Acquisition footprint pressing method and device based on principal component analysis and dimension rearrangement

Abstract

The invention relates to the technical field of seismic data noise suppression, and particularly discloses a principal component analysis and dimension rearrangement acquisition footprint suppression method and device, and the method comprises the steps: carrying out the multiple principal component analysis of observation data, and obtaining a multi-stage noise model; fusing the multi-stage noise models together to obtain a footprint collection model; and carrying out dimension rotation and filtering on the footprint acquisition model. According to the method, the seismic data principal component analysis theory and the data dimension rearrangement thought are utilized, depiction and suppression processing of the seismic acquisition footprint abnormal energy noise are achieved, and the method has the advantages of being efficient and robust.

Description

Technical Field

[0001] This invention relates to the field of seismic data noise suppression technology, specifically to a method and apparatus for suppressing acquisition footprints using principal component analysis and dimensional rearrangement. Background Technology

[0002] Seismic acquisition footprints are very common in conventional 3D seismic acquisition data for land and ocean. This is usually due to the design of the seismic acquisition observation system and the coupling of the deployment of seismic acquisition and receiving devices.

[0003] The presence of acquisition footprints poses a challenge to reservoir characterization and quantitative interpretation because they manifest as linear amplitude anomalies in the seismic acquisition and observation system, hindering the characterization of the true properties of seismic data.

[0004] Currently, most methods for weakening or suppressing collected footprints are based on the periodicity of the footprints and frequency wavenumber filtering or curve domain suppression, which have achieved certain results, but are insufficient in terms of amplitude preservation or operability.

[0005] Based on this technical background, this invention studies a method and apparatus for collecting footprints using principal component analysis and dimensional rearrangement. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a method and apparatus for suppressing acquisition footprints based on principal component analysis and dimensional rearrangement. This method utilizes the theory of principal component analysis of seismic data and the idea of ​​data dimensional rearrangement to characterize and suppress anomalous energy noise in seismic acquisition footprints, exhibiting high efficiency and robustness.

[0007] To achieve the above objectives, a first aspect of the present invention provides a method for suppressing acquisition footprints in principal component analysis and dimensional rearrangement, comprising:

[0008] A multi-level noise model was obtained by performing multiple principal component analyses on the observed data;

[0009] The multi-level noise models are fused together to obtain the footprint model;

[0010] The collected footprint model is rotated and filtered.

[0011] A second aspect of the present invention provides a footprint pressing device for principal component analysis and dimensional rearrangement, comprising:

[0012] The analysis module is used to perform multiple principal component analyses on the observed data to obtain a multi-level noise model;

[0013] The fusion module is used to fuse the multi-level noise models together to obtain the collected footprint model;

[0014] The rotation filtering module is used to rotate and filter the collected footprint model in dimensions.

[0015] A third aspect of the present invention provides an electronic device, the electronic device comprising:

[0016] Memory, which stores executable instructions;

[0017] A processor that executes the executable instructions in the memory to implement the acquisition footprint suppression method for principal component analysis and dimensional rearrangement as described in the first aspect.

[0018] A fourth aspect of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the acquisition footprint suppression method for principal component analysis and dimensional rearrangement as described in the first aspect.

[0019] The beneficial effects of this invention include:

[0020] (1) The acquisition footprint suppression method proposed in this invention utilizes the principal component analysis and dimension rearrangement theory of seismic data and the idea of ​​data dimension rearrangement to realize the characterization and suppression of abnormal energy noise of seismic acquisition footprints, and has the characteristics of high efficiency and robustness.

[0021] (2) The acquisition footprint suppression method proposed in this invention based on principal component analysis and dimension rearrangement is applicable to all land and marine seismic data acquisition footprint processing. It can also be extended to noise suppression processing under any amplitude strip anomaly conditions. It is highly efficient and has high application value.

[0022] Other features and advantages of the present invention will be described in detail in the following detailed description section. Attached Figure Description

[0023] The above and other objects, features and advantages of the present invention will become more apparent from the more detailed description of exemplary embodiments of the invention in conjunction with the accompanying drawings.

[0024] Figure 1 This is a flowchart illustrating the footprint suppression method for principal component analysis and dimensional rearrangement proposed in this invention.

[0025] Figure 2 This is a flowchart illustrating a specific implementation of the footprint suppression method for principal component analysis and dimensional rearrangement proposed in this invention.

[0026] Figure 3 This is a schematic diagram showing the comparison of slices before and after footprint suppression in a specific embodiment of the footprint suppression method for principal component analysis and dimensional rearrangement proposed in this invention.

