Approximate real earth surface prestack depth migration reference surface construction method, storage medium and approximate real earth surface prestack depth migration reference surface construction system

Abstract

The invention provides an approximate real earth surface prestack depth migration reference surface construction method, and relates to the technical field of petroleum seismic exploration, and the method comprises the steps: constructing a small smooth earth surface reference surface, and eliminating the earth surface topographic change with the scale range within 10 meters through small-scale smoothing processing; elevation interpolation and slope calculation: carrying out B-spline interpolation on the smoothed surface elevation and calculating an elevation slope value; elevation slope smoothing and normalization are carried out, Robust local weighting linear regression fitting and smoothing processing are carried out on the slope value, and linear normalization is carried out; and performing weighted downward movement on the slope, and constructing a new offset reference surface by taking the smoothed and normalized slope value as a weight. According to the method, weighting processing is carried out on the down-shift amount of the reference surface by using the surface elevation slope, so that the change trend of the surface elevation is maintained to the greatest extent, the transformation of actual wave field travel time is reduced, and the pre-stack depth migration result is more accurate and reliable.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of oil seismic exploration technology, in particular to a near true surface pre-stack depth migration datum surface construction method, a storage medium and a system. BACKGROUND

[0002] Conventional pre-stack time migration is often difficult to accurately determine the structure form in complex areas, while pre-stack depth migration is an effective seismic data imaging method for determining complex underground geological structure form and depth, which can greatly reduce the exploration risk. In the pre-stack depth migration technology, the construction of the migration datum surface is very important, which will directly affect the subsequent imaging effect and accuracy.

[0003] At present, there are two main methods to construct the migration datum surface: one is to use the conventional RG floating surface as the datum surface, and the other is to use the actual ground surface after small-scale smoothing as the datum surface. Using the conventional RG floating surface as the datum surface in pre-stack depth migration has poor effect, because it will greatly modify the actual wave field travel time, resulting in structural distortion phenomenon in the migration result; when using the actual ground small smoothing surface to construct the near true surface pre-stack depth migration datum surface, the problem of "piercing the air" will occur in the complex surface topography area, that is, the datum surface will pierce the air and be above the actual ground. In order to solve this problem, the conventional method is to move the datum surface down by a certain distance. However, this method will cause the datum surface to move too much in the area where the ground surface changes gently, and then affect the accurate restoration of the wave field. SUMMARY

[0004] The purpose of the present application is to provide a near true surface pre-stack depth migration datum surface construction method, a storage medium and a system, which can construct a migration datum surface that can maintain the trend of the change of the ground elevation and avoid the problem of "piercing the air".

[0005] To achieve the above purpose, the present application realizes by the following technical scheme:

[0006] A near true surface pre-stack depth migration datum surface construction method, comprising the following steps,

[0007] S1: small smoothing ground datum surface construction, small-scale smoothing the true ground elevation;

[0008] S2: elevation interpolation and slope calculation, B-spline interpolation is performed on the small smoothing ground elevation, and the interpolated data is differentially calculated with the original data to obtain the elevation slope value;

[0009] S3: elevation slope smoothing and normalization, Robust local weighted linear regression fitting and linear normalization are performed on the elevation slope value to ensure the smoothness and stability of the slope value;

[0010] S4: slope weighted downshift, using the smoothed and normalized surface elevation slope value as weight, weighted downshift to build the datum surface.

[0011] Further: the specific operation of step S1 is to smooth the actual surface elevation according to the actual data processing designed horizontal resolution, and eliminate the surface topographic change with a scale range of 10-15 meters. By smoothing the surface topographic change with a scale range of 10-15 meters, the calculation amount can be reduced while the accuracy of subsequent calculation can be ensured.

[0012] Further: the horizontal resolution range is 12.5-25 meters.

[0013] Further: in step S2, the B-spline interpolation increases the horizontal grid offset on the basis of the surface elevation model processed in step S1 to form a new horizontal grid. The purpose of this is to ensure that the interpolation is all interpolation, and to avoid forming extrapolation to introduce new errors.

