A method and apparatus for seismic data processing based on three-dimensional predictive deconvolution

Abstract

A method and device for processing seismic data based on three-dimensional predictive deconvolution, the method comprising: determining an operator for three-dimensional predictive deconvolution based on a target seismic data of a study area and a pre-established three-dimensional predictive deconvolution model; and performing three-dimensional predictive deconvolution processing on the target seismic data according to the operator for three-dimensional predictive deconvolution to obtain processed seismic data.

Description

Technical Field

[0001] This article relates to the field of seismic data processing technology, and in particular to a seismic data processing method and apparatus based on three-dimensional predictive deconvolution. Background Technology

[0002] Multiples are a difficult and challenging problem in seismic data processing, especially in marine data processing. Therefore, there is an urgent need for a technique to suppress shallow-water multiples in marine seismic data processing to improve data resolution. Predictive deconvolution differs from other multiple denoising methods such as parabolic Radon transform and SRME. Suppressing multiples does not require auxiliary information such as velocity, and the prediction step size is easily obtained through autocorrelation. Furthermore, predictive deconvolution has the advantage of simple parameter settings, making it a commonly used method for suppressing multiples. It has been well applied in eliminating false reflections, mixed echoes, or other simple forms of multiples. Some seismic data processing methods use one-dimensional predictive deconvolution; therefore, implementing a three-dimensional predictive deconvolution method for seismic data processing is a pressing issue. Summary of the Invention

[0003] This application provides a seismic data processing method and apparatus based on three-dimensional predictive deconvolution. This method can consider multiple directions of three-dimensional seismic data, predict the three-dimensional predictive deconvolution operator, and then perform seismic data processing to obtain seismic data with greater spatial coherence. At the same time, compared with the one-dimensional predictive deconvolution seismic data processing method, it improves the resolution of seismic data and enhances the quality of seismic data processing and migration imaging.

[0004] In a first aspect, this application provides a seismic data processing method based on three-dimensional predictive deconvolution, the method comprising:

[0005] Based on the target seismic data of the study area, the operators for three-dimensional predictive deconvolution are determined using a pre-established three-dimensional predictive deconvolution model;

[0006] The target seismic data is subjected to three-dimensional predictive deconvolution processing based on the three-dimensional predictive deconvolution operator to obtain processed seismic data.

[0007] In one exemplary embodiment, the target seismic data is three-dimensional common-sensor point seismic gather data.

[0008] In one exemplary embodiment, the pre-established 3D predictive deconvolution model is:

[0009]

[0010] In the above model, d b,m,jThis represents the common-receiver gather data with shot line number b, shot point number m, time direction sample point number j, and containing multiples; p b,m,j This represents the common-receiver gather data of shot line number b, shot point number m, time direction sample point number j, and without multiples; f c,n,k d represents the operator for 3D predictive deconvolution; c represents the sample point number of the 3D predictive operator corresponding to b, n represents the sample point number of the 3D predictive operator corresponding to m, and k represents the sample point number of the 3D predictive operator corresponding to j; w represents the number of half-width points in the Crossline direction, u represents the number of half-width points in the Inline direction; g represents the step size for 3D predictive deconvolution calculation; d b-c,m-n,j-k This represents the common-detector seismic gather data with shot line number bc, shot point number mn, time direction sample point number jk, and containing multiple waves.

[0011] In one exemplary embodiment, the process of establishing the three-dimensional predictive deconvolution model includes:

[0012] One-dimensional seismic data is determined based on the reflection coefficient and the seismic wavelet; wherein the reflection coefficient is a white noise sequence and the seismic wavelet is a wavelet with minimum phase.

[0013] Based on the statistical characteristics of earthquake data, a one-dimensional predictive deconvolution model is determined;

[0014] The three-dimensional predictive deconvolution model is determined based on the one-dimensional predictive deconvolution model and the three-dimensional properties of the three-dimensional seismic data.

[0015] In one exemplary embodiment, determining one-dimensional seismic data based on the reflection coefficient and the seismic wavelet includes:

[0016] One-dimensional seismic data are calculated using the seismic convolution formula based on the reflection coefficient and seismic wavelet. The seismic convolution formula is as follows:

[0017] d(t) = b(t) * f(t)

[0018] In the above formula, d(t) represents the seismic data in the time domain, b(t) represents the seismic wavelet in the time domain, f(t) represents the reflection coefficient in the time domain, and t represents the time of the seismic data.

[0019] In one exemplary embodiment, determining a one-dimensional predictive deconvolution model based on the statistical characteristics of seismic data includes:

[0020] Determine the relationship between earthquake data and time based on the statistical characteristics of earthquake data;

[0021] A one-dimensional predictive deconvolution model is established based on the determined relationship between earthquake data and time.

