the second amplitude M C2 The correlation coefficient between (t, w) is given by COV, where COV represents the covariance and DK represents the variance, and -W ≤ w ≤ W.

[0457] S26033. Among the differential amplitude correlation coefficients corresponding to multiple frequencies, the frequencies that satisfy the first condition are determined as the effective logging frequencies.

[0458] Optionally, the differential amplitude correlation coefficient corresponding to the frequency satisfies the formula: dρ 12 (w)=ρ 12 (w)-ρ 12 (w+1), w>0; where dρ 12 (w) represents the difference magnitude correlation coefficient corresponding to w, ρ 12 (w) represents the correlation coefficient between the first and second amplitudes corresponding to w, ρ 12 (w+1) represents the correlation coefficient between the first and second amplitudes corresponding to w+1.

[0459] In some embodiments, the first condition is expressed as a formula: Where w3 represents the effective logging frequency, argmax represents the maximum value operation, w represents the frequency within the effective frequency range, and dρ 12(w) represents the differential amplitude correlation coefficient corresponding to w, || represents taking the absolute value, and a represents the first preset value.

[0460] Optionally, the first preset value is the threshold for identifying a significant decrease in the correlation of the signal, which can be set to a normal number that is close to 0.

[0461] S2604. Based on the effective low frequency of seismic activity, the effective high frequency of seismic activity, and the effective frequency of well logging, the first preset frequency band is divided into a low frequency band, an effective seismic frequency band, an effective well logging frequency band, and a high frequency band.

[0462] Figure 28 is a schematic diagram of the four frequency bands provided in the embodiments of this application. As shown in Figure 28, band A represents the low-frequency band, with a frequency range of 0-F1; band B represents the effective seismic frequency band, with a frequency range of F1-F2; band C represents the effective logging frequency band, with a frequency range of F2-F3; and band D represents the high-frequency band, with a frequency range of F3 to the high-frequency end. Among them, F1 is the effective low-frequency seismic band, F2 is the effective high-frequency seismic band, and F3 is the effective logging frequency.

[0463] Figure 29 is a schematic flowchart of the method for determining the initial inversion result of the full frequency according to an embodiment of this application. As shown in Figure 29, the method includes:

[0464] S2901. Based on the second logging data and the second well bypass seismic data of multiple candidate sample wells, determine the reservoir model structure variation function corresponding to the multiple candidate sample wells.

[0465] In some embodiments, the reservoir model structure variation function corresponding to the Xth candidate sample well among a plurality of candidate sample wells satisfies the formula: W1*F X (dS)=a X ×(dS) 2 +b X ×(dS)+b X Where W1 represents the seismic wavelet, F X The structure variation function of the reservoir model corresponding to the Xth candidate sample well is represented by *, where * indicates convolution operation, and a X ,b X ,c X The coefficients in the structure variation function of the reservoir model corresponding to the Xth candidate sample well are used. dS represents the difference data between the second well bypass seismic data of the Xth candidate sample well and the second well bypass seismic data of the Yth candidate sample well among multiple candidate sample wells. X and Y are different, and × represents multiplication operation.

[0466] The following explanation uses the Xth candidate sample as an example to illustrate the reservoir model structure variation function corresponding to the Xth candidate sample well:

[0467] When the number of candidate sample wells is K3, the second logging data and second well bypass seismic data of the Xth candidate sample well, as well as the second logging data and second well bypass seismic data of the Yth candidate sample well (excluding the Xth candidate sample well), are input into the standard linear equation a×(S X -S Y ) 2 +b×(S X -S Y )+c=W1*(M X -M Y ), to obtain (K3-1) standard linear equations, where Y≠X;

[0468] Solve the system of standard linear equations consisting of (K3-1) standard linear equations to determine the value of a. X The value of b is b X The value of c is c X To obtain the reservoir model structure variation function corresponding to the Xth candidate sample well.

[0469] It should be noted that a1, b1, and c1 in the reservoir model structure variation function may differ for different candidate sample wells.

[0470] When the number of candidate sample wells is K3, taking X=5 as an example, the system of standard linear equations consisting of (K3-1) standard linear equations is as follows:

[0471] Solving the standard linear equations, we can determine that the value of 'a' is a5, the value of 'b' is b5, and the value of 'c' is c5. Furthermore, substituting a5, b5, and c5 into the standard linear equations, we obtain the reservoir model structure variation function corresponding to the 5th candidate sample well, which satisfies the formula: W1*F5(dS)=a5×(dS)2+b5×(dS)+c5.

[0472] Figure 30 is a schematic diagram of the seismic waveform structure and reservoir model structure of well 1 and well 2 provided in the embodiments of this application.

[0473] S2902. Based on the reservoir model structure variation function corresponding to the first number of first sample wells among multiple candidate sample wells, determine the full-frequency initial inversion result of the seismic trace to be inverted.

[0474] In some embodiments, for each first sample well, the structure variation function of the reservoir model corresponding to the first sample well is solved by Markov chain Monte Carlo stochastic simulation to obtain multiple feasible solutions corresponding to the first sample well; the probability of each possible solution is determined by Bayesian analysis.

[0475] The Bayesian model average is performed on the second-highest number of feasible solutions with the highest probability to obtain the full-frequency initial inversion results of the seismic trace to be inverted.

[0476] Optionally, the multiple feasible solutions corresponding to the first sample well can be expressed as S int The sum of the differences between the logging data of the nearest sample well and the model simulation of the reservoir model's structure variation function, the i-th possible solution satisfies the formula: Among them, M i Let M represent the i-th possible solution. dis Indicates with S int Logging data from the nearest sample well. This represents the model difference compared to the reservoir model's structure variation function simulation, where dS = W1*M dis -S int。

[0477] Optionally, the probability P of the i-th possible solution i Satisfying the formula:

[0478] Where P() represents the law of total probability, P(|) represents the law of conditional probability, and F g1 F represents the reservoir model structure variation function corresponding to the first sample well g1. g2 M represents the reservoir model structure variation function corresponding to the first sample well g2. g1 This represents a possible solution to the well logging data. M represents the logging data corresponding to the first sample well g1. g2 This represents a possible solution to the well logging data. This represents the logging data corresponding to the first sample well g2, and Kc represents the first quantity.

[0479] The Bayesian model average of the top I feasible solutions with the highest probabilities is obtained, and the top I feasible solutions with the highest probabilities satisfy the formula: Where m represents the initial full-frequency inversion result of the seismic trace to be inverted, and E(M) represents the expectation of the first I feasible solutions.

[0480] Optionally, the initial inversion result of the full frequency of the seismic trace to be inverted is the well logging data of the seismic trace to be inverted.

[0481] Figure 31 is a schematic diagram of the final well logging data of the seismic trace to be inverted obtained by averaging the probabilities of the six feasible solutions using a Bayesian model, as provided in an embodiment of this application. As shown in Figure 31, the final well logging data of the seismic trace to be inverted is obtained by averaging the probabilities of the six feasible solutions using a Bayesian model.

[0482] Figure 32 shows the well logging data of the seismic trace to be inverted provided in an embodiment of this application. Figure 33 shows the seismic waveform of the seismic trace to be inverted provided in an embodiment of this application.

[0483] Comparing Figures 32 and 33, it can be seen that the resolution of the well logging data of the seismic trace to be inverted is much higher than that of the seismic data, that is, a high vertical resolution full-frequency initial inversion model is obtained.

[0484] Furthermore, S2901-S2902 is executed for each seismic trace to be inverted in the preset work area to obtain the logging data of each seismic trace to be inverted, and to establish the full-frequency initial inversion model of the preset work area.

[0485] Figure 34 is a cross-sectional view of the full-frequency initial inversion model of the preset work area provided in the embodiment of this application.

[0486] Figure 35 is a schematic flowchart of a method for correcting the initial inversion result of the full frequency according to an embodiment of this application. As shown in Figure 35, the method includes:

[0487] S3501. Obtain the first root mean square velocity corresponding to the seismic trace to be inverted.

[0488] Optionally, the raw acquired seismic data can be obtained;

[0489] The original acquired seismic data were processed to obtain the first root mean square velocity corresponding to multiple seismic traces to be inverted and the first seismic data of multiple seismic traces to be inverted.

[0490] From the first root mean square velocities corresponding to multiple seismic traces to be inverted, obtain the first root mean square velocity corresponding to the seismic trace to be inverted required in S3501.

