A method and system for improving resolution of post-stack data

Abstract

The present invention belongs to the technical field of seismic data processing, and specifically discloses a method and system for improving the resolution of post-stack data. Under the premise of statistical signal-to-noise ratio spectrum, the present invention performs dominant frequency band analysis on the preferred regional data and estimates the time-varying wavelet amplitude spectrum, designs an operator according to the octave in the S domain to perform frequency band widening according to the octave, and statistically calculates the energy change function of each channel of the post-stack data set in frequency bands based on the data after spectrum expansion, fits the amplitude compensation coefficient that varies with time / depth, and comprehensively applies it to the data to be adjusted in different regional locations, further eliminating the wavelet differences of post-stack data between different regions, obtaining a high-resolution post-stack data body that tends to be consistent in space, and further improving the accuracy of post-stack inversion and reservoir prediction. The present invention is suitable for improving the resolution of post-stack data in the process of seismic data processing.

Description

Technical Field

[0001] The present invention belongs to the technical field of seismic data processing, and relates to a method and system for improving the resolution of post-stack data. Background Art

[0002] When post-stack seismic data are processed using conventional resolution enhancement methods for spectrum broadening, the effects of pre-stack processing parameters and unresolved near-surface absorption attenuation can easily lead to poor spatial consistency of seismic wavelets and low signal-to-noise ratios. This makes it difficult to effectively reflect the lateral distribution of seismic responses in lithologic formations, further impacting subsequent reservoir prediction. Therefore, improvements and optimization of post-stack resolution enhancement methods are needed.

[0003] Current methods for improving the resolution of post-stack data based on deconvolution methods mainly assume that the seismic wavelet is a stationary signal and the reflection sequence is a white noise sequence. Single-channel or multi-channel spectral statistics are performed on the post-stack data, and then the data is filtered by the inverse operator obtained based on the expected wavelet calculation. Since this method does not take into account the changes in the signal-to-noise ratio in different frequency bands and the energy characteristics of the dominant frequency band of the data itself, it is very easy to cause the defect of significantly reducing the signal-to-noise ratio after improving the resolution. At the same time, since the spatial stability assumption cannot be met, the data in different regions cannot obtain the same high-resolution results. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for improving the resolution of post-stack data, so as to eliminate the wavelet differences of post-stack data between different regions, obtain a spatially consistent high-resolution post-stack data volume, and further improve the accuracy of post-stack inversion and reservoir prediction;

[0005] A second object of the present invention is to provide a system for improving the resolution of post-stack data for implementing the above method.

[0006] In order to achieve the above-mentioned purpose, the present invention adopts the following technical solutions:

[0007] A method for improving the resolution of post-stack data is performed in the following order:

[0008] S1. Perform signal-to-noise ratio analysis of the target layer based on the input post-stack data and select high signal-to-noise ratio areas;

[0009] S2. Perform spectrum analysis in the high signal-to-noise ratio region to select broadband model traces and transform them into the S domain. Fit the time-varying wavelet amplitude spectrum with the trace gather amplitude spectrum to obtain the estimated time-varying wavelet amplitude spectrum.

[0010] S3. In the S domain, an inverse operator is designed based on the estimated time-varying wavelet amplitude spectrum, and the spectrum is broadened to high and low frequency bands according to the octave band to obtain the model channel spectrum, that is, the high-resolution model data in the S domain;

[0011] S4, constructing a frequency division filter function in the S domain, performing frequency division filtering on the high-resolution model data and the gather to be optimized using the frequency division filter function in the S domain, and obtaining the model data and the data to be optimized after frequency division;

[0012] S5. In each frequency range, use an exponential polynomial to fit the absorption attenuation coefficient of the frequency range to obtain adjusted model data;

[0013] S6. Using the adjusted model data as a reference, frequency band energy correction is performed on the data to be optimized by frequency division and space division to obtain spatially uniform high-resolution data, i.e., the adjusted seismic data.

[0014] S7. Reconstruct the adjusted seismic data and complete the amplitude and frequency consistency adjustment.