[0014] Further: in step S2, the formula for calculating the difference of the elevation slope value is:

[0015]

[0016] wherein:

[0017] h(x+Δx) is the small smooth surface datum elevation value at the interpolation grid point x+Δx obtained by interpolation;

[0018] h(x) is the small smooth surface datum elevation value at the horizontal grid point x;

[0019] Δx is the horizontal grid offset, in meters;

[0020] h'(x) is the surface elevation slope value corresponding to the horizontal distance x.

[0021] Further: the formula for calculating the elevation slope value at the last point of the new horizontal grid is modified as:

[0022]

[0023] h' n (x) is the elevation slope value at the last grid point, dimensionless;

[0024] h n (x n ) is the elevation value at the last grid point, in meters;

[0025] h n (x nwherein h(x) is the elevation value at the last grid point by interpolation, in meters.

[0026] Further, in step S4, the new offset reference surface is constructed using a weighted calculation formula:

[0027] h new (x)=h0(x)+k·Δh

[0028] wherein:

[0029] h0(x) is the small smooth ground surface reference elevation value at the horizontal grid point x;

[0030] k is the elevation slope value after smoothing and normalization;

[0031] Δh is the reference surface vertical downward displacement, in meters;

[0032] h new (x) is the new reference surface elevation value at the horizontal grid point x by slope weighting.

[0033] The application also provides a storage medium, wherein the storage medium stores program codes for realizing the above method.

[0034] The application also provides an approximate true ground surface pre-stack depth migration reference surface construction system, comprising:

[0035] a model acquisition module for small-scale smoothing of the true ground surface elevation to construct a small smooth ground surface;

[0036] an interpolation and slope calculation module for B-spline interpolation of the small-scale smooth ground surface and calculation of the slope value of the small-scale smooth ground surface;

[0037] a slope smoothing and normalization module for Robust local weighted linear regression fitting and linear normalization processing of the slope value;

[0038] a reference surface construction module for construction of a new offset reference surface according to the normalized slope value as a weight.

[0039] Compared with the prior art, the application has the following beneficial effects:

[0040] Firstly, the application uses the ground surface elevation slope to weight the reference surface downward displacement, maximally maintains the change trend of the ground surface elevation, reduces the modification to the actual wave field travel time, makes the result of the pre-stack depth migration more accurate, and reduces the occurrence of structural distortion phenomenon.

[0041] Secondly, the method of the present application is more close to the small smooth datum surface by using the slope weighted datum surface compared with the constant downward datum surface, but it is strictly below the actual height datum surface, avoiding the problem of "empty" of the small smooth datum surface.

[0042] Thirdly, the method of the present application has lower requirement on data quality, and is suitable for pre-stack depth migration processing in various complex structure areas, especially in the case of large surface relief and high exploration precision requirement, and can better meet the actual production demand. BRIEF DESCRIPTION OF DRAWINGS

[0043] Figure 1 The flowchart of the method for constructing the approximate true surface pre-stack depth migration datum surface based on slope weighting provided by one of the embodiments of the present application is shown in the figure.

[0044] Figure 2 The small scale smoothing diagram of the true surface height provided by one of the embodiments of the present application is shown in the figure.

[0045] Figure 3 The diagram for calculating the surface height interpolation and height slope provided by one of the embodiments of the present application is shown in the figure.

[0046] Figure 4 The effect diagram of the slope weighted datum surface before and after smoothing provided by one of the embodiments of the present application is shown in the figure.

[0047] Figure 5 The effect diagram of the slope weighted datum surface before and after normalization provided by one of the embodiments of the present application is shown in the figure.

[0048] Figure 6 The effect comparison diagram of the slope weighted datum surface, the actual height, the small smooth and the constant downward datum surface provided by one of the embodiments of the present application is shown in the figure. DETAILED DESCRIPTION

[0049] The technical solutions of the present application will be described clearly and completely in combination with the drawings. Obviously, the described embodiments are part of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all the other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the present application.