[0491] S3502. Based on the first root mean square velocity corresponding to the seismic trace to be inverted and the second logging data of multiple candidate sample wells, determine the first impedance data corresponding to the seismic trace to be inverted.

[0492] In some embodiments, S3502 includes S35021 to S35026.

[0493] S35021. For each candidate sample well, convert the DT in the second logging data of the sample well into the logging layer velocity corresponding to the sample well.

[0494] S35022. Interpolate the logging velocities corresponding to multiple candidate sample wells to the logging velocities corresponding to the seismic traces to be inverted.

[0495] S35023. Perform full three-dimensional velocity enhancement processing on the first root mean square velocity to obtain the second root mean square velocity corresponding to the seismic trace to be inverted.

[0496] Optionally, the full 3D velocity enhancement processing includes a moving integral flattening algorithm, designed to address the difficulty in flattening seismic trace aggregations due to significant shifts in the same phase axis. The moving integral flattening algorithm optimizes seismic data processing by precisely determining the minute shifts between adjacent seismic traces and progressively integrating and extrapolating.

[0497] Optionally, the moving integral leveling algorithm specifically includes: determining the leveling amount between each pair of adjacent tracks to obtain a first leveling amount; generating a locally leveled stacked track set based on the first leveling amount; determining the leveling amount between adjacent tracks again based on the locally leveled stacked track set to obtain a second leveling amount; integrating the second leveling amount step by step and extrapolating it to the entire stacked track set, thereby achieving leveling of the entire track set.

[0498] S35024. Using the Dix formula, the second root mean square velocity is converted into the first layer velocity corresponding to the seismic trace to be inverted.

[0499] S35025. Based on the logging layer velocity corresponding to the seismic trace to be inverted, perform well-controlled layer velocity background trend correction on the first layer velocity to obtain the second layer velocity corresponding to the seismic trace to be inverted.

[0500] Optionally, the first layer velocity is corrected for the background trend of well-controlled layer velocity based on the well logging layer velocity corresponding to the seismic trace to be inverted using the DIX formula, so as to obtain the second layer velocity corresponding to the seismic trace to be inverted.

[0501] S35026. Convert the second-layer velocity into the first impedance data corresponding to the seismic trace to be inverted.

[0502] Figure 36 shows the resolution of the first root mean square velocity provided in an embodiment of this application. As shown in Figure 36, the lateral resolution of the first root mean square velocity is 160×160 channels (2000m×2000m), and the longitudinal resolution is approximately 200ms. Figure 37 shows the resolution of the second root mean square velocity provided in an embodiment of this application. As shown in Figure 37, the lateral resolution of the second root mean square velocity is 1×1 channel (12.5m×12.5m), and the longitudinal resolution is approximately 4ms. Comparing Figures 36 and 37, it can be seen that the longitudinal and lateral resolutions of the second root mean square velocity are significantly improved.

[0503] Figure 38 is a cross-sectional view of the first root mean square velocity provided in an embodiment of this application. Figure 39 is a cross-sectional view of the second root mean square velocity provided in an embodiment of this application. Comparing Figures 38 and 39, it can be seen that the second root mean square velocity has richer details in the shallow and target layers, that is, the vertical resolution is improved.

[0504] Figure 40 is a plan view of the first root mean square velocity provided in an embodiment of this application. Figure 41 is a plan view of the second root mean square velocity provided in an embodiment of this application. Comparing Figures 40 and 41, it can be seen that after velocity enhancement processing, the velocity variation details in the Ma17-Ma157 well are richer, and the boundary between low and high velocity changes in the Ma136-Ma3 well also has subtle changes, with the velocity near the Ma3 well becoming higher.

[0505] Figure 42 is a comparison diagram of the well logging velocity corresponding to the seismic trace to be inverted and the first layer velocity provided in the embodiment of this application. As shown in Figure 42, the first layer velocity differs significantly from the well logging velocity. The resolution of the first layer velocity is 2-3 Hz, which is low and cannot reflect the abnormal velocity changes caused by changes in lithology or fluid.

[0506] Figure 43 is a comparison chart of the well logging velocity and the second layer velocity corresponding to the seismic trace to be inverted provided in the embodiment of this application. As shown in Figure 43, the second layer velocity shows the same trend as the well logging velocity, and the resolution of the second layer velocity is 6-8 Hz, which is a significant improvement in resolution.

[0507] Figure 44 is a comparison diagram of the well logging velocity and the first layer velocity profile provided in the embodiment of this application. Figure 45 is a comparison diagram of the well logging velocity and the second layer velocity profile provided in the embodiment of this application. Comparing Figures 44 and 45, it can be seen that the first layer velocity and the well logging velocity profile have significant differences, while the second layer velocity and the well logging velocity profile are more consistent. The second layer velocity not only compensates for the missing low-frequency information in the seismic data, but also clearly shows the velocity structure of different stratigraphic systems on the seismic profile. Vertically, the hierarchical relationship of the second layer velocity is clearer, and horizontally, the variation characteristics of the second layer velocity are more accurate.

[0508] S3503. Using a high-precision sequence stratigraphic framework, determine the data corresponding to the low-frequency band in the initial full-frequency inversion results. Then, using the data corresponding to the low-frequency band in the first impedance data, correct the data corresponding to the low-frequency band in the initial full-frequency inversion results to obtain the first corrected full-frequency inversion data.

[0509] Figure 46 is a schematic diagram of the low-frequency band data in the first impedance data provided in the embodiment of this application. Figure 47 is a schematic diagram of the first corrected full-frequency inversion data provided in the embodiment of this application.

[0510] S3504. Based on the second seismic data of the seismic trace to be inverted, and the second logging data and second well bypass seismic data of the first sample wells of the first number of samples, determine the second impedance data corresponding to the seismic trace to be inverted.

[0511] In some embodiments, S3504 includes S35041 to S35045.

[0512] S35041. Based on the second seismic data of the seismic trace to be inverted, determine the maximum amplitude spectrum corresponding to the seismic trace to be inverted.

[0513] In some embodiments, multiple amplitude spectra corresponding to the seismic trace to be inverted are obtained based on the second seismic data of the seismic trace to be inverted;

[0514] Extract multiple maximum amplitudes and their corresponding frequencies from multiple amplitude spectra;

[0515] According to the order of frequency from smallest to largest, the maximum amplitude and the frequency combination corresponding to the maximum amplitude are determined as the maximum amplitude spectrum corresponding to the seismic trace to be inverted.

[0516] Optionally, the maximum amplitude spectrum corresponding to the seismic trace to be inverted satisfies the formula: Where Q represents the maximum amplitude spectrum corresponding to the seismic trace to be inverted, GST[n, w] represents the amplitude spectrum corresponding to the seismic trace to be inverted, w represents the frequency, and n represents the index of the sampling point data.

[0517] S35042. Determine the first characteristic parameter corresponding to the seismic trace to be inverted based on the maximum amplitude spectrum.

[0518] The first characteristic parameters include the centroid parameter, the mean parameter, the variance parameter, and the variable variance parameter.

[0519] S35043. For each first sample well, based on the second well bypass seismic data of the first sample well, determine the maximum amplitude spectrum corresponding to the first sample well, and based on the maximum amplitude spectrum corresponding to the first sample well, determine the second characteristic parameter corresponding to the first sample well.

[0520] The second characteristic parameters include centroid parameter, mean parameter, variance parameter, and variable variance parameter.

[0521] The calculation methods for the first and second characteristic parameters can be found in S2201, and will not be repeated here.

[0522] S35044. Based on the first characteristic parameter corresponding to the seismic trace to be inverted and the second characteristic parameter corresponding to the first number of first sample wells, perform efficient simulated annealing cluster analysis to obtain the second characteristic parameter with the largest characteristic distance from the first characteristic parameter.

[0523] S35045. Use the second logging data of the first sample well corresponding to the second characteristic parameter with the largest characteristic distance from the first characteristic parameter to determine the second impedance data corresponding to the seismic trace to be inverted.

[0524] Optionally, the specific process of efficient simulated annealing clustering analysis can be found in relevant techniques, which will not be elaborated here.

[0525] S3505. Using a high-precision sequence stratigraphic framework, determine the data corresponding to the effective seismic frequency band in the first corrected full-frequency inversion data. Then, using the data corresponding to the effective seismic frequency band in the second impedance data, correct the data corresponding to the effective seismic frequency band in the first corrected full-frequency inversion data to obtain the second corrected full-frequency inversion data.