[0015] As a limitation, when selecting the broadband model channel in step S2, the stacked multi-channel data of multiple wells in the study area are selected as the reference model for amplitude and frequency adjustment, and transformed into the S domain according to formula ①

[0016]

[0017] Among them, S(τ,f) represents the data after S domain transformation, x(t) represents the time domain data, is the Gaussian window function, f represents the frequency, τ and t represent the time variables, and i represents the imaginary unit;

[0018] The spectrum of a data channel after S transformation is expressed as s(τ, f, x j ), j represents the track number.

[0019] As a further limitation, step S3 is performed in the following order:

[0020] S31. According to formula ②, the estimated time-varying wavelet amplitude spectrum obtained after S-domain transformation is filtered and weighted to eliminate the local abnormal attenuation of the wavelet while retaining the background attenuation.

[0021]

[0022] Among them, A lf (τ, f, x j ) is the autocorrelation spectrum of the data after low-pass filtering of channel j, a j is the weighting coefficient of channel j;

[0023] S32. In different ranges of the time-varying wavelet amplitude spectrum, different inverse operator parameters are designed according to the octave band. The inverse operator is expressed as formula ③

[0024]

[0025] Among them, S fw (max) represents the peak energy of the time-varying wavelet amplitude spectrum; η is 1 in the dominant frequency band and a positive real number less than 1 in the extended octave band; ε represents the stability factor;

[0026] S33, using formula ③, apply the operator to the broadband model channel selected in step S2 to obtain the high-resolution model data in the S domain shown in formula ④

[0027]

[0028] As a further limitation, in step S4, frequency division filtering is performed according to formula ⑤

[0029] X pj (τ, f i )=S pj (τ, f i )*w(f i ) Formula ⑤

[0030] Among them, S pj (τ, f i ) represents the S-transformation of the input model data; X pj (τ, f i ) is the model data after frequency division filtering; ω(f i ) is the frequency division filter function; f i is the i-th frequency band.

[0031] As a further limitation, in step S4, before frequency division filtering, the model data and the track set to be optimized are first divided into several frequency bands within the frequency band range commonly seen in seismic data, wherein the 1ms sampling is divided into 21-29 frequency bands, and the 2ms sampling is divided into 15-17 frequency bands.

[0032] As a further limitation, in step S5, the data after frequency spreading in step S3 is used as a model, and the objective function is defined as follows in the least squares sense:

[0033]

[0034] Among them, a M (t,f i )=a1+a2t+…a n t n , represents the absorption attenuation coefficient of the model data after frequency division.

[0035] As a further limitation, in step S6, based on the obtained amplitude energy absorption attenuation coefficient of each frequency band of the adjusted model data, formula ⑦ is used for the input data to perform spatially unified gather band energy adjustment in different frequency bands within the time window to complete the band energy correction.

[0036]

[0037] Where D(τ, f) ji is the output data after amplitude energy adjustment based on the model channel, d(τ, f) ji The gather data to be optimized.

[0038] As a further limitation, in step S7, during the reconstruction process, the adjusted seismic data is inversely transformed back to the time-space domain according to formula ⑧

[0039]

[0040] Formula ⑧ is the inverse S transform of the S transform described in formula ①.

[0041] A system for improving the resolution of post-stack data, used for implementing the above-mentioned method for improving the resolution of post-stack data, comprising: a signal-to-noise ratio analysis module, a spectrum analysis module, an S-transformation module, a fitting module, a frequency division filtering module, an adjustment module, a correction module, and a reconstruction module;

[0042] Among them, the signal-to-noise ratio analysis module performs signal-to-noise ratio analysis of the target layer based on the input post-stack data, selects the high signal-to-noise ratio area, and outputs it to the spectrum analysis module;

[0043] The spectrum analysis module is used to perform spectrum analysis in the high signal-to-noise ratio area to select the broadband model channel and output it to the S-transform module;

[0044] The S-transform module is used to perform S-transform based on the received data and to design the inverse operator;

[0045] The fitting module is used to fit the time-varying wavelet amplitude spectrum through the gather amplitude spectrum;

[0046] The frequency division filter module is used to construct a frequency division filter function and perform frequency division filtering on the received data;

[0047] The adjustment module uses an exponential polynomial to fit the absorption attenuation coefficient of each frequency band and outputs the adjusted data to the correction module;

[0048] The correction module is used to correct the data to be optimized and output it to the reconstruction module;

[0049] The reconstruction module is used to reconstruct the adjusted seismic data.