[0027] Figure 4 This is a schematic diagram of a footprint model in a specific implementation of the footprint suppression method based on principal component analysis and dimensional rearrangement proposed in this invention.

[0028] Figure 5 The diagram illustrates the energy anomalies of the footprints collected using the principal component analysis and dimensional rearrangement method proposed in this invention, presented in slice form. Detailed Implementation

[0029] Preferred embodiments of the invention will now be described in more detail. While preferred embodiments of the invention are described below, it should be understood that the invention can be implemented in various forms and should not be limited to the embodiments set forth herein.

[0030] This invention provides a method for suppressing acquisition footprints in principal component analysis and dimensional rearrangement, such as... Figure 1 As shown, it includes:

[0031] A multi-level noise model was obtained by performing multiple principal component analyses on the observed data;

[0032] The multi-level noise models are fused together to obtain the footprint model;

[0033] The collected footprint model is rotated and filtered.

[0034] This invention utilizes principal component analysis theory of seismic data and the idea of ​​data dimension rearrangement to characterize and suppress abnormal energy noise in seismic acquisition footprints, exhibiting high efficiency and robustness.

[0035] According to the present invention, a multi-level noise model is obtained by performing multiple principal component analyses on the observation data, including:

[0036] The process of separating the effective signal and noise components from the observation data is described as an optimization problem;

[0037] By using singular value decomposition to solve the optimization problem, a multi-level noise model is obtained.

[0038] According to the present invention, the expression for the observation data is:

[0039] D = S + N;

[0040] Where S is the effective signal, N is the noise component, and D is the observation data.

[0041] According to the present invention, the expression for the optimization problem is:

[0042]

[0043] Where k is a unit vector that satisfies singular value decomposition.

[0044] According to the present invention, the formula used to solve the optimization problem through singular value decomposition is as follows:

[0045] D=U∑V * ;

[0046] Where U is an m*m left singular vector matrix, ∑() is a positive semi-definite m*n diagonal matrix, and V * V is the transpose of an n*n right singular vector matrix.

[0047] Preferably, the multi-level noise model is m1, m2, ..., m i ;

[0048] Where, m i This is the noise model obtained from the i-th principal component analysis;

[0049] The expression for the footprint collection model obtained by fusing the multi-level noise models together is as follows:

[0050] m=∑m i ;

[0051] Where m represents the footprint model.

[0052] According to the present invention, dimensional rotation and filtering of the collected footprint model includes:

[0053] The energy anomalies in the collected footprint models are displayed and filtered in slice format, such as... Figure 5 As shown.

[0054] This invention is applicable to the processing of all land and marine seismic data acquisition footprints, and can also be extended to noise suppression processing under any amplitude strip anomaly conditions. It is highly efficient and has high application value.

[0055] The present invention will be described in more detail below through embodiments.

[0056] Example 1:

[0057] like Figure 2 As shown, this embodiment proposes a method for suppressing acquisition footprints using principal component analysis and dimensional rearrangement. It combines iterative principal component analysis and frequency filtering techniques of dimensional rearrangement. The signal is preserved through frequency filtering of dimensional rearrangement and filtering of principal lines and tie lines. The application and number of iterations of principal component analysis depend on the noise level.

[0058] The specific steps of this method are as follows:

[0059] Step 1: Estimate the amplitude of the collected footprints using principal component analysis techniques in a cascaded or parallel manner;

[0060] Orthogonal transformation is used to convert the observations of a set of potentially correlated variables into the values ​​of a set of linearly uncorrelated variables called principal components. The first principal component has the largest possible variance, while the subsequent components have larger possible variances under the constraint that they are orthogonal (i.e. uncorrelated) to the preceding components. The number of principal components is less than or equal to the number of original variables.

[0061] Seismic data D = S + N, where S is the effective signal and N is the noise component. The process of separating S and N from the observation data D using principal component analysis can be described as an optimization problem:

[0062]

[0063] This optimization problem can be solved using singular value decomposition:

[0064] D=U∑V * Where U is an m*m left singular vector matrix, Σ() is a positive semi-definite m*n diagonal matrix, and V * V is the transpose of an n*n right singular vector matrix.

[0065] Thus, a multi-level noise model m1, m2, ..., m can be obtained. i , where m is the noise model obtained from each principal component analysis;

[0066] Step 2: Combine all the collected footprint amplitudes obtained in the previous step;

[0067] This step involves analyzing, filtering, and fusing all the footprint models obtained in the first step: m = Σm i

[0068] Step 3: Collect the footprint model, perform dimensional rotation, and filter:

[0069] To better protect the effective signal and perform filtering, the algorithm considers rotating the final noise model data, that is, displaying the energy anomalies of the collected footprints in a slice format, such as... Figure 5 As shown.