[0022] The one-dimensional predictive deconvolution model is as follows:

[0023]

[0024] In the above one-dimensional predictive deconvolution model, d j This represents the seismic gather data of the common-detector point with sample number j in the time direction and containing multiple waves; p j This represents the seismic gather data of the common receiver point with sample number j in the time direction and without multiples; d j-k This represents the seismic gather data of common-detector points with time-direction sample number jk that contain multiples; f k This represents an operator for predicting deconvolution in one dimension.

[0025] In a second aspect, embodiments of the present invention provide a seismic data processing device based on three-dimensional predictive deconvolution, the device comprising: an operator module for determining three-dimensional predictive deconvolution and a seismic data processing module;

[0026] The module for determining the operator of the three-dimensional predicted deconvolution is used to determine the operator of the three-dimensional predicted deconvolution based on the target seismic data of the study area and using a pre-established three-dimensional predicted deconvolution model.

[0027] The seismic data processing module is used to perform three-dimensional predictive deconvolution processing on the target seismic data according to the three-dimensional predictive deconvolution operator to obtain processed seismic data.

[0028] In one exemplary embodiment, the pre-established 3D predictive deconvolution model is:

[0029]

[0030] In the above model, d b,m,j This represents the common-receiver gather data with shot line number b, shot point number m, time direction sample point number j, and containing multiples; p b,m,j This represents the common-receiver gather data of shot line number b, shot point number m, time direction sample point number j, and without multiples; f c,n,k d represents the operator for 3D predictive deconvolution; c represents the sample point number of the 3D predictive operator corresponding to b, n represents the sample point number of the 3D predictive operator corresponding to m, and k represents the sample point number of the 3D predictive operator corresponding to j; w represents the number of half-width points in the Crossline direction, u represents the number of half-width points in the Inline direction; g represents the step size for 3D predictive deconvolution calculation; d b-c,m-n,j-k This represents the common-detector seismic gather data with shot line number bc, shot point number mn, time direction sample point number jk, and containing multiple waves.

[0031] In one exemplary embodiment, the process of establishing the three-dimensional predictive deconvolution model includes:

[0032] One-dimensional seismic data is determined based on the reflection coefficient and the seismic wavelet; wherein the reflection coefficient is a white noise sequence and the seismic wavelet is a wavelet with minimum phase.

[0033] Based on the statistical characteristics of earthquake data, a one-dimensional predictive deconvolution model is determined;

[0034] The three-dimensional predictive deconvolution model is determined based on the one-dimensional predictive deconvolution model and the three-dimensional properties of the three-dimensional seismic data.

[0035] In one exemplary embodiment, determining one-dimensional seismic data based on the reflection coefficient and the seismic wavelet includes:

[0036] One-dimensional seismic data are calculated using the seismic convolution formula based on the reflection coefficient and seismic wavelet. The seismic convolution formula is as follows:

[0037] d(t) = b(t) * f(t)

[0038] In the above formula, d(t) represents the seismic data in the time domain, b(t) represents the seismic wavelet in the time domain, f(t) represents the reflection coefficient in the time domain, and t represents the time of the seismic data.

[0039] In one exemplary embodiment, determining a one-dimensional predictive deconvolution model based on the statistical characteristics of seismic data includes:

[0040] Determine the relationship between earthquake data and time based on the statistical characteristics of earthquake data;

[0041] A one-dimensional predictive deconvolution model is established based on the determined relationship between earthquake data and time.

[0042] The one-dimensional predictive deconvolution model is as follows:

[0043]

[0044] In the above one-dimensional predictive deconvolution model, d j This represents the seismic gather data of the common-detector point with sample number j in the time direction and containing multiple waves; p j This represents the seismic gather data of the common receiver point with sample number j in the time direction and without multiples; d j-k This represents the seismic gather data of common-detector points with time-direction sample number jk that contain multiples; f k This represents an operator for predicting deconvolution in one dimension.

[0045] Thirdly, embodiments of the present invention provide a computer program product with seismic data processing function based on three-dimensional predictive deconvolution, comprising a computer program / instruction, characterized in that, when the computer program / instruction is executed by a processor, it implements the seismic data processing method based on three-dimensional predictive deconvolution as described in any of the above embodiments.

[0046] Compared with related technologies, this application provides a seismic data processing method and apparatus based on three-dimensional predictive deconvolution. The method includes: determining a three-dimensional predictive deconvolution operator based on target seismic data of a study area using a pre-established three-dimensional predictive deconvolution model; and performing three-dimensional predictive deconvolution processing on the target seismic data according to the three-dimensional predictive deconvolution operator to obtain processed seismic data. This application considers multiple directions of three-dimensional seismic data, predicts a three-dimensional predictive deconvolution operator, and then performs seismic data processing to obtain seismic data with better spatial coherence. Simultaneously, compared with one-dimensional predictive deconvolution seismic data processing methods, it improves the resolution of seismic data and enhances the quality of seismic data processing and migration imaging.