[0050] In the description of the present application, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer" and the like indicate the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the present application. In addition, the terms "first", "second", "third" are only for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0051] In view of the fact that it is difficult to meet the accuracy requirements or better solve the "empty space" problem when the RG floating surface or small smooth surface ground is used as the depth migration reference surface, the present application provides a slope-weighted approximate true ground pre-stack depth migration reference surface construction method, which will be described in detail below in conjunction with embodiments.

[0052] As shown in Figure 1 , a flowchart of a slope-weighted approximate true ground pre-stack depth migration reference surface construction method, comprising the following steps S1 to S4.

[0053] Step S1, small smooth ground reference surface construction, mainly smooths the true ground elevation to obtain a small smooth surface model.

[0054] In this step, according to the actual data processing design of the lateral resolution, such as the range between 12.5 meters and 25 meters, the actual ground elevation is processed by moving average smoothing according to the lateral scale, and the ground topographic changes within 15 meters are eliminated, as shown in Figure 2 , which can reduce the amount of calculation while ensuring the accuracy of subsequent calculations.

[0055] Step S2, elevation interpolation and slope calculation. The small smooth ground elevation data is interpolated by B-spline interpolation, and the interpolated data is differentiated with the original data to accurately calculate the ground elevation slope value.

[0056] In this step, the grid offset in the horizontal direction is set to 0.001m (Δx), and for a horizontal grid of length n, the first n-1 new horizontal grids are obtained by adding the offset to the original horizontal grid (x i , i = 1, 2,..., n-1), and for the nth new depth grid, the new horizontal grid (x i -Δx) is obtained by subtracting the offset from the original horizontal grid (x n ). n Figure 3 ​The purpose of this is to ensure that all interpolations are interpolations, avoiding the formation of extrapolations that introduce new errors. Then, the new ground elevation value h(x+Δx) at the new horizontal grid point is obtained by internal interpolation. The interpolation method uses B-spline interpolation. After obtaining the velocity value h(x+Δx) at the new horizontal grid point by B-spline interpolation, the elevation slope value is calculated by the following difference formula:

[0057]

[0058] wherein:

[0059] h(x+Δx) is the ground elevation value at the interpolation grid point x+Δx obtained by interpolation;

[0060] h(x) is the small smooth ground elevation value at the horizontal grid point x;

[0061] Δx is the horizontal grid point offset, which is set to 0.001 m in the embodiment of the present application;

[0062] h′(x) is the ground elevation slope value corresponding to the horizontal distance x.

[0063] As mentioned earlier, in order to ensure that all interpolation points are interpolations, the elevation slope value calculation formula at the last point is modified as follows:

[0064]

[0065] Therefore, the elevation slope value at each spatial coordinate point (x i ,y i ) can be obtained by B-spline interpolation and gradient calculation formula, as shown in Figure 3

[0066] Step S3, elevation slope smoothing and normalization, the smooth and stable elevation slope value is obtained by Robust local weighted linear regression fitting and linear normalization of the elevation slope value.

[0067] In this step, the elevation slope value calculated in step S2 is smoothed and normalized. The purpose of smoothing the elevation slope value is mainly to make the slope value change relatively smooth, avoiding unstable calculation results when using the slope value for subsequent weighted calculation. As shown in Figure 4 , directly using the unsmoothed slope value for weighting, the new datum plane obtained will change very sharply and basically lose the spatial variation characteristics of the old datum plane, while using the smoothed slope value for weighting, the new datum plane obtained basically retains the spatial variation characteristics of the old datum plane. The smoothed elevation slope value is further linearly normalized to limit the value range of the elevation slope value to a specified range. As shown in Figure 5 ​As shown, using the normalized slope value weighting can make the new datum closer to the small-scale smooth datum.