[0050] Due to the adoption of the above technical solution, the present invention has achieved the following technical advancements compared with the prior art:

[0051] (1) Based on the statistical signal-to-noise ratio spectrum, the present invention optimizes regional data for dominant frequency band analysis and estimates the time-varying wavelet amplitude spectrum, designs an operator in the S domain according to the octave band to perform frequency band broadening according to the octave band, and statistically calculates the energy change function of each trace of the post-stack data set in frequency bands based on the data after spectrum expansion, fits the amplitude compensation coefficient that varies with time / depth, and comprehensively applies it to the data to be adjusted in different regional locations, further eliminating the wavelet differences of post-stack data between different regions, obtaining a high-resolution post-stack data volume that tends to be consistent in space, and further improving the accuracy of post-stack inversion and reservoir prediction;

[0052] (2) The present invention selects high signal-to-noise ratio main frequency band data through frequency band signal-to-noise ratio analysis, performs S-domain transformation and estimates time-varying wavelets, applies inverse operators according to octaves, and obtains broadband model data; the amplitude energy of the overall data in the S-space domain is adjusted with reference to the broadband model data to obtain the final broadband result, which to a certain extent avoids the assumptions that are difficult to meet in deconvolution methods, constrains the frequency band extension degree through frequency band signal-to-noise ratio analysis, and obtains high-resolution data with a high signal-to-noise ratio, effectively improving the ability to identify thin reservoirs and laying the foundation for subsequent reservoir description and well deployment;

[0053] (3) The present invention adopts S-domain transformation, combines the advantages of Fourier transform and wavelet transform, and can provide a fine time-frequency spectrum;

[0054] (4) The present invention uses a specific function to perform frequency division filtering on the model data and the gather to be optimized in the S domain. Compared with methods such as the wavelet domain, the advantage of the present invention is the smaller Gibbs effect and frequency leakage.

[0055] (5) The present invention designs a purely data-driven S-domain spreading operator based on the post-stack data with different characteristics in different regions. By optimizing parameters, the vertical resolution of the post-stack data is improved without damaging the signal-to-noise ratio, and the amplitude energy is adjusted in the spatial domain so that the data obtains consistent broadband properties in the spatial direction. This method has good applicability to post-stack data, and on this basis can effectively improve the accuracy of post-stack inversion and reservoir prediction.

[0056] The invention belongs to the technical field of seismic data processing and can improve the resolution of post-stack data. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention.

[0058] In the attached figure:

[0059] Figure 1 This is a flow chart of Example 1 of the present invention;

[0060] Figure 2 This is a cross-sectional comparison diagram of the model gather before and after S-domain frequency spreading in Example 1 of the present invention;

[0061] Figure 2 a is a cross-sectional view of the model gather before S-domain frequency spreading in Example 1 of the present invention;

[0062] Figure 2 b is a cross-sectional view of the model gather after S-domain frequency spreading in Example 1 of the present invention;

[0063] Figure 3 This is a comparison diagram of the target layer spectrum before and after S-domain spectrum spreading of the model gather in Example 1 of the present invention;

[0064] Figure 3 a is a spectrum diagram of the target layer before S-domain spectrum spreading of the model gather in Example 1 of the present invention;

[0065] Figure 3 b is the spectrum diagram of the target layer after S-domain spreading of the model gather in Example 1 of the present invention;

[0066] Figure 4 This is a statistical comparison diagram of the signal-to-noise ratio of the target layer spectrum before and after S-domain spectrum spreading of the model channel in Example 1 of the present invention;

[0067] Figure 4 a is a statistical diagram of the spectrum signal-to-noise ratio along the target layer before S-domain spectrum spreading of the model channel in Example 1 of the present invention;

[0068] Figure 4 b is a statistical diagram of the signal-to-noise ratio of the target layer spectrum along the layer after S-domain spectrum spreading of the model channel in Example 1 of the present invention;

[0069] Figure 5 This is a cross-section comparison diagram before and after improving the signal-to-noise ratio based on post-stack data of the model trace in Example 1 of the present invention;