[0526] Figure 48 is a schematic diagram of the data corresponding to the effective seismic frequency band in the second impedance data provided in the embodiment of this application. Figure 49 is a schematic diagram of the second corrected full-frequency inversion data provided in the embodiment of this application.

[0527] S3506. Based on the second seismic data of multiple seismic traces to be inverted, determine the third impedance data corresponding to the seismic traces to be inverted.

[0528] In some embodiments, S3506 includes S35061 to S35066.

[0529] S35061. Determine the seismic data difference space based on the second seismic data of multiple seismic traces to be inverted.

[0530] Alternatively, the seismic data difference space can be represented as: Ω={dS z , z = 1, 2, 3, ..., Z; where Ω represents the seismic data difference space, Z represents the total number of seismic traces to be inverted, and dS z This represents the difference of the z-th seismic trace to be inverted.

[0531] S35062. Obtain the impedance difference space corresponding to the seismic data difference space.

[0532] In some embodiments, the value range interval is determined based on the maximum and minimum values ​​in the seismic data difference space;

[0533] Divide the value range into multiple sub-intervals at equal intervals;

[0534] The probability distribution of data belonging to each sub-interval in the difference space of the statistical earthquake data is obtained to determine the probability of each sub-interval.

[0535] The initial state dS0 is obtained based on the probability of each sub-interval;

[0536] Based on the probability of each sub-interval, determine the probability transition matrix for the migration of sample point data at index n to sample point data at index n+1;

[0537] Perform the generation operation of the j-th element in the impedance difference space: using the Markov chain Monte Carlo algorithm, generate the (j-1)-th element dY. j-1 By processing the probability transition matrix, the j-th element dY in the impedance difference space is obtained. jAdd j+1 until the impedance difference space {dY} is obtained. j , j = 1, 2, 3, ..., J}; Initially, j = 1, the (j-1)th element dY j-1 It is dS0.

[0538] For example, the maximum value in the difference space of seismic data is represented as A. max The minimum value is represented by A. min The range is represented as [A] max A min Optionally, [A] max A min The interval is divided into K2 subintervals, and the k2th subinterval represents: A k2 k2 = 1, 2, 3, ..., K2, where K2 can be an integer less than 256.

[0539] The probability transition matrix for migrating from sample point data at index n to sample point data at index n+1 can be represented as: For example, P(A1|A2) represents dS z (n)∈A2 migrates to dS z The probability of (n+1)∈A1 is the probability of the sample point data with index n migrating to the sample point data n+1 with index n+1.

[0540] S35063. Obtain the reservoir model structure variation function corresponding to each first sample well.

[0541] S35064. Construct the objective function based on the reservoir model structure variation function corresponding to each first sample well.

[0542] The objective function can be expressed as: Where v takes the value of an integer between 1 and the first quantity, L v Let F represent the v-th coefficient of the objective function. v Let dY represent the reservoir model structure variation function corresponding to the vth first sample well. j This represents the j-th element in the impedance difference.

[0543] S35065. Based on the elements in the impedance difference space, input the objective function and optimize the objective function to obtain the coefficients of the objective function and the minimum element in the impedance difference space.

[0544] Optionally, {dY j Substitute the elements of {j = 1, 2, 3, ..., J} into the objective function in sequence, and optimize the objective function to obtain the coefficients of the objective function as L. v The smallest element in the impedance difference space is dY x .

[0545] S35066. Based on the coefficients and minimum element of the objective function, determine the third impedance data corresponding to the seismic trace to be inverted.

[0546] The third impedance data satisfies the formula: Where m3 represents the third impedance data, M v F represents the second logging data of the vth first sample well. v Let dY represent the reservoir model structure variation function corresponding to the vth first sample well. x This represents the smallest element.

[0547] S3507. Using a high-precision sequence stratigraphic framework, determine the data corresponding to the effective logging frequency band in the second corrected full-frequency inversion data. Then, using the data corresponding to the effective logging frequency band in the third impedance data, correct the data corresponding to the effective logging frequency band in the second corrected full-frequency inversion data to obtain the third corrected full-frequency inversion data.

[0548] Figure 50 is a schematic diagram of the data corresponding to the effective frequency band of well logging in the third impedance data provided in the embodiment of this application. Figure 51 is a schematic diagram of the third corrected full-frequency inversion data provided in the embodiment of this application.

[0549] S3508. Based on the second seismic data of the seismic trace to be inverted, and the second logging data and second well bypass seismic data of the first sample wells of the first number of samples, determine the fourth impedance data corresponding to the seismic trace to be inverted.

[0550] In some embodiments, S3508 includes the following S35081 to S35087.

[0551] S35081. For each first sample well, determine the fifth impedance data based on the second logging data of the first sample well.

[0552] The wave impedance data Z is established using the DEN and sonic velocity data V from the second logging data, where Z = V × DEN, and V = 100000 / DT. It should be noted that the wave impedance in Figure 12 is Z at this location.

[0553] S35082. Perform statistical processing on the fifth impedance data of the first number of first sample wells to obtain the cumulative probability distribution function (CDF).

[0554] In some embodiments, S35082 includes:

[0555] Sort the impedance values ​​in the fifth impedance data from smallest to largest;

[0556] Starting with the first impedance value, for each impedance value, determine the number of impedance values ​​that are less than or equal to that value;

[0557] Divide the number corresponding to each impedance value by the total number of impedance values ​​to obtain the percentage of the cumulative distribution corresponding to each impedance value;

[0558] Each impedance value corresponds one-to-one with the percentage of its cumulative distribution, forming a CDF.

[0559] For example, the impedance values ​​in the fifth impedance data are: 2, 3, 4, 4, 6, 8;

[0560] For an impedance value of 2, there is one data point that is less than or equal to 2, therefore the CDF value at x = 2 is 1 / 6;

[0561] For an impedance value of 3, there are 2 data points that are less than or equal to 3, therefore the CDF value at x = 3 is 2 / 6;

[0562] For an impedance value of 4, there are 4 data points less than or equal to 4, therefore the CDF value at x = 4 is 4 / 6;

[0563] For an impedance value of 6, there are 5 data points less than or equal to 6, therefore the CDF value at x = 6 is 5 / 6;

[0564] For an impedance value of 8, there are 6 data points less than or equal to 8, therefore the CDF value at x = 8 is 6 / 6;

[0565] By connecting these points, a CDF diagram can be drawn, which shows the cumulative distribution of each impedance value.

[0566] CDF can be used to analyze the distribution characteristics of data. In the CDF curve, a gentler slope indicates that the data changes relatively smoothly, that is, the impedance values ​​in this range are relatively concentrated; a steeper slope indicates that the data changes drastically, that is, the impedance values ​​in this range are relatively dispersed.

[0567] CDF can also be used to guide the probability distribution assumptions of impedance models, serving as input for subsequent statistical simulation algorithms.

[0568] S35083. For each first sample well, perform a standard normal transformation on the fifth impedance data of the first sample well according to CDF to obtain the first normal distribution impedance data.

[0569] In some embodiments, the mean and variance are determined according to the CDF; and the fifth impedance data of the first sample well are subjected to a standard normal transformation based on the mean and variance to obtain the first normally distributed impedance data.

[0570] Optionally, the standard normal transform satisfies the formula: Where z1 represents the first normally distributed impedance data, U represents the fifth impedance data of the first sample well, μ represents the mean, and σ1 represents the standard deviation. (U-μ) makes the mean of the first normally distributed impedance data zero. This ensures that the first normally distributed impedance data has a unit standard deviation.

[0571] S35084. Using a collaborative filtering algorithm, the second well bypass seismic data of the first number of first sample wells and the second seismic data of the seismic trace to be inverted are processed to determine multiple second sample wells from the first number of first sample wells.

[0572] In the embodiments of this application, in addition to collaborative filtering algorithms, cosine similarity, Pearson correlation coefficient, Euclidean distance, etc. can also be used to determine multiple second sample wells from the first number of first sample wells based on the second well bypass seismic data of the first number of first sample wells and the second seismic data of the seismic trace to be inverted.

[0573] Optionally, the random sequence generated by the random residual function satisfies the following two conditions:

[0574] The mean of the random sequence is 0;

[0575] The variance of a random sequence is equal to the Kriging variance.

