Multi-horizon automatic pickup method based on sparse two-stage dynamic warping algorithm

By adopting a sparse two-stage dynamic regularization algorithm in hierarchy, using sequence similarity constraints search range and hypothesis test, the problem of low efficiency and accuracy of dynamic time regularization algorithm is solved, and accurate hierarchical recognition in complex geological environments is achieved and fault treatment improve the efficiency and accuracy of hierarchical tracking.

CN119001845AActive Publication Date: 2024-11-22CHINA NAT PETROLEUM CORP +2

Patent Information

Application Number
CN202310561497.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2023-05-18
Publication Date
2024-11-22
Estimated Expiration
2043-05-18

AI Technical Summary

Technical Problem

The existing dynamic time regularization algorithm has low efficiency and accuracy in hierarchical tracking, especially in complex geological areas, and there are fault problems, which affects the tracking effect.

Method used

The two-stage dynamic regularization algorithm based on sparse is adopted to constrain the search range through sequence similarity, reduce the search path, and improve accuracy by combining hypothesis testing. The similarity and backtracking path are calculated using the sparse matrix and distance formula to achieve multi-layer bit automatic pickup.

Benefits of technology

Improves the efficiency and accuracy of hierarchy tracking, and can accurately identify hierarchy in complex geological environments, save time, and improves tracking accuracy in the presence of faults.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119001845A_ABST
    Figure CN119001845A_ABST
Patent Text Reader

Abstract

The invention discloses a multi-horizon automatic pickup method based on a sparse two-stage dynamic warping algorithm. The multi-horizon automatic pickup method comprises the following steps in sequence: S1, acquiring the similarity and a backtracking path of two sequences; s2, clustering seismic traces; and S3, carrying out multi-layer automatic picking. The invention provides a multi-horizon automatic pickup method, horizon tracking work is carried out by utilizing a sparse two-stage dynamic warping algorithm, and on the basis of a traditional dynamic warping algorithm, constraint is added according to the similarity of data to reduce a search range and a second-order distance formula, so that the horizon tracking efficiency is improved. The problem that a classical dynamic time warping algorithm is low in efficiency and precision in horizon tracking application is solved. The method is suitable for multi-horizon automatic picking based on a sparse two-stage dynamic warping algorithm, accurate horizon tracking work can be obtained in a complex geological environment, and horizon picking efficiency and accuracy are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the field of geophysical exploration technology, and relates to layer tracking of two-dimensional seismic profiles and three-dimensional seismic bodies, and specifically is a multi-layer automatic picking method based on a sparse two-stage dynamic regularization algorithm. Background Art

[0002] Stratum tracing is the process of extracting the event axis from the reflection layer in the obtained seismic image. Stratum tracing is an important part of seismic interpretation and the basic work of seismic image interpretation. The earliest stratigraphic tracing work relied on manual calibration, but due to the large and complicated workload, it also required personnel with certain geological experience, which had great limitations. With the development of technology, automatic stratigraphic tracing technology is currently a commonly used method. Automatic tracing technology can not only identify stratigraphic layers on two-dimensional profiles, but also obtain three-dimensional stratigraphic tracing results, which is convenient for research.

[0003] At present, the automatic layer tracking technology includes methods based on coherence, edge detection operators, waveform similarity and deep learning methods. Anderson (1983) and others proposed to apply the waveform matching method to layer tracking; Li Peng (2010) proposed to apply five edge detection operator methods to layer tracking work, and conducted comparative analysis to draw conclusions; Li Xuefeng (2011) used a coherence-based method to analyze the coherent profile layer tracking interpretation and workflow. The above is the relevant research on layer tracking by domestic and foreign scholars using other related algorithms. At present, there are relatively few related studies on the use of dynamic warping algorithms. Song Jin (2017) proposed to introduce the dynamic time warping algorithm into the seismic layer interpretation work, and combined the algorithm with regularization to process complex seismic data; Aina JuellBugge (2019) proposed to use the classic DTW combined with the instantaneous phase to extract the target layer and realize layer tracking on three-dimensional seismic images. According to previous studies, there are relatively few applications of dynamic time warping algorithms for horizon tracking, and even fewer methods for identification in complex geological areas that take both time and accuracy into account. Summary of the invention