[0070] Figure 5 a is a cross-section diagram before improving the signal-to-noise ratio based on the post-stack data of the model trace in Example 1 of the present invention;

[0071] Figure 5 b is a cross-sectional view after improving the signal-to-noise ratio based on the post-stack data of the model trace according to Example 1 of the present invention;

[0072] Figure 6 This is a diagram showing the target layer well-seismic calibration result after improving the signal-to-noise ratio of post-stack data in Example 1 of the present invention;

[0073] Figure 7 This is a diagram of the post-stack inversion result after the frequency-boosting processing of the gather in Example 1 of the present invention. DETAILED DESCRIPTION

[0074] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention and are not used to limit the present invention.

[0075] Example 1: A method for improving resolution of post-stack data

[0076] This embodiment provides an example of adjusting the amplitude spectrum energy based on the S-domain transform spectrum extension for post-stack data, thereby achieving high-resolution processing of post-stack data and improving the accuracy of post-stack inversion. Figure 1 As shown, this embodiment is carried out in the following steps in sequence:

[0077] S1. Perform signal-to-noise ratio analysis of the target layer based on the input post-stack data and select high signal-to-noise ratio areas;

[0078] The high signal-to-noise ratio area is determined according to the data. In this embodiment, the high signal-to-noise ratio area is the top 10% selected as a reference after performing a target layer signal-to-noise ratio analysis.

[0079] S2. Perform spectrum analysis in the high signal-to-noise ratio region to select broadband model traces and transform them into the S domain. Fit the time-varying wavelet amplitude spectrum with the trace gather amplitude spectrum to obtain the estimated time-varying wavelet amplitude spectrum.

[0080] The high signal-to-noise ratio region is the nature of the data itself. Selecting the broadband region during spectrum analysis in the high signal-to-noise ratio region is a progressive screening process of selecting the best from the best. Ultimately, the high signal-to-noise ratio broadband data is selected as the model data, or reference data.

[0081] S3. In the S domain, an inverse operator is designed based on the estimated time-varying wavelet amplitude spectrum, and the spectrum is broadened to high and low frequency bands according to the octave band to obtain the model channel spectrum, that is, the high-resolution model data in the S domain;

[0082] S4, constructing a frequency division filter function in the S domain, performing frequency division filtering on the high-resolution model data and the gather to be optimized using the frequency division filter function in the S domain, and obtaining the model data and the data to be optimized after frequency division;

[0083] S5. For the divided model data, in each frequency range, use an exponential polynomial to fit the absorption attenuation coefficient of the frequency band to obtain adjusted model data;

[0084] S6. Using the adjusted model data as a reference, frequency band energy correction is performed on the data to be optimized by frequency division and space division to obtain spatially uniform high-resolution data, i.e., the adjusted seismic data.

[0085] S7. Reconstruct the adjusted seismic data and complete the amplitude and frequency consistency adjustment.

[0086] Specifically, when selecting the broadband model channel in step S2, the stacked multi-channel data of multiple wells in the study area are selected as the reference model for amplitude and frequency adjustment. When there are fewer wells, the reference channel gathers of the non-well control area can be added and transformed into the S domain according to formula ①.

[0087]

[0088] Among them, S(τ,f) represents the data after S domain transformation, x(t) represents the time domain data, is the Gaussian window function, f represents the frequency, τ and t represent the time variables, and i represents the imaginary unit;

[0089] The S domain combines the advantages of Fourier transform and wavelet transform, and can provide a fine time-frequency spectrum. The spectrum of a data channel after S transform is expressed as s(τ, f, x j ), j represents the track number.

[0090] Then, step S3 is performed in the following order:

[0091] S31. According to formula ②, the estimated time-varying wavelet amplitude spectrum obtained after S-domain transformation is filtered and weighted to eliminate the local abnormal attenuation of the wavelet while retaining the background attenuation.