[0576] Optionally, the random residual function can be tested to verify whether the generated random sequence reasonably reflects the uncertainty of the actual geological conditions and ensures that it meets the requirements of the geological model.

[0577] S35085. For each second sample well, perform Kriging interpolation on the first normal distribution impedance data corresponding to the second sample well to obtain the second normal distribution impedance data; generate a random sequence through a random residual function; insert the random sequence into the second normal distribution impedance data to obtain random simulated impedance data.

[0578] S35086. Based on the random simulated impedance data of multiple second sample wells, determine the random simulated impedance data of the seismic trace to be inverted.

[0579] By using a sequential Gaussian stochastic process, the stochastic simulated impedance data of multiple second sample wells are processed to obtain the stochastic simulated impedance data of the seismic trace to be inverted.

[0580] S35087. Using CDF, perform inverse transformation on the stochastic simulated impedance data of the seismic trace to be inverted to obtain the fourth impedance data.

[0581] S3509. Using a high-precision sequence stratigraphic framework, determine the data corresponding to the high-frequency band in the third corrected full-frequency inversion data. Then, using the data corresponding to the high-frequency band in the fourth impedance data, correct the data corresponding to the high-frequency band in the third corrected full-frequency inversion data to obtain the full-frequency target inversion result.

[0582] Figure 52 is a schematic diagram of the high-frequency band data in the fourth impedance data provided in the embodiment of this application. Figure 53 is a schematic diagram of the fourth frequency inversion data provided in the embodiment of this application.

[0583] Optionally, by establishing a full-lifecycle, full-frequency inversion effect quality control index system that covers multiple key links from rock physical analysis and seismic horizon interpretation to inversion accuracy analysis, the entire lifecycle of the full-frequency inversion process can be monitored and quality controlled to ensure the accuracy and reliability of the final obtained underground geological structure and physical property parameters.

[0584] Optionally, the full life cycle full frequency inversion effect quality control index system includes the following six indicators: histogram statistical lithological discrimination degree, synthetic record correlation coefficient, characteristic distance between the seismic trace to be inverted and the selected sample well, seismic interpretation layer and drilling geological stratification error, the coincidence rate between the inverted pseudo-well curve and the original logging curve, and the coincidence rate between the inverted predicted sand body thickness and the logging interpretation thickness.

[0585] Optionally, the seismic inversion process can be improved based on indicators to enhance the accuracy of the full-frequency target inversion results.

[0586] Based on the same technical concept, this application also provides a well-seismic fusion full-frequency inversion device. This well-seismic fusion full-frequency inversion device can achieve the functions of the aforementioned embodiments. The well-seismic fusion full-frequency inversion device provided in this application will be described below with reference to FIG54.

[0587] Figure 54 is a schematic diagram of the well-seismic fusion full-frequency inversion device provided in an embodiment of this application. As shown in Figure 54, the well-seismic fusion full-frequency inversion device 540 includes:

[0588] The acquisition module 5401 is used to acquire the first seismic data of multiple seismic traces to be inverted and the first logging data of multiple wells. The multiple seismic traces to be inverted and the multiple wells are in a preset work area. The first seismic data of the multiple seismic traces to be inverted includes the first well bypass seismic data of the multiple wells.

[0589] Filtering module 5402 is used to filter the first seismic data of the plurality of seismic traces to be inverted to obtain the second seismic data of the plurality of seismic traces to be inverted, wherein the second seismic data of the plurality of seismic traces to be inverted includes the second well bypass seismic data of the plurality of wells;

[0590] The cleaning module 5403 is used to clean the first logging data of the multiple wells to obtain the second logging data of the multiple wells.

[0591] The first determining module 5404 is used to determine multiple candidate sample wells from the multiple wells based on the second logging data of the multiple wells;

[0592] Module 5405 is used to establish a high-precision sequence stratigraphic framework for the preset work area based on the second seismic data of the multiple seismic traces to be inverted and the second logging data of the multiple sample wells to be selected.

[0593] The second determining module 5406 is used to determine the characteristic distance between the second seismic data of the seismic trace to be inverted and the second well bypass seismic data of each candidate sample well for each seismic trace to be inverted;

[0594] The third determining module 5407 is used to determine the first number of first sample wells with the smallest feature distance among the plurality of candidate sample wells;

[0595] The fourth determining module 5408 is used to determine at least two frequency bands based on the high-precision sequence stratigraphic framework, the second seismic data of the seismic trace to be inverted, and the second well bypass seismic data of the first number of first sample wells;

[0596] The fifth determining module 5409 is used to determine the full-frequency initial inversion result of the seismic trace to be inverted based on the second logging data and the second well bypass seismic data of the first sample wells of the first number of sample wells;

[0597] The correction module 5410 is used to correct the initial full-frequency inversion result based on the high-precision sequence stratigraphic framework and the at least two frequency bands to obtain the full-frequency target inversion result of the seismic trace to be inverted.

[0598] It should be noted that the well-seismic fusion full-frequency inversion device 540 provided in this application embodiment can realize all the method steps implemented in the above method embodiment and can achieve the same technical effect. Here, the parts that are the same as those in the method embodiment and the beneficial effects will not be described in detail.

[0599] In some embodiments, the second determining module 5406 is specifically used for:

[0600] Based on the second seismic data of the seismic trace to be inverted, determine the characteristic parameters corresponding to the seismic trace to be inverted;

[0601] For each candidate sample well, the characteristic parameters corresponding to the candidate sample well are determined based on the second well bypass seismic data of the candidate sample well; the distance between the characteristic parameters corresponding to the seismic trace to be inverted and the characteristic parameters corresponding to the candidate sample well is determined as the characteristic distance between the second seismic data of the seismic trace to be inverted and the second well bypass seismic data of the candidate sample well.

[0602] In some embodiments, the characteristic parameters corresponding to the seismic trace to be inverted include at least two of the following: centroid parameter, mean parameter, variance parameter, and variable variance parameter.

[0603] In some embodiments, the centroid parameter satisfies the following formula: Wherein, C represents the centroid parameter, N represents the total number of sampling points in the second seismic data of the seismic trace to be inverted, s(n) represents the sampling point data with index n in the second seismic data of the seismic trace to be inverted, and ∑ represents the summation operation.

[0604] In some embodiments, the average value parameter satisfies the following formula: in, The average value parameter is denoted by s(n), which represents the sampling point data with index n in the second seismic data of the seismic trace to be inverted.

[0605] In some embodiments, the variance parameter satisfies the following formula: Wherein, σ represents the variance parameter.

[0606] In some embodiments, the variance parameter satisfies the following formula: Wherein, D represents the variable variance parameter, r(h) represents the mean of the sum of squares of the differences between the sample data s(n) with index n and the sample data s(n+h) with index (n+h) when there is an interval of (h-1) sample data points, and N1 represents the number of data points in the second seismic data of the seismic trace to be inverted that are spaced apart by (h-1) sample data points.

[0607] In some embodiments, the at least two frequency bands include: a low-frequency band, a seismic effective frequency band, a well logging effective frequency band, and a high-frequency band; the fourth determining module 5408 is specifically used for:

[0608] Based on the high-precision sequence stratigraphic framework and the second seismic data of the seismic trace to be inverted, the effective low-frequency and effective high-frequency seismic data are determined.

[0609] Select the two target wells with the largest feature distance from the first number of first sample wells;

[0610] Based on the high-precision sequence stratigraphic framework and the second well bypass seismic data of the two target wells, the effective logging frequency is determined;

[0611] Based on the effective low frequency of the seismic earthquake, the effective high frequency of the seismic earthquake, and the effective frequency of the well logging, the first preset frequency band is divided into the low frequency band, the effective frequency band of the seismic earthquake, the effective frequency band of the well logging, and the high frequency band.

[0612] In some embodiments, the fourth determining module 5408 is specifically used for:

[0613] Using the high-precision sequence stratigraphic framework, the seismic data corresponding to the target layer is obtained from the second seismic data of the seismic trace to be inverted;

[0614] The seismic data corresponding to the target layer are subjected to spectral scanning to obtain amplitude and frequency information;

[0615] The first frequency corresponding to the preset amplitude in the amplitude frequency information is determined as the effective low frequency of the earthquake, and the second frequency corresponding to the preset amplitude is determined as the effective high frequency of the earthquake, wherein the first frequency is less than the second frequency.

