A method for virtually creating three-dimensional data from two-dimensional seismic data

Abstract

The application discloses a method for converting two-dimensional seismic data into virtual three-dimensional data, which comprises the following steps: step one, combining the distribution of the target area, the basic parameters of the three-dimensional survey network are determined; step two, analyzing the original two-dimensional seismic data, the interval distribution range of the two-dimensional seismic data amplitude value domain is counted, and the consistency processing is performed to make the seismic data amplitude change constant in the standard interval, thereby avoiding the generation of the two-dimensional line closure error; step three, the two-dimensional seismic data is projected into the corresponding three-dimensional space based on the Natural-Neighbors algorithm, and the conversion from two-dimensional to three-dimensional is realized; step four, the converted three-dimensional data body is restored and output to the determined three-dimensional survey network, and the whole process of the two-dimensional seismic data virtual three-dimensional conversion is completed. The Natural-Neighbors algorithm is used to project the sparse two-dimensional line data into the three-dimensional space, so that the lateral distribution analysis and attribute extraction of the seismic data can be realized, and the underground structure form and fracture distribution can be determined.

Description

Technical Field

[0001] This invention relates to a method for creating virtual three-dimensional data from two-dimensional seismic data, belonging to the technical field of seismic exploration. Background Technology

[0002] Seismic exploration originated in the mid-19th century. With improvements in seismic data acquisition methods and equipment, and advancements in processing and interpretation techniques, seismic data has been used for reservoir description and detection. However, due to the influence of acquisitions from different periods and units, as well as various excitation and reception factors, two-dimensional seismic data exhibits significant differences in energy, frequency, and signal-to-noise ratio. Without eliminating these differences, the problem of seismic data stitching in this region cannot be effectively solved, making it difficult to perform detailed structural interpretation and reservoir inversion. With the development of three-dimensional seismic exploration technology, detailed three-dimensional interpretation effectively avoids the problem of two-dimensional seismic closure error correction, while providing a three-dimensional and in-depth understanding of the geometric morphology, spatial distribution, and interrelationships of subsurface geological structures, offering strong evidence for predicting the distribution of subsurface resources.

[0003] However, sometimes the exploration level of the target area, regional climate, and engineering operations limit the scope of exploration to two-dimensional seismic exploration. Therefore, if we want to carry out detailed lateral structural interpretation and reservoir prediction based on two-dimensional seismic data, we need to virtualize the two-dimensional seismic data into three dimensions.

[0004] Current mainstream interpretation software directly builds velocity models based on 3D seismic data, employing three main modeling methods: migration velocity analysis, reflection tomography, and waveform inversion. Migration velocity analysis, a product of pre-stack migration technology, is highly dependent on the modeling method for complex stratigraphic structures. Reflection tomography faces difficulties in obtaining mid-wavenumber velocity components, and as a generalized linear inversion algorithm, its objective function has multiple minima, significantly limiting its practical application. Waveform inversion typically exhibits a strongly nonlinear relationship between the objective function and velocity perturbations, making the inversion results highly dependent on the initial model selection. However, when dealing with 2D data, it's impossible to directly establish velocity models using the aforementioned three methods. Therefore, it's necessary to utilize sparse 2D seismic data to create a virtual 3D data volume, fully leveraging the advantages of 3D data volumes for lateral distribution analysis and eliminating closure errors. This is a prerequisite for conducting detailed 3D spatial interpretation of subsurface targets. Summary of the Invention

[0005] The purpose of this invention is to provide a method for simulating three-dimensional seismic data from two-dimensional seismic data. This method can project sparse two-dimensional seismic line data into three-dimensional space using the Natural-Neighbors algorithm, thereby enabling lateral distribution analysis and attribute extraction of seismic data.

[0006] To achieve the above objectives, the technical solution adopted by this invention is: a method for simulating three-dimensional data from two-dimensional seismic data, comprising: defining a three-dimensional space, performing consistency processing on the original two-dimensional seismic data, performing data conversion based on the Natural-Neighbors algorithm, and constructing three-dimensional seismic data. The specific steps are as follows:

[0007] Step 1: Based on the distribution of survey lines in the target area, determine the basic parameters of the three-dimensional survey network;

[0008] Step 2: Analyze the original 2D seismic data, statistically analyze the range of amplitude values ​​in the 2D seismic data, and use consistency processing to keep the amplitude variation of the seismic data constant within the standard range, thus avoiding the occurrence of 2D seismic line closure error.

[0009] Step 3: Based on the Natural-Neighbors algorithm, project the two-dimensional seismic data into the corresponding three-dimensional space to achieve the transformation from two-dimensional to three-dimensional.

[0010] Step 4: Restore the converted 3D data volume and output it to the planned 3D seismic network to complete the entire process of virtual 3D conversion of 2D seismic data.

[0011] Furthermore, in step one, the basic parameters include the range of line numbers, grid size, and four-point coordinates of the three-dimensional survey network.

[0012] Furthermore, in step two, the amplitude energy variation range of all two-dimensional seismic data is analyzed, and a standard amplitude value range is set. If an amplitude anomaly value exceeding the range appears, it is processed for consistency to ensure that the amplitude energy of all seismic data is within the same value range.