[0092]

[0093] Among them, A lf (τ, f, x j ) is the autocorrelation spectrum of the data after low-pass filtering of channel j, a j is the weighting coefficient of channel j;

[0094] S32. In different ranges of the time-varying wavelet amplitude spectrum, different inverse operator parameters are designed according to the octave band. The inverse operator is expressed as formula ③

[0095]

[0096] Among them, S fw (max) represents the peak energy of the time-varying wavelet amplitude spectrum; η is 1 in the dominant frequency band and a positive real number less than 1 in the extended octave band; in this embodiment, the dominant frequency band is the frequency band range agreed upon by -20 dB; ε represents the stability factor, which is generally small and ranges from [0,1];

[0097] S33, using formula ③, apply the operator to the broadband model channel selected in step S2 to obtain the high-resolution model data in the S domain shown in formula ④

[0098]

[0099] In step S4, before frequency filtering, the model data and the gather to be optimized are divided into several frequency segments within the common frequency band of seismic data. For 1ms sampling, the frequency bands are 21-29, and for 2ms sampling, the frequency bands are 15-17. The common frequency band of seismic data is determined based on the actual data. In this embodiment, according to current broadband acquisition standards, the frequency band is set to approximately 1.5-80 Hz. The specific implementation process requires analysis during actual operation.

[0100] Specifically, frequency division filtering is performed according to formula ⑤

[0101] X pj (τ, f i )=S pj (τ, f i )*w(f i ) Formula ⑤

[0102] Among them, S pj (τ, f i ) represents the S-transformation of the input model data; X pj (τ, f i ) is the model data after frequency division filtering; ω(f i ) is the frequency division filter function; f i is the i-th frequency band.

[0103] Assuming that the energy attenuation of the 3D work area is stable on a large scale underground, most of the energy differences between post-stack gathers still come from the surface consistency energy compensation and the residual terms of wavelet processing in pre-stack processing. The main differences come from the drastic changes in near-surface conditions. The differences in complex underground structures or lithology have been basically resolved in pre-stack processing. Therefore, in step S5, we use the data after frequency extension in step S3 as a model, and define the objective function in the least squares sense as

[0104]

[0105] Among them, a M (t,f i )=a1+a2t+…a n t n , represents the absorption attenuation coefficient of the model data after frequency division.

[0106] In step S6, based on the obtained amplitude energy absorption attenuation coefficient of each frequency band of the adjusted model data, the input data is subjected to spatially unified gather band energy adjustment in different frequency bands within the time window using formula ⑦, that is, the amplitude spectrum of the optimized data is adjusted to complete the band energy correction.

[0107]

[0108] Where D(τ, f) ji is the output data after amplitude energy adjustment based on the model channel, d(τ, f) ji is the gather data to be optimized, a M is the absorption attenuation coefficient of the model data after frequency division.

[0109] Finally, in step S7, during the reconstruction process, the adjusted seismic data is inversely transformed back to the time-space domain according to formula ⑧

[0110]

[0111] Formula ⑧ is the inverse S transform of the S transform described in formula ①.

[0112] In many blocks such as the Huabei Oilfield Exploration Area and the Songliao Basin, conventional deconvolution and spectral whitening methods based on the steady-state convolution model rely on the reflection coefficient of white noise and the assumption of steady-state spatial wavelets. While actually improving data resolution, they also reduce the signal-to-noise ratio of the data, seriously affecting the accuracy of subsequent pre-stack inversion and AVO analysis, and thus require optimization.

[0113] The method provided in this embodiment was applied in a certain work area in the North China Oilfield Exploration Area and the Songliao Basin. Figure 2 This is a comparison diagram of the model gather profiles before and after S-domain frequency expansion; Figure 2 a is the profile of the model gather before S-domain frequency spreading; Figure 2 b is the cross-section of the model gather after S-domain frequency expansion. Figure 2 b It can be seen that compared with the original model trace, the formation wave groups are significantly increased and the profile resolution is significantly improved.

[0114] Figure 3 The comparison diagram of the target layer spectrum before and after the S-domain spreading of the model gather is shown in Fig. Figure 3 a is the spectrum of the target layer before S-domain spreading of the model gather; Figure 3 b is the spectrum of the target layer after the model gather is spread in the S domain. By comparison, it can be seen that the absolute width of the frequency band is significantly increased after the spread, and the energy of each frequency band is relatively consistent.