[0068] Step S4, slope weighting downward movement, mainly by the smooth and normalized slope value as a weight to the small-scale smooth datum weighting downward movement, to build a new datum.

[0069] In this step, we will step S3 in the smooth and normalized slope value as a weight, and set the datum vertical downward movement multiplied by the weighted datum vertical downward movement, and then applied to the current small-scale smooth datum, thus obtaining a new datum. As shown Figure 6 As shown, the slope weighting datum compared to constant downward movement datum, it is more close to the small-scale smooth datum, but at the same time it is strictly below the actual height datum, avoiding the problem of small-scale smooth datum "wear".

[0070] In this embodiment, first using the actual ground elevation small-scale smoothing, obtain small-scale smooth datum; then using the method of B-spline interpolation on the horizontal direction of the datum interpolation, so as to accurately calculate the elevation slope value; then the calculated elevation slope value is smoothed and normalized to ensure the smooth and stable of the calculated elevation slope value; finally, using the smooth and normalized elevation slope value as a weight value, weighted calculation of the new datum.

[0071] The new offset datum adopts the weighted calculation formula:

[0072] h new (x)=h0(x)+k·Δh

[0073] Wherein:

[0074] h0(x) is the small-scale smooth ground datum elevation value at the horizontal grid point x;

[0075] k is the smooth and normalized elevation slope value;

[0076] Δh is the datum vertical downward movement, unit: meter;

[0077] h new (x) is the new datum elevation value at the horizontal grid point x by slope weighting

[0078] The method can help to establish the approximate true ground prestack depth migration datum, which plays a guiding and restraining role for the subsequent prestack depth migration and other work.

[0079] In another embodiment, the application also provides a storage medium, which stores the program code for realizing the above method.

[0080] In another embodiment, the application also provides a near-truth pre-stack depth migration base surface construction system, comprising:

[0081] a model obtaining module, configured to perform small-scale smoothing on the real surface elevation, and construct a small-scale smoothed surface base surface;

[0082] an interpolation and slope calculation module, configured to perform B-spline interpolation on the small-scale smoothed surface base surface, and calculate slope values of the small-scale smoothed surface base surface;

[0083] a slope smoothing and normalization module, configured to perform Robust local weighted linear regression fitting on the slope values, and perform linear normalization processing on the slope values;

[0084] a base surface construction module, configured to construct a new migration base surface according to the normalized slope values as weights

[0085] The near-truth pre-stack depth migration base surface construction method based on slope weighting provided by the embodiment of the application can be specific hardware on a device or software or firmware installed on the device, etc. The device provided by the embodiment of the application has the same implementation principle and generated technical effects as the foregoing method embodiment. For brevity and conciseness, the part of the device embodiment not mentioned can be referred to the corresponding content in the foregoing method embodiment. Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the system, device and unit described above can be referred to the corresponding processes in the foregoing method embodiment, which will not be repeated here.

[0086] In the embodiments provided by the application, it should be understood that the disclosed device and method can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There can be another division manner in actual implementation. For example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections between the units can be indirect couplings or communication connections through some interfaces, devices or units, and can be electrical, mechanical or in other forms.

[0087] The units described as separate components can or can not be physically separate, and the components displayed as units can or can not be physical units, that is, they can be located in one place, or can be distributed on a plurality of network units. According to actual needs, some or all of the units can be selected to achieve the purpose of the embodiment scheme.

[0088] In addition, each function unit in the embodiments provided by the present application can be integrated in one processing unit, or each unit can be physically present individually, or two or more units can be integrated in one unit.

[0089] If the functions are realized in the form of software function units and sold or used as independent products, they can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application essentially or the parts that contribute to the prior art or parts of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various program code storage media.