[0004] The purpose of the present invention is to provide a multi-layer automatic picking method based on a sparse two-stage dynamic warping algorithm. By utilizing the improved dynamic warping algorithm to perform layer tracking, on the basis of the traditional dynamic warping algorithm, constraints are added according to the similarity of the data itself to reduce the search range, so as to solve the problems of low efficiency and precision of the classic dynamic time warping algorithm in layer tracking applications, and to achieve the purpose of saving time by saving space; and taking into account the existence of faults in the actual data tracking process, the idea of ​​hypothesis testing is added to improve the accuracy of layer tracking, thereby obtaining better tracking effects.

[0005] In order to achieve the above object, the technical solution adopted by the present invention is:

[0006] A multi-layer position automatic picking method based on a sparse two-stage dynamic regularization algorithm. The search path is obtained in the final distance matrix, so the idea of ​​this improved algorithm is to first set a sparse matrix (this algorithm can consider that each position in the matrix is ​​empty), and then further set the position in the matrix according to the sequence similarity, that is, the example given below (the non-empty position in the matrix is ​​the search path range). The method includes the following steps in sequence:

[0007] S1. Obtain the similarity and backtracking path of two sequences

[0008] S11. Normalize the original sequence X and the original sequence Y

[0009] Normalize the original sequence X and the original sequence Y to obtain sequence X′ and sequence Y′;

[0010]

[0011] Where S′ is the normalized sequence, S i is the value of the i-th point in the sequence, max and min are the maximum and minimum values ​​in the sequence respectively;

[0012] S12. Set the window range and number

[0013] Set parameter θ, the range is 0-1, set the window range and number according to the set parameter θ, determine the number and range of windows and the points on the sequence in each window, the points of the sequence belonging to the same window are similar points, and then obtain the search range of the original sequence X and the original sequence Y;

[0014] S13. Set the backtracking path range based on similar points

[0015] S131. Set the initial sparse search path matrix M, the size of M is n*m, n is the length of the original sequence X, m is the length of the original sequence Y, and the value of each cell position in the matrix is ​​0, that is, a sparse cell;

[0016] An m×n sparse matrix is ​​a rectangular array of m rows and n columns of elements, and all elements in the matrix are zero;

[0017] Because the initial distance matrix is ​​set as a sparse matrix (the values ​​in the matrix are all 0), which is equivalent to the initial search path. In the later stage, according to the similarity and distance formula, the value at a certain coordinate (the position in the matrix) is calculated. This position with a value is the search path. If there is no value and it is still 0 before (the value in the sparse matrix is ​​0), it is not the final search path range;

[0018] S132. According to step S12, the search range of the original sequence X and the original sequence Y is obtained, and the positions corresponding to the similar points in each window in the sequence X′ and the sequence Y′ can correspond to the positions of the initial matrix M, that is, the initial search path range obtained according to the sequence similarity;

[0019] S133. Since the initial search path range involves the original sparse cells around the corresponding position in the matrix, in order to ensure the subsequent minimum path backtracking, it is necessary to set these sparse cells as the search path range;

[0020] S14. Calculate the distance matrix based on the second-order formula

[0021] The search ranges of sequence X and sequence Y obtained in step S13 are used to calculate the alignment distance according to the alignment distance formula, and then the cumulative distance is calculated according to the cumulative distance formula to find the optimal alignment path, that is, the backtracking path range;

[0022]

[0023] Where, d ij is the alignment distance, x i is the i-th value of sequence X, y j is the j-th value of sequence Y, i and j represent the i-th and j-th corresponding matrix elements of sequence X and sequence Y respectively;

[0024]

[0025] Where D ij is the cumulative distance, d ij is the alignment distance, i and j represent the i-th and j-th corresponding matrix units of sequence X and sequence Y respectively;

[0026] S15. The similarity and backtracking path of the two sequences (i.e., the corresponding coordinates of the two sequences) are obtained according to the distance matrix;

[0027] S2. Seismic trace clustering

[0028] Based on the idea of ​​statistical clustering, the similarity between seismic traces is obtained through a sparse two-stage dynamic regularization algorithm, and similar ones are classified into one category to obtain the clustering result of seismic traces. One of the seismic traces is matched with seismic traces of different categories.