[0047] Other features and advantages of this application will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the application. Other advantages of this application can be realized and obtained by means of the solutions described in the description and the accompanying drawings. Attached Figure Description

[0048] The accompanying drawings are used to provide an understanding of the technical solutions of this application and constitute a part of the specification. They are used together with the embodiments of this application to explain the technical solutions of this application and do not constitute a limitation on the technical solutions of this application.

[0049] Figure 1 This is a flowchart of a seismic data processing method based on three-dimensional predictive deconvolution according to an embodiment of this application;

[0050] Figure 2 These are schematic diagrams comparing one-dimensional and three-dimensional predictive deconvolution operators in some exemplary embodiments;

[0051] Figure 3 This is a schematic diagram of a seismic data processing device based on three-dimensional predictive deconvolution according to an embodiment of this application;

[0052] Figure 4 These are schematic diagrams comparing the effects of 3D model prediction deconvolution processing in some exemplary embodiments;

[0053] Figure 5 These are schematic diagrams illustrating the far-middle-near arrangement before the deconvolution process in some exemplary embodiments;

[0054] Figure 6These are schematic diagrams illustrating the predicted far-middle-near arrangement effect after deconvolution processing in some exemplary embodiments. Detailed Implementation

[0055] This application describes several embodiments, but these descriptions are exemplary and not restrictive, and it will be apparent to those skilled in the art that many more embodiments and implementations are possible within the scope of the embodiments described herein. Although many possible combinations of features are shown in the drawings and discussed in the detailed description, many other combinations of the disclosed features are also possible. Unless specifically limited, any feature or element of any embodiment may be used in combination with, or may replace, any feature or element of any other embodiment.

[0056] This application includes and contemplates combinations of features and elements known to those skilled in the art. The embodiments, features, and elements disclosed in this application may also be combined with any conventional features or elements to form a unique inventive scheme as defined by the claims. Any feature or element of any embodiment may also be combined with features or elements from other inventive schemes to form another unique inventive scheme as defined by the claims. Therefore, it should be understood that any feature shown and / or discussed in this application may be implemented individually or in any suitable combination. Therefore, the embodiments are not limited except by the limitations imposed by the appended claims and their equivalents. Furthermore, various modifications and changes may be made within the scope of the appended claims.

[0057] Furthermore, in describing representative embodiments, the specification may have presented methods and / or processes as a specific sequence of steps. However, the method or process should not be limited to the specific order of steps described herein, to the extent that it does not depend on such a specific order. As will be understood by those skilled in the art, other sequences of steps are also possible. Therefore, the specific order of steps set forth in the specification should not be construed as a limitation of the claims. Moreover, the claims concerning the method and / or process should not be limited to the steps performed in the written order, and those skilled in the art will readily understand that these orders can be varied and still remain within the spirit and scope of the embodiments of this application.

[0058] This invention provides a seismic data processing method based on three-dimensional predictive deconvolution, such as... Figure 1 As shown, the method includes steps S100-S110, as detailed below:

[0059] Step S100. Based on the target seismic data of the study area, determine the operator of the three-dimensional predictive deconvolution using a pre-established three-dimensional predictive deconvolution model;

[0060] Step S110. Perform three-dimensional predictive deconvolution processing on the target seismic data according to the three-dimensional predictive deconvolution operator to obtain processed seismic data.

[0061] This embodiment focuses on a method for suppressing shallow-water multiples during marine seismic data processing, which theoretically can suppress seabed surface multiples. Specifically, it effectively suppresses shallow-water multiples, i.e., seabed reflection multiples in water depths of 50-200 meters. However, in deep water, seabed multiples no longer exhibit periodicity, and therefore cannot be effectively suppressed.

[0062] In one exemplary embodiment, the target seismic data is three-dimensional common-receiver point seismic gather data. The target seismic data can be processed by removing the common-receiver point gathers after the direct wave. If the common-receiver point gathers after removing the direct wave are used for convolution, the influence of the primary wave introduced by the direct wave and operator convolution can be avoided.