[0013] Furthermore, the expected value range is set between 0 and 350.

[0014] Furthermore, in step three, it is assumed that data point p i It is p j When p is the K-nearest neighbor, j Also p i The nearest neighbor is called p. i and p j They are natural neighbors and satisfy the following relation:

[0015] pj∈KNNk(pi)∩piKNNk(pj)→pi∈NaN(pj),pj∈NaN(p i )

[0016] Based on the concept of natural neighbors, a natural neighbor search algorithm is introduced to project 2D seismic data into a corresponding 3D space. During the transformation process, any point in the 3D space satisfies the above formula, p.i NN represents the number of natural neighbors of the i-th data point during the search process. K (p i ) represents data point p i Find the Kth nearest neighbor, iterate K+1 times, until NN K (p i ) = NN K+1 (p i The iteration ends, and K is the feature value of the natural neighbors of the current dataset.

[0017] Furthermore, in step four, the three-dimensional data is output to the three-dimensional survey network to give the three-dimensional data geological meaning. Based on the basic parameters set in step one, the three-dimensional data is exported to construct three-dimensional seismic data.

[0018] The beneficial effects of this invention are: the method of virtual three-dimensional data based on two-dimensional seismic data using the Natural-Neighbors algorithm is able to analyze seismic data using a three-dimensional fine interpretation method when only a small number of sparse two-dimensional seismic lines can be obtained in the work area, and to determine the underground structural morphology and fault distribution, which provides the possibility for reservoir prediction based on two-dimensional seismic data. Attached Figure Description

[0019] Figure 1 It involves determining the parameters for the 3D survey network, including the basic parameters for the 3D survey network of the application work area.

[0020] Figure 2 This is a schematic diagram of the three-dimensional survey network. In this application, 6 sparse two-dimensional survey lines are selected.

[0021] Figure 3 It is a virtual 3D seismic data time slice, a time slice of a 3D seismic data volume obtained through a 3D conversion process.

[0022] Figure 4 It is a 3D seismic data profile, a seismic profile at any location after the 3D seismic data volume is obtained through a 3D conversion process.

[0023] Figure 5 This is a flowchart of the method of the present invention. Detailed Implementation

[0024] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0025] An example provided by this invention: a method for simulating three-dimensional data from two-dimensional seismic data, comprising the following steps:

[0026] Step 1: Based on the distribution of two-dimensional survey lines in a certain work area, define a three-dimensional survey network and determine its basic parameters, including the grid size, line number range, and four-point coordinates of the three-dimensional survey network.

[0027] Step 2: Analyze the amplitude range of all two-dimensional seismic data lines and unify the amplitude range of all two-dimensional seismic data lines to the expected range through consistency processing.

[0028] Step 3: Combine the Natural-Neighbors algorithm to project the two-dimensional survey line into three-dimensional space, realizing the conversion of two-dimensional data into three-dimensional data.

[0029] K-Nearest Neighbors defines a local subset of data consisting of a target and its k nearest neighbors. Reverse Nearest Neighbors is a derivative of K-Nearest Neighbors, which seeks the set of neighbors that considers the target data as its nearest neighbor. Natural Neighbors assumes that data points in core locations should have more neighbors than data points in other non-core locations.

[0030] Therefore, we can assume that the data point p i It is p j When p is the K-nearest neighbor, j Also p i The nearest neighbor is called p. i and p j They are natural neighbors and satisfy the following relation:

[0031] pj∈KNNk(pi)∩piKNNk(pj)→pi∈NaN(pj),pj∈NaN(p i )

[0032] Natural neighbors reinforce the concept of mutuality between data, contain more information, and compared to K-nearest neighbors and reverse nearest neighbors, the search domain of natural neighbors does not require specifying other parameters and can be adaptively obtained during the search process.

[0033] Based on the concept of natural neighbors, a natural neighbor search algorithm is introduced to project 2D seismic data into a corresponding 3D space. During the transformation process, any point in the 3D space satisfies the above formula, p. i NN represents the number of natural neighbors of the i-th data point during the search process. K (p i ) represents data point p i Find the Kth nearest neighbor, iterate K+1 times, until NN K (p i ) = NN K+1 (p iThe iteration ends, and K is the feature value of the natural neighbors of the current dataset. Therefore, we can adaptively obtain the search range without any parameter constraints. The obtained feature values ​​are used as data points in three-dimensional space to realize the transformation from two-dimensional to three-dimensional.

[0034] Step 4: Project the obtained 3D seismic data onto the 3D seismic network proposed in Step 1 to obtain the corresponding 3D seismic data, thus completing the entire process of virtual 3D data from 2D seismic data.

[0035] The specific steps are as follows:

[0036] Selecting seismic work area A, which has 6 two-dimensional survey lines, this study analyzes the length, coordinate range, and extension range of all survey lines to propose basic parameters for a three-dimensional seismic network. The coordinates of four points are: A (63840, 4221089), B (316523, 4221089), C (316523, 4382024), and D (63840, 4382024). The line and trace numbers range from 0 to 1000, and the grid size is 250m * 160m. Figure 1 As shown.