[0070] In this embodiment, Figure 3 To compare the footprint slices before and after compression using iterative acquisition based on principal component analysis and data rearrangement filtering. Figure 4 The image shows a footprint model suppressed using this technology. It can be seen that the method in this embodiment can effectively characterize and suppress abnormal energy noise in seismic footprints.

[0071] Example 2:

[0072] This embodiment provides a method for suppressing acquisition footprints in principal component analysis and dimensional rearrangement, such as... Figure 1As shown, it includes:

[0073] A multi-level noise model was obtained by performing multiple principal component analyses on the observed data;

[0074] The multi-level noise models are fused together to obtain the footprint model;

[0075] The collected footprint model is dimensionally rotated and filtered.

[0076] In this embodiment, the multi-level noise model obtained by performing multiple principal component analyses on the observed data includes:

[0077] The process of separating the effective signal and noise components from the observation data is described as an optimization problem;

[0078] By solving the optimization problem using singular value decomposition, a multi-level noise model is obtained.

[0079] In this embodiment, the expression for the observation data is:

[0080] D = S + N;

[0081] Where S is the effective signal, N is the noise component, and D is the observation data;

[0082] In this embodiment, the expression for the optimization problem is:

[0083]

[0084] Where k is a unit vector that satisfies singular value decomposition;

[0085] In this embodiment, the formula used to solve the optimization problem through singular value decomposition is as follows:

[0086] D=UΣV * ;

[0087] Where U is an m*m left singular vector matrix, Σ() is a positive semi-definite m*n diagonal matrix, and V * V is the transpose of an n*n right singular vector matrix.

[0088] In this embodiment, the multi-level noise model is m1, m2, ..., m i ;

[0089] Where, m i This is the noise model obtained from the i-th principal component analysis;

[0090] The expression for the footprint collection model obtained by fusing the multi-level noise models together is as follows:

[0091] m=Σm i ;

[0092] Where m represents the footprint model being collected;

[0093] In this embodiment, dimensional rotation and filtering of the collected footprint model includes:

[0094] The energy anomalies in the collected footprint models are displayed and filtered in slice format, such as... Figure 5 As shown;

[0095] This invention is applicable to the processing of all land and marine seismic data acquisition footprints, and can also be extended to noise suppression processing under any amplitude strip anomaly conditions. It is highly efficient and has high application value.

[0096] Example 3:

[0097] This embodiment provides a footprint pressing device for principal component analysis and dimensional rearrangement, including:

[0098] The analysis module is used to perform multiple principal component analyses on the observed data to obtain a multi-level noise model;

[0099] The fusion module is used to fuse multi-level noise models together to obtain the collected footprint model;

[0100] The rotation filtering module is used to rotate and filter the collected footprint model in dimensions.

[0101] In this embodiment, the multi-level noise model obtained by performing multiple principal component analyses on the observed data includes:

[0102] The process of separating the effective signal and noise components from the observation data is described as an optimization problem;

[0103] By solving the optimization problem using singular value decomposition, a multi-level noise model is obtained.

[0104] In this embodiment, the expression for the observation data is:

[0105] D = S + N;

[0106] Where S is the effective signal, N is the noise component, and D is the observation data;

[0107] In this embodiment, the expression for the optimization problem is:

[0108]

[0109] Where k is a unit vector that satisfies singular value decomposition;

[0110] In this embodiment, the formula used to solve the optimization problem through singular value decomposition is as follows:

[0111] D=UΣV * ;

[0112] Where U is an m*m left singular vector matrix, Σ() is a positive semi-definite m*n diagonal matrix, and V * V is the transpose of an n*n right singular vector matrix.

[0113] In this embodiment, the multi-level noise model is m1, m2, ..., m i ;

[0114] Where, m i This is the noise model obtained from the i-th principal component analysis;

[0115] The expression for the footprint collection model obtained by fusing the multi-level noise models together is as follows:

[0116] m=∑m i ;

[0117] Where m represents the footprint model being collected;

[0118] In this embodiment, dimensional rotation and filtering of the collected footprint model includes:

[0119] The energy anomalies in the collected footprint models are displayed and filtered in slice format, such as... Figure 5 As shown;

[0120] This invention is applicable to the processing of all land and marine seismic data acquisition footprints, and can also be extended to noise suppression processing under any amplitude strip anomaly conditions. It is highly efficient and has high application value.