[0063] In one exemplary embodiment, the pre-established 3D predictive deconvolution model is as follows:

[0064]

[0065] In the above model, d b,m,j This represents the common-receiver gather data with shot line number b, shot point number m, time direction sample point number j, and containing multiples; p b,m,j This represents the common-receiver gather data of shot line number b, shot point number m, time direction sample point number j, and without multiples; f c,n,k d represents the operator for 3D predictive deconvolution; c represents the sample point number of the 3D predictive operator corresponding to b, n represents the sample point number of the 3D predictive operator corresponding to m, and k represents the sample point number of the 3D predictive operator corresponding to j; w represents the number of half-width points in the Crossline direction, u represents the number of half-width points in the Inline direction; g represents the step size for 3D predictive deconvolution calculation; d b-c,m-n,j-k This represents the common-detector seismic gather data with shot line number bc, shot point number mn, time direction sample point number jk, and containing multiple waves.

[0066] In one exemplary embodiment, the process of establishing a 3D predictive deconvolution model includes:

[0067] Step 1. Determine one-dimensional seismic data based on the reflection coefficient and the seismic wavelet; wherein the reflection coefficient is a white noise sequence and the seismic wavelet is a wavelet with minimum phase.

[0068] Step 2. Determine the one-dimensional predictive deconvolution model based on the statistical characteristics of the seismic data;

[0069] Step 3. Determine the three-dimensional predictive deconvolution model based on the one-dimensional predictive deconvolution model and the three-dimensional properties of the three-dimensional seismic data.

[0070] In one exemplary embodiment, determining one-dimensional seismic data based on reflection coefficients and seismic wavelets includes:

[0071] One-dimensional seismic data are calculated using the seismic convolution formula based on the reflection coefficient and seismic wavelet. The seismic convolution formula is as follows:

[0072] d(t) = b(t) * f(t)

[0073] In the above formula, d(t) represents the seismic data in the time domain, b(t) represents the seismic wavelet in the time domain, f(t) represents the reflection coefficient in the time domain, and t represents the time of the seismic data.

[0074] In one exemplary embodiment, determining a one-dimensional predictive deconvolution model based on the statistical characteristics of seismic data includes:

[0075] Step 1. Determine the relationship between earthquake data and time based on the statistical characteristics of the earthquake data;

[0076] Step 2. Establish a one-dimensional predictive deconvolution model based on the determined relationship between earthquake data and time;

[0077] In this step, based on the assumption that the seismic wavelet is of minimum phase and the reflection coefficient is a white noise sequence, the seismic data can be approximated as a stationary random process, that is, the statistical characteristics of the seismic data are independent of time.

[0078] The one-dimensional predictive deconvolution model is as follows:

[0079]

[0080] In the above one-dimensional predictive deconvolution model, d j This represents the seismic gather data of the common-detector point with sample number j in the time direction and containing multiple waves; p j This represents the seismic gather data of the common receiver point with sample number j in the time direction and without multiples; d j-k This represents the seismic gather data of common-detector points with time-direction sample number jk that contain multiples; f k This represents an operator for predicting deconvolution in one dimension. For example... Figure 2 The diagram shows a comparison between the one-dimensional and three-dimensional predictive deconvolution operators. Figure 2 Figure (a) shows a one-dimensional operator. Figure 2 Figure (b) shows the three-dimensional operator, where gap represents the prediction step size g, which is generally taken to be slightly smaller than the two-way vertical travel time of the water layer (in milliseconds); the prediction operator length... Prediction operator width n2op, dx is the detector point spacing; z is the water depth; theta is the maximum reflection angle; noise is the added white noise parameter, which is generally 1%.

[0081] In step S100, based on the target seismic data of the study area, the operator for the three-dimensional predictive deconvolution is determined using a pre-established three-dimensional predictive deconvolution model; wherein, the process of determining the three-dimensional predictive deconvolution operator is as follows:

[0082]

[0083]

[0084] In this embodiment, the least squares method is used to solve for the operator and construct the E(f) function, where the operator f is the independent variable of the function E. Different operators yield different values ​​of the E(f) function. The operator f that best satisfies the condition is the one whose E(f) function value is minimized. Since the E(f) function has only one minimum value, its derivative can be set to zero to obtain the corresponding operator f. That is, assuming this derivative is zero, the operator f of the 3D predictive deconvolution is calculated. c,n,k .

[0085] In step S110, the target seismic data is subjected to three-dimensional predictive deconvolution processing according to the three-dimensional predictive deconvolution operator to obtain processed seismic data. This process is as follows: after calculating the three-dimensional predictive deconvolution operator f... c,n,k Then, using the target seismic data d b,m,j Subtract the 3D prediction deconvolution operator f c,n,k and earthquake data d b,m,j The result of convolution can be used to obtain suppressed multiple waves p b,m,j Subsequent earthquake data.

[0086] The method provided in this application has the following technical effects:

[0087] 1. It can suppress shallow water multiples in ocean data. This predictive deconvolution method for removing multiples can fill the gaps in the suppression of multiples in the stacking process.