[0616] In some embodiments, the fourth determining module 5408 is specifically used for:

[0617] For each target well, using the high-precision sequence stratigraphic framework, the second-well bypass seismic data corresponding to the target layer of the target well is obtained from the second-well bypass seismic data of the target well; the second-well bypass seismic data corresponding to the target layer is transformed from the time domain to the frequency domain to obtain the transformation result; the frequency domain data corresponding to the effective frequency range is determined from the transformation result.

[0618] Based on the frequency domain data corresponding to the effective frequency range of the two target wells, determine the differential amplitude correlation coefficients corresponding to multiple frequencies in the effective frequency range;

[0619] Based on the differential amplitude correlation coefficients corresponding to the plurality of frequencies, the frequencies that satisfy the first condition among the plurality of frequencies are determined as the effective logging frequencies.

[0620] In some embodiments, the two target wells include a first target well and a second target well, wherein the frequency domain data corresponding to the effective frequency range of the first target well includes a first amplitude corresponding to the plurality of frequencies, and the frequency domain data corresponding to the effective frequency range of the second target well includes a second amplitude corresponding to the plurality of frequencies; the fourth determining module 5408 is specifically used for:

[0621] For each frequency, a correlation coefficient is determined between the first amplitude and the second amplitude corresponding to that frequency; the difference between the correlation coefficient between the first amplitude and the second amplitude corresponding to that frequency and the correlation coefficient between the first amplitude and the second amplitude corresponding to the next frequency is determined as the differential amplitude correlation coefficient corresponding to that frequency.

[0622] In some embodiments, the first condition is expressed as the following formula: Where w3 represents the effective logging frequency, argmax represents the maximum value operation, w represents the frequency within the effective frequency range, and dρ 12 (w) represents the differential amplitude correlation coefficient corresponding to w, || represents taking the absolute value, and a represents the first preset value.

[0623] In some embodiments, the fifth determining module 5409 is specifically used for:

[0624] Based on the second logging data and second well bypass seismic data of multiple candidate sample wells, the reservoir model structure variation function corresponding to the multiple candidate sample wells is determined;

[0625] Based on the reservoir model structure variation function corresponding to the first number of first sample wells among the plurality of candidate sample wells, the full-frequency initial inversion result of the seismic trace to be inverted is determined.

[0626] In some embodiments, the reservoir model structure variation function corresponding to the Xth candidate sample well among the plurality of candidate sample wells satisfies the following formula: W1*F X (dS)=a X ×(dS) 2 +b X ×(dS)+b X Where W1 represents the seismic wavelet, F X The structure variation function of the reservoir model corresponding to the Xth candidate sample well is represented by *, where * denotes convolution operation, and a X ,b X ,c X Based on the coefficients in the structural variation function of the reservoir model corresponding to the Xth candidate sample well, dS represents the difference data between the second well bypass seismic data of the Xth candidate sample well and the second well bypass seismic data of the Yth candidate sample well among the plurality of candidate sample wells. X and Y are different, and × represents multiplication operation.

[0627] In some embodiments, the fifth determining module 5409 is specifically used for:

[0628] For each first sample well, the structure variation function of the reservoir model corresponding to the first sample well is solved by Markov chain Monte Carlo stochastic simulation to obtain multiple feasible solutions for the first sample well; the probability of each possible solution is determined by Bayesian analysis.

[0629] The Bayesian model average is performed on the second-highest number of feasible solutions with the highest probability to obtain the full-frequency initial inversion result of the seismic trace to be inverted.

[0630] In some embodiments, the at least two frequency bands include: a low-frequency band, a seismic effective frequency band, a well logging effective frequency band, and a high-frequency band; the correction module 5410 is specifically used for:

[0631] Obtain the first root mean square velocity corresponding to the seismic trace to be inverted;

[0632] Based on the first root mean square velocity corresponding to the seismic trace to be inverted and the second logging data of the multiple candidate sample wells, the first impedance data corresponding to the seismic trace to be inverted is determined.

[0633] Using the high-precision sequence stratigraphic framework, the data corresponding to the low-frequency band is determined in the full-frequency initial inversion result. Using the data corresponding to the low-frequency band in the first impedance data, the data corresponding to the low-frequency band in the full-frequency initial inversion result is corrected to obtain the first corrected full-frequency inversion data.

[0634] Based on the second seismic data of the seismic trace to be inverted, and the second logging data and second well bypass seismic data of the first sample wells of the first number of samples, the second impedance data corresponding to the seismic trace to be inverted is determined.

[0635] Using the high-precision sequence stratigraphic framework, the data corresponding to the effective seismic frequency band is determined in the first corrected full-frequency inversion data. Using the data corresponding to the effective seismic frequency band in the second impedance data, the data corresponding to the effective seismic frequency band in the first corrected full-frequency inversion data is corrected to obtain the second corrected full-frequency inversion data.

[0636] Based on the second seismic data of the plurality of seismic traces to be inverted, determine the third impedance data corresponding to the seismic traces to be inverted;

[0637] Using the high-precision sequence stratigraphic framework, the data corresponding to the effective logging frequency band is determined in the second corrected full-frequency inversion data. Using the data corresponding to the effective logging frequency band in the third impedance data, the data corresponding to the effective logging frequency band in the second corrected full-frequency inversion data is corrected to obtain the third corrected full-frequency inversion data.

[0638] Based on the second seismic data of the seismic trace to be inverted, and the second logging data and second well bypass seismic data of the first sample wells of the first number of samples, the fourth impedance data corresponding to the seismic trace to be inverted is determined.

[0639] Using the high-precision sequence stratigraphic framework, the data corresponding to the effective frequency band of the well logging is determined in the second corrected full-frequency inversion data. Using the data corresponding to the high-frequency band in the fourth impedance data, the data corresponding to the high-frequency band in the third corrected full-frequency inversion data is corrected to obtain the full-frequency target inversion result.

[0640] In some embodiments, the correction module 5410 is specifically used for:

[0641] For each candidate sample well, the DT in the second logging data of the candidate sample well is converted into the logging layer velocity corresponding to the candidate sample well;

[0642] The logging layer velocities corresponding to the multiple candidate sample wells are interpolated to the logging layer velocities corresponding to the seismic trace to be inverted.

[0643] The first root mean square velocity is subjected to full three-dimensional velocity enhancement processing to obtain the second root mean square velocity corresponding to the seismic trace to be inverted.

[0644] The second root mean square velocity is converted into the first layer velocity corresponding to the seismic trace to be inverted using the Dix formula.

[0645] Based on the logging layer velocity corresponding to the seismic trace to be inverted, the first layer velocity is corrected for the background trend of the well-controlled layer velocity to obtain the second layer velocity corresponding to the seismic trace to be inverted.

[0646] The second layer velocity is converted into the first impedance data corresponding to the seismic trace to be inverted.

[0647] In some embodiments, the correction module 5410 is specifically used for:

[0648] Based on the second seismic data of the seismic trace to be inverted, determine the maximum amplitude spectrum corresponding to the seismic trace to be inverted;

[0649] Based on the maximum amplitude spectrum corresponding to the seismic trace to be inverted, determine the first characteristic parameter corresponding to the seismic trace to be inverted;

[0650] For each first sample well, the maximum amplitude spectrum corresponding to the first sample well is determined based on the second well bypass seismic data of the first sample well, and the second characteristic parameter corresponding to the first sample well is determined based on the maximum amplitude spectrum of the first sample well.

[0651] Based on the first characteristic parameter corresponding to the seismic trace to be inverted and the second characteristic parameter corresponding to the first number of first sample wells, efficient simulated annealing clustering analysis is performed to obtain the second characteristic parameter with the largest characteristic distance from the first characteristic parameter.

[0652] The second logging data of the first sample well corresponding to the second feature parameter with the largest feature distance from the first feature parameter is used to determine the second impedance data corresponding to the seismic trace to be inverted.

[0653] In some embodiments, the correction module 5410 is specifically used for:

[0654] Based on the second seismic data of the seismic trace to be inverted, obtain multiple amplitude spectra corresponding to the seismic trace to be inverted;

[0655] Extract multiple maximum amplitudes and the frequencies corresponding to the multiple maximum amplitudes from the multiple amplitude spectra;

[0656] The maximum amplitude and the frequency corresponding to the maximum amplitude are combined in ascending order of frequency to form the maximum amplitude spectrum corresponding to the seismic trace to be inverted.

