A method and apparatus for seismic facies classification of seismic data

Abstract

The application provides a seismic facies classification method and device for seismic data, and the method comprises the following steps: determining the similarity between each data and other data in a seismic data set in a target work area according to the kernel function of each data, so as to determine the similarity between all data in the seismic data set; dividing the seismic data set in the target work area into K cluster data according to the similarity between all data, so as to generate a division result; wherein K is equal to the number of seismic facies types in the target work area; and calibrating the division result according to the petrophysical parameters and / or sedimentary facies of the target work area, so as to determine the seismic facies classification result of the seismic data set. According to the transverse variability of the seismic signal in a certain target layer, the mean shift algorithm is used to classify the seismic trace shape, the classification result forms discrete'seismic facies', and the petrophysical parameters or the sedimentary facies plane distribution rule is described by using the seismic data.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of oil and gas exploration, in particular to the technical field of seismic data processing for oil and gas exploration, and specifically relates to a seismic facies classification method and device for seismic data. BACKGROUND

[0002] In the exploration and development of underground sedimentary minerals such as oil and coal, sedimentary facies research is of great significance. However, since the target layer is deeply buried underground, the research methods and means used are quite different from those used in outcrop area sedimentary facies research.

[0003] In the prior art, only through rock data can the sedimentary facies marks of the target be directly observed in underground facies analysis, and drilling coring is generally not continuous, and the full-well coring rate of a single exploration well is usually only a few percent to tens of percent, which causes great difficulty in sedimentary facies research. Although well facies analysis using electric logging data can make continuous sedimentary facies interpretation for the whole well, it has strong multi-solution,

[0004] More importantly, even if the single-well facies analysis data are sufficient, the information obtained by using the traditional research method is still only a part of the information, and important information such as stratigraphic stacking pattern and sedimentary body shape is not utilized. Further, even if the interpretation is completely correct, it is still only a "one-sided view". In order to further grasp the planar distribution characteristics of sedimentary facies, a large number of sufficiently dense drill holes are required, which is difficult to meet at the exploration stage. Therefore, in addition to the above two methods, it is also urgently needed to obtain more information from other data to improve the accuracy of sedimentary facies interpretation. Therefore, a new means and method for better grasping the planar variation characteristics of sedimentary facies with a small number of drill holes is urgently needed.

[0005] Seismic facies analysis is just to meet the above urgent needs. Seismic facies is the appearance of reflected waves on the seismic reflection time section. Seismic facies analysis is the interpretation and inference of sedimentary facies according to the characteristics of seismic facies. In oil exploration and the exploration of some coalfields and salt mines, seismic exploration data are essential and important basic data. These data can be obtained at the early stage of exploration, and generally cover the entire oil and gas field basin, and have very rich stratigraphic, structural and sedimentary facies information, so they are very valuable basic data in underground geological analysis. Seismic facies analysis, as an important part of seismic stratigraphy, was born around 1977 and quickly spread around the world. In the past ten years, it has been continuously developed and improved in extensive practice, and has become an indispensable sharp weapon in underground facies analysis.

[0006] The method of seismic facies analysis is to identify the unique seismic reflection wave group characteristics and its shape combination in each sequence, and to assign certain geological implications, and then to carry out the interpretation of sedimentary facies, which is called seismic facies analysis.

[0007] Seismic waveform is the basic property of seismic data, which contains all qualitative and quantitative information such as reflection pattern, phase, frequency and amplitude, and is the overall characteristics of seismic information. The dynamic changes of the waveform contain rich internal information, which can truly reflect the characteristics of the underground structure. The waveform classification method is the most commonly used method of seismic facies analysis. Through the classification of seismic signal waveform, the division of seismic facies can be realized.

[0008] There are two methods for seismic facies analysis and identification. The first method is to observe the seismic reflection characteristics by artificial observation, and to compare with the established standard seismic facies characteristics to determine which seismic facies it belongs to. This method is generally applied to local seismic data interpretation and analysis, and the interpretation and identification accuracy is relatively low. The other method is to use seismic data processing technology, computer technology and certain mathematical methods to analyze and calculate the seismic data body or seismic attribute data, and to extract the seismic facies which can reflect the change of sedimentary facies. This is an efficient, advanced and quantitative seismic facies identification method.