[0090] Related term explanation:

[0091] Pre-stack depth migration imaging: Seismic data stacking refers to the process of stacking received seismic trace set signals to form a two-dimensional or three-dimensional data body. Before stacking, it is pre-stack, and after stacking, it is post-stack. The pre-stack data contains more seismic information, and the signal-to-noise ratio of the post-stack data is high. Migration imaging refers to the process of applying corresponding geophysical theory to back-propagate the multiple coverage seismic trace set data observed on the ground, eliminating the propagation effect of seismic waves, and obtaining the image of the underground medium model based on a certain mathematical and physical model (acoustic medium, elastic medium, etc.). Migration imaging can be divided into time migration and depth migration. Roughly speaking, time migration refers to a migration algorithm that does not consider ray bending, and depth migration refers to a migration algorithm that considers ray bending. Pre-stack depth migration imaging refers to the technology of first performing migration processing on seismic trace sets and then stacking.

[0092] Migration reference surface: The migration reference surface is a horizontal reference surface selected for migration imaging and seismic data interpretation results unified mapping. In the case of large surface relief and high exploration accuracy requirements, a floating reference surface is applied, which is a curved surface obtained after smoothing the terrain.

[0093] B-spline interpolation: B-spline interpolation is a commonly used interpolation method with high interpolation accuracy and good approximation properties. It is based on spline functions, and by giving a set of function values on a set of discrete points, a B-spline function is constructed to satisfy the interpolation conditions at the given discrete points.

[0094] The B-spline interpolation formula is as follows:

[0095]

[0096] wherein:

[0097] P(t) is a parametric equation of B-spline interpolation function;

[0098] t is an interpolation node;

[0099] P is control point data;

[0100] B i,k (t) is the k-th B-spline basis function of the i-th data.

[0101] In the above formula, B i,k (t) satisfies the following recursive formula:

[0102] When k>0,

[0103] When k=0,

[0104] The B-spline interpolation algorithm has many advantages. First, it has good locality and flexibility, which allows different types of interpolation effects to be achieved by adjusting the order of the basis function (k), suitable for various different data distributions and shapes, not only for regular control point distribution, but also for irregular control point distribution, and even for constructing curves or surfaces with complex topological structure. Second, the curve or surface generated by the B-spline interpolation algorithm is smooth, and the interpolation function is continuous, i.e. continuous within each interval. In the case of noisy data or incomplete data, the algorithm can still produce reasonable interpolation results.

[0105] Linear normalization: Normalization is a common data preprocessing method, the main principle is to scale the original data to the range of [0, 1] according to the proportion, so that the distribution of data is more uniform. In this invention, the use of normalization is mainly based on the following two advantages: (1) Eliminate the influence of dimension. Different characteristics of original data may have different dimensions and units, and some characteristics with larger dimensions may dominate, affecting the accuracy of the results. Normalization can map all data to the same scale, thereby eliminating such dimensional differences, making the comparison and operation between different characteristics more fair and meaningful; (2) Improve model stability. Normalization can reduce the influence of noise and outliers in the data on the model, thereby improving the stability and robustness of the model.

[0106] The linear normalization formula is as follows:

[0107]

[0108] wherein:

[0109] x is the original data;

[0110] x norm is the linear normalized data.

[0111] In this invention, we do not want to normalize the data to the range of [0, 1], but to the range of the specified interval [a, b], so the above formula is modified as follows:

[0112]

[0113] x norm_ab = a + k(x - x min )

[0114] Where:

[0115] a and b are the specified value range;

[0116] k is the correction coefficient;

[0117] x is the original data;

[0118] x norm_ab is the data normalized to the specified interval range.

[0119] Local weighted linear regression fitting: a linear regression algorithm, mainly used for data smoothing. This method gives lower weight to outliers in regression, and gives zero weight to data outside six times the mean absolute deviation, so as to remove outliers in data and smooth the effect. The formula of local weighted linear regression is as follows:

[0120]

[0121] Where:

[0122] f(θ) is the loss function;

[0123] θ is the prediction coefficient;

[0124] y true is the true value;

[0125] x is the sample feature value;

[0126] ω is the weight value;

[0127] Where the superscript T represents the transpose operation of the matrix, and the subscript i refers to the i-th data (true value, sample feature value or weight value).