[0657] In some embodiments, the correction module 5410 is specifically used for:

[0658] Based on the second seismic data of the multiple seismic traces to be inverted, the seismic data difference space is determined;

[0659] Obtain the impedance difference space corresponding to the seismic data difference space;

[0660] Obtain the reservoir model structure variation function corresponding to each first sample well;

[0661] Based on the reservoir model structure variation function corresponding to each first sample well, an objective function is constructed.

[0662] Based on the elements in the impedance difference, the objective function is input, and the objective function is optimized to obtain the coefficients of the objective function and the minimum element in the impedance difference space.

[0663] Based on the coefficients of the objective function and the minimum element, the third impedance data corresponding to the seismic trace to be inverted is determined.

[0664] In some embodiments, the third impedance data satisfies the following formula: Where m3 represents the third impedance data, v takes the value of an integer between 1 and the first quantity, and L v M represents the v-th coefficient of the objective function. v F represents the second logging data of the vth first sample well.v Let dY represent the reservoir model structure variation function corresponding to the vth first sample well. x This represents the smallest element.

[0665] In some embodiments, the correction module 5410 is specifically used for:

[0666] For each first sample well, the fifth impedance data is determined based on the second logging data of the first sample well;

[0667] Statistical processing is performed on the fifth impedance data of the first number of first sample wells to obtain the cumulative probability distribution function;

[0668] For each first sample well, the fifth impedance data of the first sample well is subjected to a standard normal transformation according to the cumulative probability distribution function to obtain the first normal distribution impedance data;

[0669] The second well bypass seismic data of the first number of first sample wells and the second seismic data of the seismic trace to be inverted are processed by a collaborative filtering algorithm to determine a number of second sample wells in the first number of first sample wells.

[0670] For each second sample well, Kriging interpolation is performed on the first normal distribution impedance data corresponding to the second sample well to obtain the second normal distribution impedance data; a random sequence is generated through a random residual function; the random sequence is inserted into the second normal distribution impedance data to obtain random simulated impedance data;

[0671] Based on the random simulated impedance data of the multiple second sample wells, determine the random simulated impedance data of the seismic trace to be inverted;

[0672] The fourth impedance data is obtained by inverse transformation of the random simulated impedance data of the seismic trace to be inverted using the cumulative probability distribution function.

[0673] In some embodiments, the establishment module 5405 is specifically used for:

[0674] Based on the second seismic data of the multiple seismic traces to be inverted and the second logging data of the multiple sample wells to be selected, well-seismic calibration is performed to obtain well-seismic calibration results;

[0675] The second seismic data of the multiple seismic traces to be inverted are subjected to intelligent fault interpretation to obtain the fault cross-sections.

[0676] Based on the second seismic data of multiple seismic traces to be inverted and the well-seismic calibration results, intelligent interpretation of the seismic horizons is performed to obtain the seismic horizons;

[0677] For the seismic profile images corresponding to the second seismic data of multiple seismic traces to be inverted, perform intelligent identification of small and medium-level sequence layers to obtain the small and medium-level sequence interface information between the seismic layers.

[0678] The high-precision sequence stratigraphic framework is established based on the fault section, the seismic horizon, and the information on the small and medium-level sequence boundaries.

[0679] It should be noted that the well-seismic fusion full-frequency inversion device 540 provided in this application embodiment can realize all the method steps implemented in the above method embodiment and can achieve the same technical effect. Here, the parts that are the same as those in the method embodiment and the beneficial effects will not be described in detail.

[0680] It should be understood that the aforementioned well-seismic fusion full-frequency inversion device 540 is embodied in the form of functional modules. The term "module" here can refer to application-specific integrated circuits (ASICs), electronic circuits, processors (e.g., shared processors, proprietary processors, or group processors) and memories for executing one or more software or firmware programs, combined logic circuits, and / or other suitable components that support the described functions.

[0681] Figure 55 is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. As shown in Figure 55, the electronic device 550 may include a memory 5501 and a processor 5502. Exemplarily, the memory 5501 and the processor 5502 are interconnected via a bus 5503.

[0682] Memory 5501 is used to store program instructions.

[0683] The processor 5502 is used to execute the program instructions stored in the memory to cause the electronic device to perform the above-described method.

[0684] All or part of the steps of each of the above method embodiments can be implemented by hardware associated with program instructions. The aforementioned program can be stored in a readable memory. When the program is executed, it performs the steps of each of the above method embodiments; and the aforementioned memory (storage medium) includes: read-only memory (ROM), RAM, flash memory, hard disk, solid-state drive, magnetic tape, floppy disk, optical disk, and any combination thereof.

[0685] This application provides a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, are used to implement the method described in the above-described method embodiments.

[0686] This application also provides a computer program product, including a computer program that, when executed by a processor, can implement the methods shown in the above-described method embodiments.

[0687] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means that implement the functions specified in one or more flowcharts and / or one or more block diagrams.

[0688] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, such that the instructions, which execute on the computer or other programmable apparatus, provide steps for implementing the functions specified in one or more flowcharts and / or one or more block diagrams.

[0689] Those skilled in the art can make various modifications and variations to the embodiments of this application without departing from the spirit and scope of this application. Therefore, if these modifications and variations to the embodiments of this application fall within the scope of the claims of this application and their equivalents, this application also intends to include these modifications and variations.

[0690] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

Claims

1. A well-seismic fusion full-frequency inversion method, characterized in that, include: Acquire first seismic data of multiple seismic traces to be inverted, and first well logging data of multiple wells, wherein the multiple seismic traces to be inverted and the multiple wells are in a preset work area, and the first seismic data of the multiple seismic traces to be inverted includes the first well bypass seismic data of the multiple wells; The first seismic data of the plurality of seismic traces to be inverted are filtered to obtain the second seismic data of the plurality of seismic traces to be inverted, wherein the second seismic data of the plurality of seismic traces to be inverted includes the second well bypass seismic data of the plurality of wells; The first logging data of the multiple wells is cleaned to obtain the second logging data of the multiple wells; Based on the second logging data of the plurality of wells, a plurality of candidate sample wells are determined from the plurality of wells; Based on the second seismic data of the multiple seismic traces to be inverted and the second logging data of the multiple sample wells to be selected, a high-precision sequence stratigraphic framework for the preset work area is established. For each seismic trace to be inverted, the characteristic distance between the second seismic data of the seismic trace to be inverted and the second well bypass seismic data of each candidate sample well is determined; Among the plurality of candidate sample wells, determine the first number of first sample wells with the smallest feature distance; Based on the high-precision sequence stratigraphic framework, the second seismic data of the seismic trace to be inverted, and the second well bypass seismic data of the first number of first sample wells, at least two frequency bands are determined. Based on the second logging data and the second well bypass seismic data of the first sample wells, the full-frequency initial inversion result of the seismic trace to be inverted is determined; Based on the high-precision sequence stratigraphic framework and the at least two frequency bands, the initial full-frequency inversion results are corrected to obtain the full-frequency target inversion results of the seismic trace to be inverted.

2. The method according to claim 1, characterized in that, Determining the characteristic distance between the second seismic data of the seismic trace to be inverted and the second well bypass seismic data of each candidate sample well includes: Based on the second seismic data of the seismic trace to be inverted, determine the characteristic parameters corresponding to the seismic trace to be inverted; For each candidate sample well, the characteristic parameters corresponding to the candidate sample well are determined based on the second well bypass seismic data of the candidate sample well; the distance between the characteristic parameters corresponding to the seismic trace to be inverted and the characteristic parameters corresponding to the candidate sample well is determined as the characteristic distance between the second seismic data of the seismic trace to be inverted and the second well bypass seismic data of the candidate sample well.

3. The method according to claim 2, characterized in that, The characteristic parameters corresponding to the seismic trace to be inverted include at least two of the following: centroid parameter, mean parameter, variance parameter, and variable variance parameter.

4. The method according to claim 3, characterized in that, The centroid parameter satisfies the following formula 1: Wherein, C represents the centroid parameter, N represents the total number of sampling points in the second seismic data of the seismic trace to be inverted, s(n) represents the sampling point data with index n in the second seismic data of the seismic trace to be inverted, and ∑ represents the summation operation.

5. The method according to claim 4, characterized in that, The average value parameter satisfies the following formula 2: in, This represents the average value parameter.

6. The method according to claim 5, characterized in that, The variance parameter satisfies the following formula 3: Wherein, σ represents the variance parameter.