[0088] 2. This method can consider multiple directions of 3D seismic data, predict the 3D predictive deconvolution operator, and then process the seismic data to obtain seismic data that is more spatially coherent.

[0089] 3. Compared with one-dimensional predictive deconvolution seismic data processing methods, this method improves the resolution of seismic data and enhances the quality of seismic data processing and migration imaging.

[0090] This disclosure also provides a seismic data processing apparatus based on three-dimensional predictive deconvolution, such as... Figure 3 As shown, the device includes: an operator module 310 for determining three-dimensional predicted deconvolution and a seismic data processing module 320;

[0091] The operator module for determining the three-dimensional predicted deconvolution is used to determine the operator for the three-dimensional predicted deconvolution based on the target seismic data of the study area and a pre-established three-dimensional predicted deconvolution model.

[0092] The seismic data processing module is used to perform three-dimensional predictive deconvolution processing on the target seismic data according to the three-dimensional predictive deconvolution operator to obtain processed seismic data.

[0093] In one exemplary embodiment, the pre-established 3D predictive deconvolution model is as follows:

[0094]

[0095] In the above model, d b,m,j This represents the common-receiver gather data with shot line number b, shot point number m, time direction sample point number j, and containing multiples; p b,m,j This represents the common-receiver gather data of shot line number b, shot point number m, time direction sample point number j, and without multiples; f c,n,k d represents the operator for 3D predictive deconvolution; c represents the sample point number of the 3D predictive operator corresponding to b, n represents the sample point number of the 3D predictive operator corresponding to m, and k represents the sample point number of the 3D predictive operator corresponding to j; w represents the number of half-width points in the Crossline direction, u represents the number of half-width points in the Inline direction; g represents the step size for 3D predictive deconvolution calculation; d b-c,m-n,j-k This represents the common-detector seismic gather data with shot line number bc, shot point number mn, time direction sample point number jk, and containing multiple waves.

[0096] In one exemplary embodiment, the process of establishing a 3D predictive deconvolution model includes:

[0097] One-dimensional seismic data is determined based on the reflection coefficient and the seismic wavelet; wherein the reflection coefficient is a white noise sequence and the seismic wavelet is a wavelet with minimum phase.

[0098] Based on the statistical characteristics of earthquake data, a one-dimensional predictive deconvolution model is determined;

[0099] The three-dimensional predictive deconvolution model is determined based on the one-dimensional predictive deconvolution model and the three-dimensional properties of the three-dimensional seismic data.

[0100] In one exemplary embodiment, determining one-dimensional seismic data based on reflection coefficients and seismic wavelets includes:

[0101] One-dimensional seismic data are calculated using the seismic convolution formula based on the reflection coefficient and seismic wavelet. The seismic convolution formula is as follows:

[0102] d(t) = b(t) * f(t)

[0103] In the above formula, d(t) represents the seismic data in the time domain, b(t) represents the seismic wavelet in the time domain, f(t) represents the reflection coefficient in the time domain, and t represents the time of the seismic data.

[0104] In one exemplary embodiment, determining a one-dimensional predictive deconvolution model based on the statistical characteristics of seismic data includes:

[0105] Determine the relationship between earthquake data and time based on the statistical characteristics of earthquake data;

[0106] A one-dimensional predictive deconvolution model is established based on the determined relationship between earthquake data and time.

[0107] The one-dimensional predictive deconvolution model is as follows:

[0108]

[0109] In the above one-dimensional predictive deconvolution model, d j This represents the seismic gather data of the common-detector point with sample number j in the time direction and containing multiple waves; p j This represents the seismic gather data of the common receiver point with sample number j in the time direction and without multiples; d j-k This represents the seismic gather data of common-detector points with time-direction sample number jk that contain multiples; f k This represents an operator for predicting deconvolution in one dimension.

[0110] This disclosure also provides a computer program product with seismic data processing capabilities based on three-dimensional predictive deconvolution, including a computer program / instruction that, when executed by a processor, implements the seismic data processing method based on three-dimensional predictive deconvolution described in any of the above embodiments.

[0111] Example 1

[0112] This example demonstrates the process of building a 3D predictive deconvolution model as follows:

[0113] The first step is to determine the seismic data by assuming that the reflection coefficient sequence is a white noise sequence and the seismic wavelet is the minimum phase.

[0114] Deconvolution is performed under the assumption that the reflection coefficient sequence is a white noise sequence and the seismic wavelet is of minimum phase. If the influence of random noise is ignored, the seismic record x(t) satisfies the following convolution model:

[0115] d(t) = b(t) * f(t)

[0116] In the above formula, d(t) represents the seismic data in the time domain, b(t) represents the seismic wavelet in the time domain, f(f) represents the reflection coefficient in the time domain, and t represents the time of the seismic data.

