Seismic attribute dimension reduction method and device, storage medium and equipment

Abstract

The invention belongs to the technical field of geophysical oil-gas exploration, and provides a seismic attribute dimension reduction method and device, a storage medium and equipment. According to the method, representative seismic attributes are obtained by compressing a seismic attribute space from a high dimension to a low dimension. According to the attribute calculation method, a seismic attribute space is compressed from a high dimension to a low dimension, redundant information in original seismic attributes is removed, and more representative seismic attributes are obtained.

Description

Technical Field

[0001] The present invention belongs to the technical field of geophysical oil and gas exploration, and particularly relates to a seismic attribute dimensionality reduction method, device, storage medium, and equipment. Background Art

[0002] With the increase in oil and gas demand, the exploration scope has gradually shifted from conventional oil and gas reservoirs to complex unconventional oil and gas reservoirs, and the difficulty of reservoir prediction and oil and gas identification has also increased accordingly. The middle and shallow channel sandstone gas reservoirs in western Sichuan are mainly distributed in the Penglaizhen and Shaximiao formations. The reservoir has strong heterogeneity, with characteristics such as low porosity, low permeability, thin single-layer thickness, and rapid vertical and horizontal changes. The gas reservoir type is mainly tight sandstone lithologic gas reservoir. There are obvious differences in the natural gas enrichment degree among different channels, the physical properties or lithologies of the same channel, and different structural positions. The gas-water relationship is extremely complex, and gas-bearing detection in this area has become the top priority of research.

[0003] Many scholars have carried out attribute operations on this area using various attribute extraction algorithms. The extraction results of multiple attributes are ambiguous, which makes the accurate identification of reservoirs somewhat ambiguous. Comprehensive utilization of multiple attributes for attribute dimensionality reduction can reflect various information such as structure, gas-bearing property, and reservoir.

[0004] Therefore, there is an urgent need for a more accurate seismic attribute dimensionality reduction method. Summary of the Invention

[0005] The purpose of the present invention is to construct a seismic attribute dimensionality reduction method to obtain more representative seismic attributes and predict the gas reservoir boundary of the target area.

[0006] In a first aspect, the present invention proposes a seismic attribute dimensionality reduction method, including: compressing the seismic attribute space from high-dimensional to low-dimensional to obtain representative seismic attributes.

[0007] As a specific implementation manner of the present invention, the seismic attributes include geometric, frequency, and amplitude attributes.

[0008] As a specific implementation manner of the present invention, the compression from high-dimensional to low-dimensional includes using the Gaussian-weighted t-SNE algorithm.

[0009] As a specific implementation manner of the present invention, the formula of the Gaussian-weighted t-SNE algorithm is

[0010]

[0011]

[0012] In the formula, p ij is the joint probability between sample points in the high-dimensional space, and q ijis the joint probability between sample points in the low-dimensional space, and T is the joint probability p between the high-dimensional and low-dimensional spaces ij and q ij is the KL divergence between them, and y i 、y j is the point where the high-dimensional space data sample is mapped onto the low-dimensional space, and d kj is the Euclidean distance of the high-dimensional space sample, and δ is its variance.

[0013] According to the present invention, the t-SNE algorithm is developed from the SNE algorithm, and its main goal is to convert a multi-dimensional data set into a low-dimensional data set. The present invention improves the t-SNE algorithm to obtain an improved Gaussian-weighted t-SNE algorithm, which has more advantages in data visualization compared with other dimensionality reduction algorithms. The Gaussian-weighted t-SNE algorithm is not a linear dimensionality reduction technology, but a non-linear dimensionality reduction algorithm, that is, it can capture the complex manifold structure of high-dimensional data.

[0014] The process of the Gaussian-weighted t-SNE algorithm is as follows:

[0015] 1. Input the sample set of high-dimensional data and determine the total number of samples;

[0016] (1) Calculate the Euclidean distance between any two samples in the high-dimensional space and normalize the result;

[0017] (2) Group the normalized Euclidean distances according to different distribution conditions, and use the empirical formula proposed by Sturges to determine the number of groups of samples

[0018]

[0019] In the formula, N is the number of data, and the rounded value of G is the theoretical number of groups;

[0020] (3) The sample group distance refers to the difference between the largest m value and the smallest m value in the same group. The sample group distance c can be determined by the maximum value, minimum value and number of groups G of the Euclidean distance of the entire data set after normalization.

[0021]

[0022] According to the different m values after Euclidean distance normalization, the distances between high-dimensional space sample pairs can be divided into sample groups of various distances, and each distance is assigned a weight value. The Gaussian function is used to weight the features of each dimension. The Gaussian function is a commonly used probability density function, which has a centrosymmetric shape. By adjusting the parameters of the Gaussian function, we can control the degree of weighting.

[0023] 5. Calculate the weighted Euclidean distance according to the weight ω:

[0024]

[0025] 6. Calculate the similarity conditional probability between samples in high-dimensional space using the weighted Euclidean distance;

[0026]

[0027]

[0028] 7. Calculate the joint probabilities p ij and q ij between them, and set their KL divergence as the objective function T:

[0029]

[0030] 8. Use the gradient descent method to optimize the objective function T by derivation to obtain the optimal solution;

[0031]

[0032] 9. Obtain the final dimensionality reduction result.