[0128] Linear regression requires the loss function f(θ) to take the minimum value, that is, to make the derivative value of the function zero:

[0129]

[0130] Finally, θ can be obtained as:

[0131] θ=(x T ωxθ) -1 x T ωy true .

[0132] The above embodiments are only for illustrating the technical concept and features of the present invention, and are intended to enable those skilled in the art to understand the content of the present invention and implement it accordingly. They should not be construed as limiting the scope of protection of the present invention. All equivalent transformations or modifications made in accordance with the spirit and essence of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A method for building a datum for near true earth prestack depth migration, characterized in that: The method comprises the following steps: S1: constructing a small-scale smooth ground surface reference plane, and performing small-scale smoothing on the real ground elevation; S2: performing B-spline interpolation on the small-scale smooth ground surface elevation, and calculating the elevation slope value by differentiating the interpolated data from the original data; S3: performing Robust local weighted linear regression fitting and linear normalization on the elevation slope value to ensure that the slope value is smooth and stable; S4: performing weighted downward movement using the smooth and normalized ground elevation slope value as the weight to construct the reference plane.

2. The method of claim 1, wherein the method further comprises: determining a reference surface for the near-surface pre-stack depth migration. The specific operation of step S1 is to perform sliding average smoothing on the actual ground elevation according to the horizontal resolution designed according to the actual data processing, and to eliminate the ground topographic changes within a scale range of 15 meters.

3. The method for constructing a pre-stack depth migration reference surface for approximate true surface as described in claim 2, characterized in that: The horizontal resolution ranges from 12.5 meters to 25 meters.

4. The method of claim 1, wherein the method further comprises: In step S2, the B-spline interpolation is used to increase the horizontal grid offset on the basis of the ground elevation model processed in step S1 to form a new horizontal grid.

5. The method of claim 4, wherein the method further comprises: In step S2, the formula for calculating the elevation slope value by differentiation is: wherein: h(x+Δx) is the small-scale smooth ground reference plane elevation value at the interpolation grid point x+Δx obtained by interpolation; h(x) is the small-scale smooth ground reference plane elevation value at the horizontal grid point x; Δx is the horizontal grid offset, in meters; h′(x) is the ground elevation slope value corresponding to the horizontal distance x.

6. The method of claim 5, wherein the method further comprises: The formula for calculating the elevation slope value at the last grid point of the new horizontal grid is modified as: h′ n (x) is the value of the elevation slope at the last grid point, dimensionless; h n (x n ) is the elevation value at the last grid point in meters; h n (x n -Δx) is the elevation value at the last grid point obtained by interpolation, in meters.

7. The method of claim 1, wherein: In step S4, the weighted calculation formula is used to construct the new offset reference plane: h new (x) = h0(x) + k - Ah wherein: h0(x) is the small-scale smooth ground reference plane elevation value at the horizontal grid point x; k is the smooth and normalized elevation slope value; Δh is the vertical downward movement of the reference plane, in meters; h new (x) is the new datum elevation value at the horizontal grid point x obtained by the slope weighting.

8. A storage medium, characterized by The storage medium stores a program code for implementing the method according to any one of claims 1-7.

9. A near true earth pre-stack depth migration datum surface construction system, characterized by, The method according to any one of claims 1-7 comprises: a model acquisition module configured to perform small-scale smoothing on the real ground elevation to construct a small-scale smooth ground reference plane; an interpolation and slope calculation module configured to perform B-spline interpolation on the small-scale smooth ground reference plane and calculate the slope value of the small-scale smooth ground reference plane; a slope smoothing and normalization module configured to perform Robust local weighted linear regression fitting on the slope value and perform linear normalization processing; a reference plane construction module configured to construct a new offset reference plane according to the normalized slope value as the weight.

Images