[0117] The second step assumes that the reflection coefficient sequence is a white noise sequence and the seismic wavelet is of minimum phase. The determined seismic data can be approximated as a stationary random process, meaning that the statistical characteristics of the seismic data are independent of time. Under these assumptions, a one-dimensional predictive deconvolution model is established:

[0118]

[0119] In the above one-dimensional predictive deconvolution model, d j This represents the seismic gather data of the common-detector point with sample number j in the time direction and containing multiple waves; p j This represents the seismic gather data of the common receiver point with sample number j in the time direction and without multiples; d j-k This represents the seismic gather data of common-detector points with time-direction sample number jk that contain multiples; f k This represents an operator for predicting deconvolution in one dimension.

[0120] Third step: Based on the one-dimensional prediction deconvolution model, derive the following prediction formula:

[0121]

[0122] Based on the above prediction formula, given the earthquake data, the operator for one-dimensional prediction of deconvolution is unknown, so we need to find p. After analysis, we can first find the operator f for one-dimensional prediction of deconvolution:

[0123]

[0124]

[0125] Where i ≥ g;

[0126] R is the autocorrelation function of D: r i =r -i =∑ j d j d j-i .

[0127] Assuming this derivative is zero, we obtain the Toeplitz equation:

[0128]

[0129] Based on the Toeplitz equations described above, an operator for predicting one-dimensional deconvolution can be obtained.

[0130] Step 4: Extend the one-dimensional predictive deconvolution to three-dimensional predictive deconvolution, resulting in the following three-dimensional predictive deconvolution model:

[0131]

[0132] In the above model, d b,m,j This represents the common-receiver gather data with shot line number b, shot point number m, time direction sample point number j, and containing multiples; p b,m,j This represents the common-receiver gather data of shot line number b, shot point number m, time direction sample point number j, and without multiples; f c,n,k d represents the operator for 3D predictive deconvolution; c represents the sample point number of the 3D predictive operator corresponding to b, n represents the sample point number of the 3D predictive operator corresponding to m, and k represents the sample point number of the 3D predictive operator corresponding to j; w represents the number of half-width points in the Crossline direction, u represents the number of half-width points in the Inline direction; g represents the step size for 3D predictive deconvolution calculation; d b-c,m-n,j-k This represents the common-detector seismic gather data with shot line number bc, shot point number mn, time direction sample point number jk, and containing multiple waves.

[0133] Example 2

[0134] Seismic data processing based on 3D predictive deconvolution was performed using seismic data from a marine block, as follows:

[0135] Step 1. Acquire marine seismic exploration data and obtain the common receiver point seismic gather d. b,m,j ;

[0136] Step 2. Determine the operator f for 3D predictive deconvolution based on the pre-established 3D predictive deconvolution model. c,n,k ;

[0137] The three-dimensional predictive deconvolution model is as follows:

[0138]

[0139] In the above model, d b,m,j This represents the common-receiver gather data with shot line number b, shot point number m, time direction sample point number j, and containing multiples; p b,m,j This represents the common-receiver gather data of shot line number b, shot point number m, time direction sample point number j, and without multiples; f c,n,kd represents the operator for 3D predictive deconvolution; c represents the sample point number of the 3D predictive operator corresponding to b, n represents the sample point number of the 3D predictive operator corresponding to m, and k represents the sample point number of the 3D predictive operator corresponding to j; w represents the number of half-width points in the Crossline direction, u represents the number of half-width points in the Inline direction; g represents the step size for 3D predictive deconvolution calculation; d b-c,m-n,j-k This represents the common-detector seismic gather data with shot line number bc, shot point number mn, time direction sample point number jk, and containing multiple waves.

[0140] Step 3. Based on the calculated 3D predicted deconvolution operator f c,n,k and earthquake data d b,m,j Convolution is performed to obtain the seismic data p after suppressing multiple waves. b,m,j .

[0141] To verify the effectiveness of the seismic data processing method based on 3D predictive deconvolution, a 3D model data containing multiple waves was forward modeled using the wave equation method for predictive deconvolution test.

[0142] Figure 4 This is a comparison chart of the effects of 3D model prediction and deconvolution processing. Figure 4 The 3D model data parameters are as follows: longitudinal sampling point ns is 251, cmp direction nx is 134; cmp_line direction ny is 134. The data models used for the 3D prediction deconvolution method are all single-layer models, that is, they contain only one primary wave and the rest are multiple wave signals.

[0143] Figure 4 This is a three-dimensional seismic data model containing only one layer, where... Figure 4 (a) in the diagram is a schematic diagram before data processing. Figure 4 (b) in the diagram shows the effect after data processing. Since it is not easy to compare and observe the overall 3D seismic data, distant, near, and middle rows are extracted from the data for separate comparative analysis. The extracted locations are important for... Figure 4 As shown by the dashed line.