[0029] S3. Multi-layer automatic picking

[0030] The matching results are interpolated to obtain the layer position, that is, the automatic picking of multiple layers is completed;

[0031] That is, the matching results and similarity clustering of seismic data are obtained through S1, and then the layer tracking is carried out. That is, the grid is first set according to the existing algorithm, and then there is a reference channel and multiple matching channels when the sequence in the grid is matched. After clustering, there is no need to match the reference channel with each matching channel. It is only necessary to match one channel belonging to the same category with the reference channel, and the rest use the results. Finally, the matching results are used to obtain the layer according to the interpolation method.

[0032] As a limitation, when performing the automatic picking of multiple layers, an improved dynamic time warping algorithm can also be used to track layers in seismic data with faults, and the hypothesis testing principle can be used to determine whether the layers on both sides of the fault belong to the same layer;

[0033] Stratum tracking requires setting up a grid according to the existing method after obtaining the three-dimensional seismic data, and using the improved algorithm of this patent to obtain the matching results of the two seismic trace data one by one in the sequence in the grid, and then obtaining the stratigraphic position based on the interpolation method studied by previous researchers based on the matching results (some coordinates are obtained, that is, some points on the corresponding seismic image).

[0034] As a further limitation, the specific steps of the hypothesis testing principle are: according to the waveform similarity principle, when judging whether the Ith layer and the Jth layer belong to the same layer, if they belong to the same layer, the distributions are the same or similar, that is, the means of the two distributions are equal.

[0035] As another limitation, in step S12, the method for setting the window range and number is: the difference between the upper limit of the window and the upper limit of the adjacent window is set to , and the difference between the lower and upper bounds within a window is set to ;

[0036] According to the method of setting the window range and number and the numerical size of the sequence itself, the number and range of windows and the points on the sequence within each window are determined.

[0037] As a third limitation, in step S2, the method for obtaining the clustering results of the seismic traces is: setting the clustering similarity size sim and the boundary range width, and clustering the seismic traces smaller than sim and width into one category, so as to obtain the clustering results of the seismic traces, in which the seismic traces in each category have a high degree of similarity.

[0038] When matching the reference and matching traces of layer tracing, the results of seismic traces belonging to the same category in all matching traces can be matched with one of the traces in the category and the reference trace using a two-stage dynamic regularization method to save time.

[0039] Due to the adoption of the above technical solution, the present invention has the following technical advances compared with the prior art:

[0040] ① The multi-layer automatic picking method based on the sparse two-stage dynamic warping algorithm provided by the present invention uses the improved dynamic warping algorithm to perform layer tracking. On the basis of the traditional dynamic warping algorithm, constraints are added according to the similarity of the data itself to reduce the search range, solving the problems of low efficiency and precision of the classic dynamic time warping algorithm in layer tracking applications;

[0041] ② The multi-layer position automatic picking method based on the sparse two-stage dynamic regularization algorithm provided by the present invention uses the similarity between the seismic trace data to set the search path range to improve the matching efficiency and the cumulative distance formula is a second-order formula to improve the matching accuracy;

[0042] ③ The multi-layer position automatic picking method based on the sparse two-stage dynamic regularization algorithm provided by the present invention achieves the purpose of saving time by saving space;

[0043] ④ The multi-layer automatic picking method based on the sparse two-stage dynamic regularization algorithm provided by the present invention takes into account the existence of faults in the actual data tracking process, adds the idea of ​​hypothesis testing, and improves the accuracy of layer tracking.