7. The method according to claim 6, characterized in that, The variable variance parameter satisfies the following formula 4: Wherein, D represents the variable variance parameter, r(h) represents the mean of the sum of squares of the differences between the sample data s(n) with index n and the sample data s(n+h) with index (n+h) when there is an interval of (h-1) sample data points, and N1 represents the number of data points in the second seismic data of the seismic trace to be inverted that are spaced apart by (h-1) sample data points.

8. The method according to any one of claims 1-7, characterized in that, The at least two frequency bands include: a low-frequency band, an effective seismic frequency band, an effective well logging frequency band, and a high-frequency band; The determination of at least two frequency bands based on the high-precision sequence stratigraphic framework, the second seismic data of the seismic trace to be inverted, and the second well bypass seismic data of the first number of first sample wells includes: Based on the high-precision sequence stratigraphic framework and the second seismic data of the seismic trace to be inverted, the effective low-frequency and effective high-frequency seismic data are determined. Select the two target wells with the largest feature distance from the first number of first sample wells; Based on the high-precision sequence stratigraphic framework and the second well bypass seismic data of the two target wells, the effective logging frequency is determined; Based on the effective low frequency of the seismic earthquake, the effective high frequency of the seismic earthquake, and the effective frequency of the well logging, the first preset frequency band is divided into the low frequency band, the effective frequency band of the seismic earthquake, the effective frequency band of the well logging, and the high frequency band.

9. The method according to claim 8, characterized in that, Based on the high-precision sequence stratigraphic framework and the second seismic data of the seismic trace to be inverted, the effective low-frequency and effective high-frequency seismic data are determined, including: Using the high-precision sequence stratigraphic framework, the seismic data corresponding to the target layer is obtained from the second seismic data of the seismic trace to be inverted; The seismic data corresponding to the target layer are subjected to spectral scanning to obtain amplitude and frequency information; The first frequency corresponding to the preset amplitude in the amplitude frequency information is determined as the effective low frequency of the earthquake, and the second frequency corresponding to the preset amplitude is determined as the effective high frequency of the earthquake, wherein the first frequency is less than the second frequency.

10. The method according to claim 8, characterized in that, Based on the high-precision sequence stratigraphic framework and the second-well bypass seismic data from the two target wells, the effective logging frequencies are determined, including: For each target well, using the high-precision sequence stratigraphic framework, the second-well bypass seismic data corresponding to the target layer of the target well is obtained from the second-well bypass seismic data of the target well; the second-well bypass seismic data corresponding to the target layer is transformed from the time domain to the frequency domain to obtain the transformation result; the frequency domain data corresponding to the effective frequency range is determined from the transformation result. Based on the frequency domain data corresponding to the effective frequency range of the two target wells, determine the differential amplitude correlation coefficients corresponding to multiple frequencies in the effective frequency range; Based on the differential amplitude correlation coefficients corresponding to the plurality of frequencies, the frequencies that satisfy the first condition among the plurality of frequencies are determined as the effective logging frequencies.

11. The method according to claim 10, characterized in that, The two target wells include a first target well and a second target well. The frequency domain data corresponding to the effective frequency range of the first target well includes a first amplitude corresponding to the plurality of frequencies, and the frequency domain data corresponding to the effective frequency range of the second target well includes a second amplitude corresponding to the plurality of frequencies. Based on the frequency domain data corresponding to the effective frequency range of the two target wells, determine the differential amplitude correlation coefficients corresponding to multiple frequencies within the effective frequency range, including: For each frequency, determine the correlation coefficient between the first amplitude and the second amplitude corresponding to that frequency; The difference between the correlation coefficient between the first amplitude and the second amplitude corresponding to the frequency and the correlation coefficient between the first amplitude and the second amplitude corresponding to the next frequency is determined as the differential amplitude correlation coefficient corresponding to the frequency.

12. The method according to claim 10, characterized in that, The first condition is expressed as follows: Formula 5: Where w3 represents the effective logging frequency, argmax represents the maximum value operation, w represents the frequency within the effective frequency range, and dρ 12 (w) represents the differential amplitude correlation coefficient corresponding to w, || represents taking the absolute value, and a represents the first preset value.

13. The method according to any one of claims 1-7, characterized in that, The step of determining the initial full-frequency inversion result of the seismic trace to be inverted based on the second logging data and the second well bypass seismic data of the first sample wells of the first number of sample wells includes: Based on the second logging data and second well bypass seismic data of multiple candidate sample wells, the reservoir model structure variation function corresponding to the multiple candidate sample wells is determined; Based on the reservoir model structure variation function corresponding to the first number of first sample wells among the plurality of candidate sample wells, the full-frequency initial inversion result of the seismic trace to be inverted is determined.

14. The method according to claim 13, characterized in that, The reservoir model structure variation function corresponding to the Xth candidate well among the plurality of candidate sample wells satisfies the following formula 6: W1*F X (dS)=a X ×(dS) 2 +b X ×(dS)+b X (Formula 6); Where W1 represents the seismic wavelet, F X The structure variation function of the reservoir model corresponding to the Xth candidate sample well is represented by *, where * denotes convolution operation, and a X ,b X ,c X Based on the coefficients in the structural variation function of the reservoir model corresponding to the Xth candidate sample well, dS represents the difference data between the second well bypass seismic data of the Xth candidate sample well and the second well bypass seismic data of the Yth candidate sample well among the plurality of candidate sample wells. X and Y are different, and × represents multiplication operation.

15. The method according to claim 13, characterized in that, Based on the reservoir model structure variation function corresponding to the first number of first sample wells among the plurality of candidate sample wells, the full-frequency initial inversion result of the seismic trace to be inverted is determined, including: For each first sample well, the structure variation function of the reservoir model corresponding to the first sample well is solved by Markov chain Monte Carlo stochastic simulation to obtain multiple feasible solutions for the first sample well; the probability of each possible solution is determined by Bayesian analysis. The Bayesian model average is performed on the second-highest number of feasible solutions with the highest probability to obtain the full-frequency initial inversion result of the seismic trace to be inverted.

16. The method according to any one of claims 1-7, characterized in that, The at least two frequency bands include: a low-frequency band, an effective seismic frequency band, an effective well logging frequency band, and a high-frequency band; Based on the high-precision sequence stratigraphic framework and the at least two frequency bands, the initial full-frequency inversion results are corrected to obtain the full-frequency target inversion results of the seismic trace to be inverted, including: Obtain the first root mean square velocity corresponding to the seismic trace to be inverted; Based on the first root mean square velocity corresponding to the seismic trace to be inverted and the second logging data of the multiple candidate sample wells, the first impedance data corresponding to the seismic trace to be inverted is determined. Using the high-precision sequence stratigraphic framework, the data corresponding to the low-frequency band is determined in the full-frequency initial inversion result. Using the data corresponding to the low-frequency band in the first impedance data, the data corresponding to the low-frequency band in the full-frequency initial inversion result is corrected to obtain the first corrected full-frequency inversion data. Based on the second seismic data of the seismic trace to be inverted, and the second logging data and second well bypass seismic data of the first sample wells of the first number of samples, the second impedance data corresponding to the seismic trace to be inverted is determined. Using the high-precision sequence stratigraphic framework, the data corresponding to the effective seismic frequency band is determined in the first corrected full-frequency inversion data. Using the data corresponding to the effective seismic frequency band in the second impedance data, the data corresponding to the effective seismic frequency band in the first corrected full-frequency inversion data is corrected to obtain the second corrected full-frequency inversion data. Based on the second seismic data of the plurality of seismic traces to be inverted, determine the third impedance data corresponding to the seismic traces to be inverted; Using the high-precision sequence stratigraphic framework, the data corresponding to the effective logging frequency band is determined in the second corrected full-frequency inversion data. Using the data corresponding to the effective logging frequency band in the third impedance data, the data corresponding to the effective logging frequency band in the second corrected full-frequency inversion data is corrected to obtain the third corrected full-frequency inversion data. Based on the second seismic data of the seismic trace to be inverted, and the second logging data and second well bypass seismic data of the first sample wells of the first number of samples, the fourth impedance data corresponding to the seismic trace to be inverted is determined. Using the high-precision sequence stratigraphic framework, the data corresponding to the effective frequency band of the well logging is determined in the second corrected full-frequency inversion data. Using the data corresponding to the high-frequency band in the fourth impedance data, the data corresponding to the high-frequency band in the third corrected full-frequency inversion data is corrected to obtain the full-frequency target inversion result.