[0144] Figure 5 This is a schematic diagram of the far-middle-near arrangement before the predicted deconvolution process. Figure 5 (a) is a near arrangement. Figure 5 (b) is a middle arrangement. Figure 5 (c) is the far arrangement. Figure 6 This is a schematic diagram of the far-middle-near arrangement after the predicted deconvolution process; Figure 6 (a) is a near arrangement. Figure 6 (b) is a middle arrangement. Figure 6 (c) is the far arrangement. Through correspondence comparison, Figure 5 (a) Closest arrangement and Figure 6The middle (a) nearest arrangement is compared. Figure 5 (b) Arrangement and Figure 6 Compare the arrangement in (b) with the following. Figure 5 Middle (c) far arrangement and Figure 6 Comparing the near-to-far arrangement, it can be seen that multiple wave suppression is achieved in all near-to-far arrangement data, although there are some differences in multiple wave suppression among the near-to-far, mid-to-far, and far-to-near arrangement data. Because some noise is generated at the data boundaries when the prediction deconvolution operator is convolved with the seismic data, from... Figure 5 A comparison of the left arrow position in Figure (a) also reveals that this is a normal phenomenon of the method. Figure 5 Figure (a) and Figure 6 A comparison of the positions of the right arrows in Figure (a) reveals that the multiple waves are suppressed to some extent.

[0145] This example provides a method for determining a 3D predictive deconvolution operator based on target seismic data of a study area using a pre-established 3D predictive deconvolution model. The target seismic data is then processed using this operator to obtain processed seismic data. This application considers multiple directions of the 3D seismic data, predicts the 3D predictive deconvolution operator, and then processes the seismic data to obtain spatially more coherent seismic data. Furthermore, compared to one-dimensional predictive deconvolution seismic data processing methods, it improves the resolution of the seismic data and enhances the quality of seismic data processing and migration imaging.

[0146] It will be understood by those skilled in the art that all or some of the steps, systems, or apparatuses disclosed above, and their functional modules / units, can be implemented as software, firmware, hardware, or suitable combinations thereof. In hardware implementations, the division between functional modules / units mentioned above does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed collaboratively by several physical components. Some or all components may be implemented as software executed by a processor, such as a digital signal processor or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit (ASIC). Such software may be distributed on a computer-readable medium, which may include computer storage media (or non-transitory media) and communication media (or transient media). As is known to those skilled in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and can be accessed by a computer. Furthermore, it is well known to those skilled in the art that communication media typically contain computer-readable instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

Claims

1. A seismic data processing method based on three-dimensional predictive deconvolution, characterized in that, The method includes: Based on the target seismic data of the study area, the operators for three-dimensional predictive deconvolution are determined using a pre-established three-dimensional predictive deconvolution model; The target seismic data is subjected to three-dimensional predictive deconvolution processing based on the three-dimensional predictive deconvolution operator to obtain processed seismic data; The target seismic data is three-dimensional common-detector point seismic gather data; The process of establishing the three-dimensional predictive deconvolution model includes: One-dimensional seismic data is determined based on the reflection coefficient and the seismic wavelet; wherein the reflection coefficient is a white noise sequence and the seismic wavelet is a wavelet with minimum phase. Based on the statistical characteristics of earthquake data, a one-dimensional predictive deconvolution model is determined; The three-dimensional predictive deconvolution model is determined based on the one-dimensional predictive deconvolution model and the three-dimensional properties of the three-dimensional seismic data; The pre-established three-dimensional predictive deconvolution model is as follows: In the above model, d b,m,j This represents the common-receiver gather data with shot line number b, shot point number m, time direction sample point number j, and containing multiples; p b,m,j This represents the common-receiver gather data of shot line number b, shot point number m, time direction sample point number j, and without multiples; f c,n,k d represents the operator for 3D predictive deconvolution; c represents the sample point number of the 3D predictive operator corresponding to b, n represents the sample point number of the 3D predictive operator corresponding to m, and k represents the sample point number of the 3D predictive operator corresponding to j; w and u are the number of half-width points in the X and Y directions, respectively; g represents the step size for 3D predictive deconvolution calculation; d b-c,m-n,j-k This represents the common-detector seismic gather data with shot line number bc, shot point number mn, time direction sample point number jk, and containing multiple waves.