[0044] The present invention is suitable for multi-layer automatic picking based on a sparse two-stage dynamic regularization algorithm, can obtain accurate layer tracking work in a complex geological environment, and improves the efficiency and accuracy of layer picking. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is a flow chart of a multi-layer position automatic picking method based on a sparse two-stage dynamic regularization algorithm designed in the embodiment;

[0046] Figure 2 is a flow chart of a sparse two-stage dynamic regularization algorithm in an embodiment;

[0047] Figure 3 is a schematic diagram of a horizon tracing result of a seismic profile in an embodiment;

[0048] Figure 4 This is a three-dimensional layer tracking effect diagram of the sparse two-stage dynamic regularization algorithm in the embodiment;

[0049] Figure 5 It is a schematic diagram of the actual application of the layer tracking of seismic trace clustering in the embodiment. DETAILED DESCRIPTION

[0050] The present invention is further described in detail below by specific examples. It should be understood that the described examples are only used to explain the present invention, and are not intended to limit the present invention.

[0051] Embodiment A multi-layer bit automatic picking method based on sparse two-stage dynamic regularization algorithm

[0052] This embodiment is a multi-layer position automatic picking method based on a sparse two-stage dynamic regularization algorithm. This embodiment is carried out in a seismic simulation data set. The workflow of this embodiment is shown in the figure below: Figure 1 As shown, according to the flowchart, multi-layer bit automatic picking based on the sparse two-stage dynamic regularization algorithm can be realized. The method includes the following steps performed in sequence:

[0053] S1. Use the sequence similarity to set the search path range. The flow chart of the sparse two-stage dynamic regularization algorithm in this step is as follows: Figure 2 Shown

[0054] S11. Normalize the original sequence X and the original sequence Y

[0055] The original sequence X is [4,0,1,2,1], and the original sequence Y is [3,5,3,2,1];

[0056] Normalize the original sequence X and the original sequence Y to obtain sequence X′ and sequence Y′;

[0057]

[0058] Where S′ is the normalized sequence, S i is the value of the i-th point in the sequence, max and min are the maximum and minimum values ​​in the sequence respectively;

[0059] The sequence X′ is [1, 0, 0.25, 0.5, 0], and the sequence Y′ is [0.5, 1, 0.5, 0.25, 0];

[0060] S12. Set the window range and number

[0061] Set the parameter θ to 1, set the window range and number according to the set parameter θ, determine the number and range of windows and the points on the sequence in each window, and the points of the sequence belonging to the same window are similar points, and then obtain the search range of the original sequence X and the original sequence Y, as shown in Table 1;

[0062] The method for setting the window range and number is: the gap between the upper bound of the window and the upper bound of the adjacent window is set to , and the difference between the lower and upper bounds within a window is set to ;

[0063] According to the above requirements and the numerical size of the sequence itself, determine the number and range of windows and the points on the sequence within each window;

[0064]

[0065] The search path range coordinates obtained above are the coordinates of the distance matrix, which correspond to the positions of the initial matrix. Because the above may only obtain part of the positions of the m*n matrix, there is no corresponding position or the original sparse state (the value is 0, which can be understood as empty), and the non-empty position is the range of the subsequent search path. When performing subsequent path search, we can only look for coordinate positions that are not empty, so this is to set the search path range according to the sequence itself.

[0066] S13. Set the backtracking path range based on similar points

[0067] S131. Set the initial sparse search path matrix M, the size of M is n*m, n is the length of the original sequence X 5, m is the length of the original sequence Y 5, and the value of each cell position in the matrix is ​​0, that is, a sparse cell;

[0068] S132. According to step S12, the search range of the original sequence X and the original sequence Y is obtained, and the positions corresponding to the similar points in each window in the sequence X′ and the sequence Y′ can correspond to the positions of the initial matrix M, that is, the initial search path range obtained according to the sequence similarity, that is, [1,0][1,2][1,3][1,4][2,0][2,2][2,3][2,4][3,0][3,2][3,3][3,4][4,0][4,2][4,3][4,4][0,1][3,1];