[0115] Figure 4 The figure is a statistical comparison of the signal-to-noise ratio of the target layer spectrum before and after the S-domain frequency spreading of the model channel; Figure 4 a is the spectrum signal-to-noise ratio statistics of the target layer before S-domain spectrum spreading of the model channel; Figure 4 (b) shows the signal-to-noise ratio statistics of the target layer spectrum after S-domain frequency spreading of the model channel. It can be seen that the data after frequency spreading significantly improves the resolution while maintaining a high signal-to-noise ratio, indicating that this method mainly enhances the effective wave energy.

[0116] Figure 5Comparison diagram of sections before and after improving the signal-to-noise ratio of post-stack data based on model traces; Figure 5 a is the cross section before improving the signal-to-noise ratio of post-stack data based on the model trace; Figure 5 b is the cross-section after improving the signal-to-noise ratio of post-stack data based on the model trace. Figure 6 The result of well-seismic calibration of the target layer after improving the signal-to-noise ratio of post-stack data. Figure 5 and Figure 6 The data before and after the time-frequency-space domain amplitude energy adjustment show that the characteristic seismic response of the thin oil shale reservoir could not be determined before the resolution was improved. After the resolution was improved, the location of the oil shale in the target layer in the data showed a good correspondence with the logging interpretation results.

[0117] Figure 7 This is the post-stack inversion result diagram after the gather frequency is increased. It can be seen that the pre-stack inversion result of the gather with improved resolution is more accurate.

[0118] Example 2: A system for improving resolution of post-stack data

[0119] This embodiment is used to implement the method provided in Example 1. This embodiment includes: a signal-to-noise ratio analysis module, a spectrum analysis module, an S-transformation module, a fitting module, a frequency division filtering module, an adjustment module, a correction module, and a reconstruction module.

[0120] Among them, the signal-to-noise ratio analysis module performs signal-to-noise ratio analysis of the target layer partition based on the input post-stack data, selects the high signal-to-noise ratio area, and outputs it to the spectrum analysis module.

[0121] The spectrum analysis module is used to perform spectrum analysis in the high signal-to-noise ratio area to select the broadband model channel and output it to the S transform module.

[0122] The S-transform module is used, on the one hand, to perform S-transform according to the received data, and on the other hand, to design an inverse operator.

[0123] The fitting module is used to fit the time-varying wavelet amplitude spectrum through the gather amplitude spectrum.

[0124] The frequency division filtering module is used to construct a frequency division filtering function and perform frequency division filtering on the received data.

[0125] The adjustment module uses an exponential polynomial to fit the absorption attenuation coefficient of each frequency band and outputs the adjusted data to the correction module.

[0126] The correction module is used to correct the data to be optimized and output it to the reconstruction module.

[0127] The reconstruction module is used to reconstruct the adjusted seismic data.

Claims

1. A method for improving the resolution of post-stack data, characterized in that: Follow these steps in order: S1. Perform signal-to-noise ratio analysis of the target layer based on the input post-stack data and select high signal-to-noise ratio areas; S2. Perform spectrum analysis in the high signal-to-noise ratio region to select broadband model traces and transform them into the S domain. Fit the time-varying wavelet amplitude spectrum with the trace gather amplitude spectrum to obtain the estimated time-varying wavelet amplitude spectrum. S3. In the S domain, an inverse operator is designed based on the estimated time-varying wavelet amplitude spectrum, and the spectrum is broadened to high and low frequency bands according to the octave band to obtain the model channel spectrum, that is, the high-resolution model data in the S domain; S4, constructing a frequency division filter function in the S domain, performing frequency division filtering on the high-resolution model data and the gather to be optimized using the frequency division filter function in the S domain, and obtaining the model data and the data to be optimized after frequency division; S5. In each frequency range, use an exponential polynomial to fit the absorption attenuation coefficient of the frequency range to obtain adjusted model data; S6. Using the adjusted model data as a reference, frequency band energy correction is performed on the data to be optimized by frequency division and space division to obtain spatially uniform high-resolution data, i.e., the adjusted seismic data. S7. Reconstruct the adjusted seismic data and complete the amplitude and frequency consistency adjustment.