[0033] As a specific implementation manner of the present invention, the method includes the following steps:

[0034] S1: Perform multi-attribute analysis through attribute optimization to optimize the favorable attributes for gas reservoir identification in the target target area;

[0035] S2: Comprehensively utilize multiple attributes for attribute dimensionality reduction;

[0036] S3: Adopt a multi-attribute dimensionality reduction technical means based on the Gaussian-weighted t-SNE algorithm to complete the boundary identification of the gas reservoir in the target target area.

[0037] As a specific implementation manner of the present invention, in the step S1, the attribute optimization method is determined according to the geological structure of the target target area in combination with the target horizon.

[0038] As a specific implementation manner of the present invention, in the step S2, the method of attribute dimensionality reduction includes low-variance filtering, high-correlation filtering, random forest, backward feature elimination, forward feature selection, missing value ratio, principal component analysis (PCA), and independent component analysis (ICA).

[0039] In a second aspect, the present invention provides a seismic attribute dimensionality reduction device, including:

[0040] An attribute collection unit, configured to collect multiple attributes of the gas reservoir in the target target area;

[0041] A dimensionality reduction unit, configured to compress the seismic attribute space from high-dimensional to low-dimensional;

[0042] An attribute extraction unit for extracting more representative seismic attributes;

[0043] A prediction unit for predicting the gas reservoir boundary of the target area.

[0044] In a third aspect, the present invention provides a computer-readable storage medium. The computer program stored in the computer-readable storage medium can be executed by one or more processors to implement the method as described in the first aspect.

[0045] In a fourth aspect, the present invention provides an electronic device, including a memory and one or more processors. A computer program is stored on the memory. When the computer program is executed by the one or more processors, the method as described in the first aspect is executed.

[0046] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0047] 1. The present invention provides an attribute calculation technology, forming a multi-attribute dimensionality reduction technology combination and application process for river channel description, improving the understanding of the development during the river channel recognition process.

[0048] 2. The prediction method of the present invention can integrate other effective seismic information into a two-dimensional data volume, significantly improving the interpretation result. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1a Amplitude attribute plan view of an embodiment of the present invention;

[0050] Figure 1b Amplitude attribute fusion diagram of an embodiment of the present invention;

[0051] Figure 2a Attribute plan view of a certain small layer of an embodiment of the present invention;

[0052] Figure 2b Attribute fusion diagram of a certain small layer of an embodiment of the present invention;

[0053] Figure 3a Attribute plan view of a certain small layer of an embodiment of the present invention;

[0054] Figure 3b Attribute fusion diagram of a certain small layer of an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0055] The present invention will be further described below in conjunction with specific embodiments, but it does not constitute any limitation to the present invention.

[0056] Embodiment

[0057] This embodiment provides a seismic attribute dimensionality reduction method. In combination with FIGS. 1-3, it is applied to a project in a certain work area. The algorithm application research is carried out using a tight sandstone in the southwest, and multi-attribute fusion is performed on 14 sub-layers of the Shaximiao Formation and used for channel characterization analysis:

[0058] S1: Conduct multi-attribute analysis through attribute optimization, and optimize the favorable attributes for gas reservoir identification in the target target area. The input attributes this time are mainly multi-scale data, amplitude data, and azimuthal data;

[0059] S2: Comprehensively utilize multiple attributes for attribute dimensionality reduction. The methods of attribute dimensionality reduction include low-variance filtering, high-correlation filtering, random forest, backward feature elimination, forward feature selection, missing value ratio, principal component analysis (PCA), and independent component analysis (ICA);

[0060] S3: Use the multi-attribute dimensionality reduction technical means based on the Gaussian-weighted t-SNE algorithm to complete the boundary identification of the gas reservoir in the target target area.

[0061] Figure 1a shows the plane map of the sub-layer amplitude attribute and Figure 1b shows the fused map of the sub-layer amplitude attribute, from Figure 1a and 1b it can be seen that compared with the conventional stacking method, the multi-attribute fusion technology based on Gaussian-weighted t-SNE can fully mine the effective information in the multi-attribute data, so that the characterized channel boundary is clearer, as shown in Figure 1b shown in the red circle in. Perform frequency division processing on the continuous data far from the stacked body, then extract sensitive attributes for Gaussian-weighted t-SNE fusion, generate the fused slice map of the sub-layer, and use it for channel characterization analysis. The results are as follows Figure 2a 、 2b and Figure 3a 、 3b shown. From the comparison of the plane maps of different layers, it can be seen that the fused map has more advantages in characterizing the channel than the conventional attribute plane map, and has a clearer and more intuitive understanding of the overall distribution state of the channel.

[0062] In one embodiment, a seismic attribute dimensionality reduction device is provided, including:

[0063] An attribute collection unit for collecting various attributes of the gas reservoir in the target target area;

[0064] A dimensionality reduction unit for compressing the seismic attribute space from high dimension to low dimension;

[0065] An attribute extraction unit for extracting more representative seismic attributes;

[0066] A prediction unit for predicting the gas reservoir boundary of the target target area.