[0069] S133. Since the initial search path range involves the original sparse cells around the corresponding position in the matrix, in order to ensure the subsequent minimum path backtracking, it is necessary to set these sparse cells as the search path range;

[0070] S14. Calculate the distance matrix based on the second-order formula

[0071] The search ranges of sequence X and sequence Y obtained in step S13 are used to calculate the alignment distance according to the alignment distance formula, and then the cumulative distance is calculated according to the cumulative distance formula to find the optimal alignment path, that is, the backtracking path range;

[0072]

[0073] Where, d ij is the alignment distance, x i is the i-th value of sequence X, yj is the j-th value of sequence Y, i and j represent the i-th and j-th corresponding matrix elements of sequence X and sequence Y respectively;

[0074]

[0075] Where D ij is the cumulative distance, d ij is the alignment distance, i and j represent the i-th and j-th corresponding matrix units of sequence X and sequence Y respectively;

[0076] S15. According to the distance matrix, the similarity of the two sequences is 5 and the backtracking path is (0,2)(1,3)(2,4)(3,4)(4,4), that is, the corresponding coordinates of the two sequences;

[0077] The result of tracing a layer of a seismic profile in this embodiment is as follows: Figure 3 shown by Figure 3 It can be seen that the improved algorithm can obtain accurate layer positions in seismic images. Figure 3 Represents a horizon tracing result of a seismic section;

[0078] S2. Use the sparse two-stage dynamic regularization algorithm to track the horizon in the seismic data with faults, and use the hypothesis testing principle to determine whether the horizons on both sides of the fault belong to the same horizon. The specific method is as follows: according to the waveform similarity principle, when determining whether the Ith and Jth layers belong to the same horizon, if they belong to the same horizon, the distributions are the same or similar, that is, the means of the two distributions are equal;

[0079] The distribution of the layers on the left and right sides of the fault can be constructed by some amplitude parameter values ​​of the layer waveform, that is, the statistical mean and variance can be calculated. According to the theory of mathematical statistics, the statistics of the two distributions can be used to derive the statistics of the new F distribution, and finally the conclusion of whether they belong to the same layer can be determined based on the significance and critical value.

[0080] The effect diagram of three-dimensional layer tracking by the sparse two-stage dynamic regularization algorithm in this embodiment is as follows: Figure 4 shown by Figure 4 It can be seen that the multi-layer position tracking method based on the sparse two-stage dynamic regularization algorithm can obtain continuous layer tracking results, and the layer continuity results in the fault part are also better;

[0081] S3. Seismic trace clustering

[0082] Based on the idea of ​​statistical clustering, the similarity between seismic traces is obtained through a sparse two-stage dynamic regularization algorithm, and similar ones are classified into one category to obtain the clustering result of seismic traces. One of the seismic traces is matched with seismic traces of different categories.

[0083] The schematic diagram of the actual application of layer tracking of seismic trace clustering in this embodiment is as follows: Figure 5 shown by Figure 5 It can be seen that after adding the clustering idea, the layer tracking results are still good, and the time is reduced while ensuring the accuracy of layer tracking to obtain the layer tracking results.