2. The method for improving resolution of post-stack data according to claim 1, characterized in that: When selecting the broadband model channel in step S2, the stacked multi-channel data of multiple wells in the study area are selected as the reference model for amplitude and frequency adjustment, and transformed into the S domain according to formula ① Among them, S(τ,f) represents the data after S domain transformation, x(t) represents the time domain data, is the Gaussian window function, f represents the frequency, τ and t represent the time variables, and i represents the imaginary unit; The spectrum of a data channel after S transformation is expressed as s(τ, f, x j ), j represents the track number.

3. The method for improving the resolution of post-stack data according to claim 2, characterized in that ,Step S3 is performed in the following order: S31. According to formula ②, the estimated time-varying wavelet amplitude spectrum obtained after S-domain transformation is filtered and weighted to eliminate the local abnormal attenuation of the wavelet while retaining the background attenuation. Among them, A lf (τ, f, x j ) is the autocorrelation spectrum of the data after low-pass filtering of channel j, a j is the weighting coefficient of channel j; S32. In different ranges of the time-varying wavelet amplitude spectrum, different inverse operator parameters are designed according to the octave band. The inverse operator is expressed as formula ③ Among them, S fw (max) represents the peak energy of the time-varying wavelet amplitude spectrum; η is 1 in the dominant frequency band and a positive real number less than 1 in the extended octave band; ε represents the stability factor; S33, using formula ③, apply the operator to the broadband model channel selected in step S2 to obtain the high-resolution model data in the S domain shown in formula ④ 4. A method for improving resolution of post-stack data according to claim 3, characterized in that In step S4, frequency division filtering is performed according to formula ⑤ X pj (τ, f i ) = S pj (τ, f i ) * w(f i ) Equation ⑤ Among them, S pj (τ, f i ) represents the S-transformation of the input model data; X pj (τ, f i ) is the model data after frequency division filtering; ω(f i ) is the frequency division filter function; f i is the i-th frequency band.

5. The method for improving resolution of post-stack data according to claim 4, characterized in that In step S4, before frequency filtering, the model data and the gather to be optimized are first divided into several frequency bands within the common frequency band range of seismic data. For 1ms sampling, the frequency bands are 21-29, and for 2ms sampling, the frequency bands are 15-17.

6. A method for improving resolution of post-stack data according to claim 4 or 5, characterized in that In step S5, the data after frequency extension in step S3 is used as the model, and the objective function is defined as Among them, a M (t,f i )=a1+a2t+…a n t n , represents the absorption attenuation coefficient of the model data after frequency division.

7. The method for improving resolution of post-stack data according to claim 6, characterized in that In step S6, based on the obtained amplitude energy absorption attenuation coefficient of each frequency band of the adjusted model data, the input data is subjected to the formula ⑦ to perform spatially unified gather band energy adjustment in different frequency bands within the time window to complete the band energy correction. Where D(τ, f) ji is the output data after amplitude energy adjustment based on the model channel, d(τ, f) ji The gather data to be optimized.

8. The method for improving resolution of post-stack data according to claim 7, characterized in that In step S7, during the reconstruction process, the adjusted seismic data is transformed back to the time-space domain according to formula ⑧ Formula ⑧ is the inverse S transform of the S transform described in formula ①.

9. A system for improving the resolution of post-stack data, used for implementing the method for improving the resolution of post-stack data according to any one of claims 1 to 8, characterized in that: The system includes: a signal-to-noise ratio analysis module, a spectrum analysis module, an S-transformation module, a fitting module, a frequency division filtering module, an adjustment module, a correction module and a reconstruction module; Among them, the signal-to-noise ratio analysis module performs signal-to-noise ratio analysis of the target layer based on the input post-stack data, selects the high signal-to-noise ratio area, and outputs it to the spectrum analysis module; The spectrum analysis module is used to perform spectrum analysis in the high signal-to-noise ratio area to select the broadband model channel and output it to the S-transform module; The S-transform module is used to perform S-transform based on the received data and to design the inverse operator; The fitting module is used to fit the time-varying wavelet amplitude spectrum through the gather amplitude spectrum; The frequency division filter module is used to construct a frequency division filter function and perform frequency division filtering on the received data; The adjustment module uses an exponential polynomial to fit the absorption attenuation coefficient of each frequency band and outputs the adjusted data to the correction module; The correction module is used to correct the data to be optimized and output it to the reconstruction module; The reconstruction module is used to reconstruct the adjusted seismic data.

Images