[0067] In one embodiment, a computer-readable storage medium is provided. The computer program stored in the computer-readable storage medium can be executed by one or more processors to implement the method described in the above embodiment.

[0068] In one embodiment, an electronic device is provided, including a memory and one or more processors. A computer program is stored on the memory. The memory and the one or more processors are communicatively connected to each other. When the computer program is executed by the one or more processors, the method described in the above embodiment is executed.

[0069] In summary, the seismic attribute dimensionality reduction method of the present invention uses the multi-attribute dimensionality reduction technical means based on the Gaussian-weighted t-SNE algorithm to complete the boundary recognition of the gas reservoir in the target area. A multi-attribute dimensionality reduction technology combination and application process for various geological structures are formed, improving the understanding of the development during the channel recognition process.

[0070] In addition, it should be understood that the methods or systems disclosed in the embodiments provided in the present application can also be implemented in other ways. The method or system embodiments described above are merely illustrative. For example, the flowcharts and block diagrams in the drawings show the possible architectures, functions, and operations of the methods and apparatuses according to multiple embodiments of the present application. In this regard, each block in the flowchart or block diagram can represent a module, a computer program segment, or a part of a computer program. A module, a computer program segment, or a part of a computer program contains one or more computer programs for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than marked in the drawings, and in fact, they may be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and the combination of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system for performing the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer programs.

[0071] In this application, the terms "comprise", "include" or any other variant thereof are intended to cover non-exclusive inclusion, such that a process, method, article or apparatus comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or apparatus. Without further limitation, an element defined by the statement "comprising one..." does not exclude the presence of additional identical elements in the process, method, device or apparatus comprising the element; if terms such as "first", "second" etc. are described for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features or implicitly specifying the order of the indicated technical features. In addition, in the description of this application, unless otherwise stated, the terms "a plurality", "multiple" mean at least two.

[0072] Finally, it should be noted that in the description of this specification, the description with reference to terms such as "one embodiment", "some embodiments", "example", "an example" or "some examples" means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of this application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0073] Although the embodiments of this application have been shown and described above, it can be understood that the above embodiments are all exemplary, and the content described is only an implementation manner adopted for the convenience of understanding this application and is not used to limit this application. Any person skilled in the art within the technical field to which this application pertains, without departing from the spirit and scope disclosed in this application, can make any modifications and changes in the form of implementation and details, but the protection scope of this application shall still be subject to the scope defined by the appended claims.

Claims

1. A method for reducing the dimension of seismic attributes, characterized in that, it includes: Compressing the seismic attribute space from high dimension to low dimension to obtain representative seismic attributes.

2. The method for reducing the dimension of seismic attributes according to claim 1, characterized in that, the seismic attributes include geometric, frequency, and amplitude attributes.

3. The method for reducing the dimension of seismic attributes according to claim 2, characterized in that, the compression from high dimension to low dimension includes the t-SNE algorithm with Gaussian weighting.

4. The method for reducing the dimension of seismic attributes according to any one of claims 1-3, characterized in that, the t-SNE algorithm with Gaussian weighting is as follows: Where p ij is the joint probability between the sample points in the high-dimensional space, q ij is the joint probability between the sample points in the low-dimensional space, T is the KL divergence between the joint probabilities p ij and q ij , y i , y j are the points where the high-dimensional space data samples are mapped onto the low-dimensional space, d kj is the Euclidean distance of the high-dimensional space samples, and δ is its variance.

5. The method for reducing the dimension of seismic attributes according to any one of claims 1-4, characterized in that, this method includes the following steps: S1: Conduct multi-attribute analysis through attribute optimization to select favorable attributes for gas reservoir identification in the target target area; S2: Comprehensively utilize multiple attributes for attribute dimension reduction; S3: Adopt the multi-attribute dimension reduction technical means based on the t-SNE algorithm with Gaussian weighting to complete the boundary identification of the gas reservoir in the target target area.

6. The method for reducing the dimension of seismic attributes according to claim 5, characterized in that, in the step S1, the attribute optimization method is determined according to the geological structure of the target target area in combination with the target horizon.

7. The method for reducing the dimension of seismic attributes according to claim 6, characterized in that, in the step S2, the methods of attribute dimension reduction include low variance filtering, high correlation filtering, random forest, backward feature elimination, forward feature selection, missing value ratio, principal component analysis (PCA), and independent component analysis (ICA).

8. A device for reducing the dimension of seismic attributes, characterized in that, it includes: An attribute collection unit for collecting various attributes of the gas reservoir in the target target area; A dimension reduction unit for compressing the seismic attribute space from high dimension to low dimension; An attribute extraction unit for extracting more representative seismic attributes; A prediction unit for predicting the gas reservoir boundary of the target target area.

9. A computer-readable storage medium, characterized in that, the computer program stored in the computer-readable storage medium can be executed by one or more processors to implement the method according to any one of claims 1-7.

10. An electronic device, characterized in that, it includes a memory and one or more processors, and a computer program is stored on the memory. When the computer program is executed by the one or more processors, it executes the method according to any one of claims 1-7.

Images