Claims

1. A multi-layer bit automatic picking method based on a sparse two-stage dynamic regularization algorithm, characterized in that: The method comprises the following steps performed in sequence: S1. Obtain the similarity and backtracking path of two sequences S11. Normalize the original sequence X and the original sequence Y Normalize the original sequence X and the original sequence Y to obtain sequence X′ and sequence Y′; Where S′ is the normalized sequence, S i is the value of the i-th point in the sequence, max and min are the maximum and minimum values ​​in the sequence respectively; S12. Set the window range and number Set parameter θ, the range is 0-1, set the window range and number according to the set parameter θ, determine the number and range of windows and the points on the sequence in each window, the points of the sequence belonging to the same window are similar points, and then obtain the search range of the original sequence X and the original sequence Y; S13. Set the backtracking path range based on similar points S131. Set the initial sparse search path matrix M, the size of M is n*m, n is the length of the original sequence X, m is the length of the original sequence Y, and the value of each cell position in the matrix is ​​0, that is, a sparse cell; S132. According to step S12, the search range of the original sequence X and the original sequence Y is obtained, and the positions corresponding to the similar points in each window in the sequence X′ and the sequence Y′ can correspond to the positions of the initial matrix M, that is, the initial search path range obtained according to the sequence similarity; S133. Since the initial search path range involves the original sparse cells around the corresponding position in the matrix, in order to ensure the subsequent minimum path backtracking, it is necessary to set these sparse cells as the search path range; S14. Calculate the distance matrix based on the second-order formula The search ranges of sequence X and sequence Y obtained in step S13 are used to calculate the alignment distance according to the alignment distance formula, and then the cumulative distance is calculated according to the cumulative distance formula to find the optimal alignment path, that is, the backtracking path range; Where, d ij is the alignment distance, x i is the i-th value of sequence X, y j is the j-th value of sequence Y, i and j represent the i-th and j-th corresponding matrix elements of sequence X and sequence Y respectively; Where D ij is the cumulative distance, d ij is the alignment distance, i and j represent the i-th and j-th corresponding matrix units of sequence X and sequence Y respectively; S15. The similarity and backtracking path of the two sequences are obtained according to the distance matrix; S2. Seismic trace clustering Based on the idea of ​​statistical clustering, the similarity between seismic traces is obtained through a sparse two-stage dynamic regularization algorithm, and similar ones are classified into one category to obtain the clustering result of seismic traces. One of the seismic traces is matched with seismic traces of different categories. S3. Multi-layer automatic picking The matching results are interpolated to obtain the layers, thus completing the automatic picking of multiple layers.

2. The multi-layer bit automatic picking method based on the sparse two-stage dynamic regularization algorithm according to claim 1 is characterized in that: When performing the above-mentioned multi-layer automatic picking, an improved dynamic time warping algorithm can also be used to track the layers in the seismic data with faults, and the hypothesis testing principle can be used to determine whether the layers on both sides of the fault belong to the same layer.

3. The multi-layer bit automatic picking method based on the sparse two-stage dynamic regularization algorithm according to claim 2 is characterized in that: The specific steps of the hypothesis testing principle are: according to the waveform similarity principle, when judging whether the Ith layer and the Jth layer belong to the same layer, if they belong to the same layer, the distributions are the same or similar, that is, the means of the two distributions are equal.

4. The multi-layer bit automatic picking method based on the sparse two-stage dynamic regularization algorithm according to claim 1 is characterized in that: In step S12, the method for setting the window range and number is: the difference between the upper limit of the window and the upper limit of the adjacent window is set to The difference between the lower and upper bounds within a window is set to According to the method of setting the window range and number and the numerical size of the sequence itself, the number and range of windows and the points on the sequence within each window are determined.

5. The multi-layer bit automatic picking method based on the sparse two-stage dynamic regularization algorithm according to any one of claims 1 to 4, characterized in that: In step S2, the method for obtaining the clustering result of the seismic traces is: setting the clustering similarity size sim and the boundary range width, and clustering the seismic traces smaller than sim and width into one category.

Citation Information

Patent Citations

  • Automatic horizon tracking method adopting dip angle propagation method

    CN104375175A

  • Three-dimensional earthquake image full-layer automatic tracking method based on coupling search

    CN105223609A

  • Complex region seismic reflection layer identification and tracking method based on machine learning

    CN110749929A

  • Seismic horizon automatic extraction device and method based on dynamic planning

    CN111158046A

  • Image processing-based three-dimensional seismic horizon and fault identification system

    WO2022160538A1

Cited By

  • Ground penetrating radar horizon tracking method and system based on dynamic time warping

    CN121679721A

  • Time interval rule experiment point location optimization method and system based on backtracking algorithm, electronic equipment and storage medium

    CN122197108A