2. The seismic data processing method based on three-dimensional predictive deconvolution according to claim 1, characterized in that, The determination of one-dimensional seismic data based on reflection coefficient and seismic wavelet includes: One-dimensional seismic data are calculated using the seismic convolution formula based on the reflection coefficient and seismic wavelet. The seismic convolution formula is as follows: d(t) = b(t) * f(t) In the above formula, d(t) represents the seismic data in the time domain, b(t) represents the seismic wavelet in the time domain, f(t) represents the reflection coefficient in the time domain, and t represents the time of the seismic data.

3. The seismic data processing method based on three-dimensional predictive deconvolution according to claim 1, characterized in that, The step of determining a one-dimensional predictive deconvolution model based on the statistical characteristics of seismic data includes: Determine the relationship between earthquake data and time based on the statistical characteristics of earthquake data; A one-dimensional predictive deconvolution model is established based on the determined relationship between earthquake data and time. The one-dimensional predictive deconvolution model is as follows: In the above one-dimensional predictive deconvolution model, d j This represents the seismic gather data of the common-detector point with sample number j in the time direction and containing multiple waves; p j This represents the seismic gather data of the common receiver point with sample number j in the time direction and without multiples; d j-k This represents the seismic gather data of common-detector points with time-direction sample number jk that contain multiples; f k This represents an operator for predicting deconvolution in one dimension.

4. A seismic data processing device based on three-dimensional predictive deconvolution, characterized in that, The device includes: an operator module for determining three-dimensional predicted deconvolution and a seismic data processing module; The module for determining the operator of the three-dimensional predicted deconvolution is used to determine the operator of the three-dimensional predicted deconvolution based on the target seismic data of the study area and using a pre-established three-dimensional predicted deconvolution model. The seismic data processing module is used to perform three-dimensional predictive deconvolution processing on the target seismic data according to the three-dimensional predictive deconvolution operator to obtain processed seismic data. The target seismic data is three-dimensional common-detector point seismic gather data; The process of establishing the three-dimensional predictive deconvolution model includes: One-dimensional seismic data is determined based on the reflection coefficient and the seismic wavelet; wherein the reflection coefficient is a white noise sequence and the seismic wavelet is a wavelet with minimum phase. Based on the statistical characteristics of earthquake data, a one-dimensional predictive deconvolution model is determined; The three-dimensional predictive deconvolution model is determined based on the one-dimensional predictive deconvolution model and the three-dimensional properties of the three-dimensional seismic data; The pre-established three-dimensional predictive deconvolution model is as follows: In the above model, d b,m,j This represents the common-receiver gather data with shot line number b, shot point number m, time direction sample point number j, and containing multiples; p b,m,j This represents the common-receiver gather data of shot line number b, shot point number m, time direction sample point number j, and without multiples; f c,n,k d represents the operator for 3D predictive deconvolution; c represents the sample point number of the 3D predictive operator corresponding to b, n represents the sample point number of the 3D predictive operator corresponding to m, and k represents the sample point number of the 3D predictive operator corresponding to j; w represents the number of half-width points in the Crossline direction, u represents the number of half-width points in the Inline direction; g represents the step size for 3D predictive deconvolution calculation; d b-c,m-n,j-k This represents the common-detector seismic gather data with shot line number bc, shot point number mn, time direction sample point number jk, and containing multiple waves.

5. The seismic data processing device based on three-dimensional predictive deconvolution according to claim 4, characterized in that, The determination of one-dimensional seismic data based on reflection coefficient and seismic wavelet includes: One-dimensional seismic data are calculated using the seismic convolution formula based on the reflection coefficient and seismic wavelet. The seismic convolution formula is as follows: d(t) = b(t) * f(t) In the above formula, d(t) represents the seismic data in the time domain, b(t) represents the seismic wavelet in the time domain, f(t) represents the reflection coefficient in the time domain, and t represents the time of the seismic data.

6. The seismic data processing device based on three-dimensional predictive deconvolution according to claim 5, characterized in that, The step of determining a one-dimensional predictive deconvolution model based on the statistical characteristics of seismic data includes: Determine the relationship between earthquake data and time based on the statistical characteristics of earthquake data; A one-dimensional predictive deconvolution model is established based on the determined relationship between earthquake data and time. The one-dimensional predictive deconvolution model is as follows: In the above one-dimensional predictive deconvolution model, d j This represents the seismic gather data of the common-detector point with sample number j in the time direction and containing multiple waves; p j This represents the seismic gather data of the common receiver point with sample number j in the time direction and without multiples; d j-k This represents the seismic gather data of common-detector points with time-direction sample number jk that contain multiples; f k This represents an operator for predicting deconvolution in one dimension.

7. A computer program product with seismic data processing capabilities based on three-dimensional predictive deconvolution, comprising a computer program / instructions, characterized in that, When the computer program / instruction is executed by the processor, it implements the seismic data processing method based on three-dimensional predictive deconvolution as described in any one of claims 1 to 3.

Images