[0121] Example 4:

[0122] This invention provides an electronic device including a memory and a processor, comprising:

[0123] Memory, which stores executable instructions;

[0124] The processor executes executable instructions in memory to implement the acquisition footprint suppression method for principal component analysis and dimension rearrangement.

[0125] This memory is used to store non-transitory computer-readable instructions. Specifically, the memory may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may, for example, include random access memory (RAM) and / or cache memory. The non-volatile memory may, for example, include read-only memory (ROM), hard disk, flash memory, etc.

[0126] The processor may be a central processing unit (CPU) or other form of processing unit with data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device to perform desired functions. In one embodiment of the invention, the processor is used to execute computer-readable instructions stored in the memory.

[0127] Those skilled in the art should understand that, in order to solve the technical problem of how to achieve a good user experience, this embodiment may also include well-known structures such as communication buses and interfaces, and these well-known structures should also be included within the protection scope of this invention.

[0128] For a detailed description of this embodiment, please refer to the corresponding descriptions in the foregoing embodiments, which will not be repeated here.

[0129] Example 5:

[0130] This invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements a method for suppressing acquisition footprints for principal component analysis and dimension rearrangement.

[0131] A computer-readable storage medium according to embodiments of the present invention stores non-transitory computer-readable instructions. When these non-transitory computer-readable instructions are executed by a processor, all or part of the steps of the methods described in the foregoing embodiments of the present invention are performed.

[0132] The aforementioned computer-readable storage media include, but are not limited to: optical storage media (e.g., CD-ROM and DVD), magneto-optical storage media (e.g., MO), magnetic storage media (e.g., magnetic tape or portable hard drive), media with built-in rewritable non-volatile memory (e.g., memory card), and media with built-in ROM (e.g., ROM cartridge).

[0133] The acquisition footprint suppression method proposed in the embodiments of the present invention utilizes the principal component analysis theory of seismic data and the idea of ​​data dimension rearrangement to characterize and suppress the abnormal energy noise of seismic acquisition footprints, and has the characteristics of high efficiency and robustness.

[0134] The various embodiments of the present invention have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments.

Claims

1. A method of acquisition footprint suppression by principal component analysis and dimensional rearrangement, characterized by, The method comprises the following steps: performing multiple principal component analysis on observation data to obtain a multi-level noise model; fusing the multi-level noise model to obtain a collection footprint model; performing dimension rotation and filtering on the collection footprint model.

2. The method of claim 1, wherein, The step of performing multiple principal component analysis on observation data to obtain a multi-level noise model comprises the following steps: describing a process of separating effective signals and noise components from the observation data as an optimization problem; solving the optimization problem by singular value decomposition to obtain the multi-level noise model.

3. The method of claim 2, wherein, The expression of the observation data is: D=S+N; wherein S is an effective signal, N is a noise component, and D is observation data.

4. The method of claim 3, wherein, The expression of the optimization problem is: wherein k is a unit vector satisfying singular value decomposition.

5. The method of claim 4, wherein, The formula used for solving the optimization problem by singular value decomposition is: D = U∑V * ; Wherein, U is an m*m left singular vector matrix, ∑() is a semi-positive definite m*n diagonal matrix, V * is the transpose of an n*n right singular vector matrix V.

6. The method of claim 5, wherein, The multi-stage noise model is m1, m2,..., m i ; wherein m i is the noise model obtained from the i-th principal component analysis; The expression of fusing the multi-level noise model to obtain a collection footprint model is: m = ∑m i ; wherein m is a collection footprint model.

7. The method of claim 1, wherein, The step of performing dimension rotation and filtering on the collection footprint model comprises the following steps: displaying and filtering energy anomalies of the collection footprint model in a slice form.

8. A principal component analysis and dimensionally rearranged acquisition footprint suppression apparatus characterized by, The method comprises the following steps: an analysis module configured to perform multiple principal component analysis on observation data to obtain a multi-level noise model; a fusion module configured to fuse the multi-level noise model to obtain a collection footprint model; a rotation and filtering module configured to perform dimension rotation and filtering on the collection footprint model.

9. An electronic device, comprising: The electronic device comprises: a memory storing executable instructions; a processor configured to execute the executable instructions in the memory to implement the principal component analysis and dimension rearrangement collection footprint suppression method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a computer program, which is executed by a processor to implement the principal component analysis and dimension rearrangement collection footprint suppression method according to any one of claims 1-7.

Images