[0037] Subsequently, the amplitude energy variation range of all 2D seismic data was analyzed, and a standard amplitude value range was set. If an amplitude anomaly value exceeding the range appeared, it was necessary to perform consistency processing to ensure that the amplitude energy of all seismic data was within the same value range.

[0038] from Figure 2 As can be seen, the amplitude energy ranges of the various measurement lines differ, presumably due to variations in acquisition time or method, leading to differences in the amplitude energy variation range. Therefore, consistency processing is required for these multiple measurement lines. The expected value range is set between 0 and 350. Therefore, for measurement lines A, B, C, D, and E, only the amplitude values ​​need to be increased by 200. For measurement lines F and G, the amplitude values ​​need to be multiplied by a coefficient of 0.01 before being increased by 200. The amplitude value range changes before and after consistency processing are shown in Tables 1 and 2.

[0039] Table 1. Statistical table of amplitude range before consistency processing

[0040] Attr Min Attr Max Line Name 127 -127 A 124 -124 B 126 -126 C 127 -127 D 127 -127 E 13266.91 -16219.49 F 13681.36 -12388.28 G

[0041] We can perform three-dimensional data simulation based on the Natural-Neighbors algorithm on the two-dimensional seismic data that has already undergone consistency processing. According to the definition of the Natural-Neighbors algorithm, we can transform two-dimensional data into three-dimensional data.

[0042] Table 2. Statistical table of amplitude range after consistency processing

[0043]

[0044]

[0045] Subsequently, by outputting the corresponding 3D data to the 3D survey network, the 3D data is given geological meaning. Based on the basic parameters set in step 1, the 3D data is exported to construct 3D seismic data, such as... Figure 3 The image shown is a slice planar view of 3D seismic data at a certain time. Figure 4 This is an arbitrary line seismic profile of 3D seismic data.

[0046] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the above embodiments do not limit the scope of protection of the present invention in any way, and all technical solutions obtained by equivalent substitution or other means fall within the scope of protection of the present invention. Parts not covered in this invention are the same as or can be implemented using existing technology.

Claims

1. A method for simulating three-dimensional data from two-dimensional seismic data, characterized in that, include: The specific steps for defining three-dimensional space, processing the consistency of raw two-dimensional seismic data, data transformation based on the Natural-Neighbors algorithm, and constructing three-dimensional seismic data are as follows: Step 1: Based on the distribution of survey lines in the target area, determine the basic parameters of the three-dimensional survey network; Step 2: Analyze the original 2D seismic data, statistically analyze the range of amplitude values ​​in the 2D seismic data, and use consistency processing to keep the amplitude variation of the seismic data constant within the standard range, thus avoiding the occurrence of 2D seismic line closure error. Step 3: Based on the Natural-Neighbors algorithm, project the two-dimensional seismic data into the corresponding three-dimensional space to achieve the transformation from two-dimensional to three-dimensional. Step 4: Restore the converted 3D data volume and output it to the planned 3D seismic network to complete the entire process of virtual 3D conversion of 2D seismic data.

2. The method for creating virtual three-dimensional data from two-dimensional seismic data according to claim 1, characterized in that, In step one, the basic parameters include the range of line numbers, grid size, and four-point coordinates of the three-dimensional survey network.

3. The method for creating virtual three-dimensional data from two-dimensional seismic data according to claim 1, characterized in that, In step two, the amplitude energy variation range of all two-dimensional seismic data is analyzed, and a standard amplitude value range is set. If an amplitude anomaly value that exceeds the range appears, it is processed for consistency to ensure that the amplitude energy of all seismic data is within the same value range.

4. The method for creating virtual three-dimensional data from two-dimensional seismic data according to claim 3, characterized in that, The expected value range is between 0 and 350.

5. The method for creating virtual three-dimensional data from two-dimensional seismic data according to claim 1, characterized in that, In step three, it is assumed that data point p i It is p j When p is the K-nearest neighbor, j Also p i The nearest neighbor is called p. i and p j They are natural neighbors and satisfy the following relation: p j ∈KNN k (p i )∩p i KNN k (p j )→p i ∈NaN(p j ),p j ∈NaN(p i ) Based on the concept of natural neighbors, a natural neighbor search algorithm is introduced to project 2D seismic data into a corresponding 3D space. During the transformation process, any point in the 3D space satisfies the above formula, p. i NN represents the number of natural neighbors of the i-th data point during the search process. K (p i ) represents data point p i Find the Kth nearest neighbor, iterate K+1 times, until NN K (p i ) = NN K+1 (p i The iteration ends, and K is the feature value of the natural neighbors of the current dataset.

6. The method for creating virtual three-dimensional data from two-dimensional seismic data according to claim 1, characterized in that, In step four, the three-dimensional data is output to the three-dimensional survey network to give it geological meaning. Based on the basic parameters set in step one, the three-dimensional data is exported to construct three-dimensional seismic data.

Images