17. The method according to claim 16, characterized in that, Based on the first root mean square velocity corresponding to the seismic trace to be inverted and the second logging data of the multiple candidate sample wells, the first impedance data corresponding to the seismic trace to be inverted is determined, including: For each candidate sample well, the DT in the second logging data of the candidate sample well is converted into the logging layer velocity corresponding to the candidate sample well; The logging layer velocities corresponding to the multiple candidate sample wells are interpolated to the logging layer velocities corresponding to the seismic trace to be inverted. The first root mean square velocity is subjected to full three-dimensional velocity enhancement processing to obtain the second root mean square velocity corresponding to the seismic trace to be inverted. The second root mean square velocity is converted into the first layer velocity corresponding to the seismic trace to be inverted using the Dix formula. Based on the logging layer velocity corresponding to the seismic trace to be inverted, the first layer velocity is corrected for the background trend of the well-controlled layer velocity to obtain the second layer velocity corresponding to the seismic trace to be inverted. The second layer velocity is converted into the first impedance data corresponding to the seismic trace to be inverted.

18. The method according to claim 16, characterized in that, Based on the second seismic data of the seismic trace to be inverted, and the second logging data and second well bypass seismic data of the first number of first sample wells, the second impedance data corresponding to the seismic trace to be inverted is determined, including: Based on the second seismic data of the seismic trace to be inverted, determine the maximum amplitude spectrum corresponding to the seismic trace to be inverted; Based on the maximum amplitude spectrum corresponding to the seismic trace to be inverted, determine the first characteristic parameter corresponding to the seismic trace to be inverted; For each first sample well, the maximum amplitude spectrum corresponding to the first sample well is determined based on the second well bypass seismic data of the first sample well, and the second characteristic parameter corresponding to the first sample well is determined based on the maximum amplitude spectrum of the first sample well. Based on the first characteristic parameter corresponding to the seismic trace to be inverted and the second characteristic parameter corresponding to the first number of first sample wells, efficient simulated annealing clustering analysis is performed to obtain the second characteristic parameter with the largest characteristic distance from the first characteristic parameter. The second logging data of the first sample well corresponding to the second feature parameter with the largest feature distance from the first feature parameter is used to determine the second impedance data corresponding to the seismic trace to be inverted.

19. The method according to claim 18, characterized in that, Based on the second seismic data of the seismic trace to be inverted, determine the maximum amplitude spectrum corresponding to the seismic trace to be inverted, including: Based on the second seismic data of the seismic trace to be inverted, obtain multiple amplitude spectra corresponding to the seismic trace to be inverted; Extract multiple maximum amplitudes and the frequencies corresponding to the multiple maximum amplitudes from the multiple amplitude spectra; The maximum amplitude and the frequency corresponding to the maximum amplitude are combined in ascending order of frequency to form the maximum amplitude spectrum corresponding to the seismic trace to be inverted.

20. The method according to claim 16, characterized in that, Based on the second seismic data of the seismic trace to be inverted, the third impedance data corresponding to the seismic trace to be inverted is determined, including: Based on the second seismic data of the multiple seismic traces to be inverted, the seismic data difference space is determined; Obtain the impedance difference space corresponding to the seismic data difference space; Obtain the reservoir model structure variation function corresponding to each first sample well; Based on the reservoir model structure variation function corresponding to each first sample well, an objective function is constructed. Based on the elements in the impedance difference, the objective function is input, and the objective function is optimized to obtain the coefficients of the objective function and the minimum element in the impedance difference space. Based on the coefficients of the objective function and the minimum element, the third impedance data corresponding to the seismic trace to be inverted is determined.

21. The method according to claim 20, characterized in that, The third impedance data satisfies the following formula 7: Where m3 represents the third impedance data, v takes the value of an integer between 1 and the first quantity, and L v M represents the v-th coefficient of the objective function. v F represents the second logging data of the vth first sample well. v Let dY represent the reservoir model structure variation function corresponding to the vth first sample well. x This represents the smallest element.

22. The method according to claim 16, characterized in that, Based on the second seismic data of the seismic trace to be inverted, and the second logging data and second well bypass seismic data of the first number of first sample wells, the fourth impedance data corresponding to the seismic trace to be inverted is determined, including: For each first sample well, the fifth impedance data is determined based on the second logging data of the first sample well; Statistical processing is performed on the fifth impedance data of the first number of first sample wells to obtain the cumulative probability distribution function; For each first sample well, the fifth impedance data of the first sample well is subjected to a standard normal transformation according to the cumulative probability distribution function to obtain the first normal distribution impedance data; The second well bypass seismic data of the first number of first sample wells and the second seismic data of the seismic trace to be inverted are processed by a collaborative filtering algorithm to determine a number of second sample wells in the first number of first sample wells. For each second sample well, Kriging interpolation is performed on the first normal distribution impedance data corresponding to the second sample well to obtain the second normal distribution impedance data; a random sequence is generated through a random residual function; the random sequence is inserted into the second normal distribution impedance data to obtain random simulated impedance data; Based on the random simulated impedance data of the multiple second sample wells, determine the random simulated impedance data of the seismic trace to be inverted; The fourth impedance data is obtained by inverse transformation of the random simulated impedance data of the seismic trace to be inverted using the cumulative probability distribution function.

23. The method according to any one of claims 1-7, characterized in that, Based on the second seismic data of the multiple seismic traces to be inverted and the second well logging data of the multiple sample wells to be selected, a high-precision sequence stratigraphic framework for the preset work area is established, including: Based on the second seismic data of the multiple seismic traces to be inverted and the second logging data of the multiple sample wells to be selected, well-seismic calibration is performed to obtain well-seismic calibration results; The second seismic data of the multiple seismic traces to be inverted are subjected to intelligent fault interpretation to obtain the fault cross-sections. Based on the second seismic data of multiple seismic traces to be inverted and the well-seismic calibration results, intelligent interpretation of the seismic horizons is performed to obtain the seismic horizons; For the seismic profile images corresponding to the second seismic data of multiple seismic traces to be inverted, perform intelligent identification of small and medium-level sequence layers to obtain the small and medium-level sequence interface information between the seismic layers. The high-precision sequence stratigraphic framework is established based on the fault section, the seismic horizon, and the information on the small and medium-level sequence boundaries.

24. A well-seismic fusion full-frequency inversion device, characterized in that, include: The acquisition module is used to acquire the first seismic data of multiple seismic traces to be inverted and the first logging data of multiple wells. The multiple seismic traces to be inverted and the multiple wells are in a preset work area. The first seismic data of the multiple seismic traces to be inverted includes the first well bypass seismic data of the multiple wells. The filtering module is used to filter the first seismic data of the plurality of seismic traces to be inverted to obtain the second seismic data of the plurality of seismic traces to be inverted, wherein the second seismic data of the plurality of seismic traces to be inverted includes the second well bypass seismic data of the plurality of wells; A cleaning module is used to clean the first logging data of the multiple wells to obtain the second logging data of the multiple wells. The first determining module is used to determine multiple candidate sample wells from the multiple wells based on the second logging data of the multiple wells; A module is established to build a high-precision sequence stratigraphic framework for the preset work area based on the second seismic data of the multiple seismic traces to be inverted and the second well logging data of the multiple sample wells to be selected. The second determining module is used to determine the characteristic distance between the second seismic data of the seismic trace to be inverted and the second well bypass seismic data of each candidate sample well for each seismic trace to be inverted; The third determining module is used to determine the first number of first sample wells with the smallest feature distance among the plurality of candidate sample wells; The fourth determining module is used to determine at least two frequency bands based on the high-precision sequence stratigraphic framework, the second seismic data of the seismic trace to be inverted, and the second well bypass seismic data of the first number of first sample wells; The fifth determining module is used to determine the full-frequency initial inversion result of the seismic trace to be inverted based on the second logging data and the second well bypass seismic data of the first sample wells of the first number of sample wells; The correction module is used to correct the initial full-frequency inversion results based on the high-precision sequence stratigraphic framework and the at least two frequency bands, so as to obtain the full-frequency target inversion results of the seismic trace to be inverted.

25. An electronic device, characterized in that, include: Memory and processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory to implement the method as described in any one of claims 1-23.

26. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method described in any one of claims 1-23.

27. A computer program product, characterized in that, Includes a computer program that, when executed by a computer, implements the method described in any one of claims 1-23.

Images