[0009] The traditional waveform classification methods include K-means and self-organizing neural network method. They all use unsupervised learning method, and the traditional K-means algorithm is a hard classification method, that is, it strictly limits the data on a certain data point to belong to the first class, and has nothing to do with other classes. In fact, this data or sample may be similar to other classes, but not the most similar. This means that the data or sample may belong to multiple classes, but the membership degree is different for different classes. Therefore, the traditional waveform classification method has some technical defects. SUMMARY

[0010] The present application belongs to the technical field of seismic data processing, and one object of the present application is to provide a more intelligent automatic seismic facies classification method for seismic data. The method compares the actual seismic data in a layer by channel, and carefully describes the lateral variation of seismic signals, so as to obtain the planar distribution rule of seismic anomaly body.

[0011] Another object of the present application is to provide a seismic facies classification device for seismic data. Still another object of the present application is to provide an electronic device comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above-mentioned seismic facies classification method for seismic data when executing the computer program. Still another object of the present application is to provide a readable medium having a computer program stored thereon, and the computer program implements the steps of the above-mentioned seismic facies classification method for seismic data when executed by a processor.

[0012] To solve the technical problems in the background art, the present application provides the following technical solutions.

[0013] In a first aspect, the present application provides a seismic facies classification method for seismic data, comprising:

[0014] determining the similarity between each data in the seismic data set in the target work area and other data according to the kernel function of the data, to determine the similarity between all data in the seismic data set;

[0015] dividing the seismic data set in the target work area into K cluster data according to the similarity between all data, to generate a division result; wherein K is equal to the number of seismic facies types in the target work area;

[0016] calibrating the division result according to the petrophysical parameters and / or sedimentary facies of the target work area, to determine the seismic facies classification result of the seismic data set.

[0017] In an embodiment of the present application, the seismic facies classification method for seismic data further comprises:

[0018] generating the kernel function of each data according to the position kernel function and data kernel function of each data in the data space.

[0019] In an embodiment of the present application, the determination of the similarity between each data in the seismic data set in the target work area and other data according to the kernel function of the data comprises:

[0020] performing the following iteration operation:

[0021] calculating the weighted average value of the kernel functions of all data points in the position kernel function in the data space;

[0022] calculating the offset of all data points in the position kernel function to the weighted average value;

[0023] moving the position of the position kernel function in the data space according to the weighted average value, until the offset is less than a first preset value;

[0024] wherein the similarity between all data in the position kernel function is the highest.

[0025] In an embodiment of the present application, the division of the seismic data set in the target work area into K cluster data according to the similarity between all data, to generate a division result, comprises:

[0026] determining the probability of each data being divided into a corresponding cluster data in the K cluster data according to the similarity between the data;

[0027] partitioning the seismic data set in the target work area into K cluster data according to the probability to generate the partitioning result.

[0028] In an embodiment of the present application, the partitioning the seismic data set in the target work area into K cluster data according to the similarity between all data to generate the partitioning result further comprises:

[0029] if the distance between the current cluster data center and another cluster data center is less than the second preset distance, the current cluster data is divided into the another cluster data.

[0030] In an embodiment of the present application, the partitioning the seismic data set in the target work area into K cluster data according to the similarity between all data to generate the partitioning result further comprises:

[0031] if the distance between the current cluster data center and another cluster data center is not less than the second preset distance, a cluster data category is added.

[0032] In an embodiment of the present application, the calibration of the partitioning result according to the rock physical parameters and / or sedimentary facies of the target work area to determine the seismic facies classification result of the seismic data set comprises:

[0033] generating the seismic facies of the work area according to the classification result;

[0034] calibrating the seismic facies according to the rock physical parameters and / or sedimentary facies to determine the final seismic facies classification result of the seismic data set.

[0035] In a second aspect, the present application provides a seismic facies classification device for seismic data, which comprises:

[0036] a data similarity determination module for determining the similarity between each data in the seismic data set in the target work area and other data according to the kernel function of the data to determine the similarity between all data in the seismic data set;

[0037] a partitioning result generation module for partitioning the seismic data set in the target work area into K cluster data according to the similarity between all data to generate the partitioning result; wherein the K is equal to the number of seismic facies categories in the target work area;

[0038] a seismic facies classification result determination module for calibrating the partitioning result according to the rock physical parameters and / or sedimentary facies of the target work area to determine the seismic facies classification result of the seismic data set.

[0039] In an embodiment of the present application, the seismic facies classification device for seismic data further comprises:

[0040] The kernel function generating module is configured to generate a kernel function of each data according to a position kernel function of each data in a data space and a data kernel function.

[0041] In an embodiment of the present application, the data similarity determining module comprises:

[0042] The iteration operation unit is configured to perform the following iteration operation:

[0043] The weighted average value calculating unit is configured to calculate a weighted average value of the kernel functions of all data points in the position kernel function in the data space.

[0044] The offset calculating unit is configured to calculate an offset of all data points in the position kernel function to the weighted average value.

[0045] The position kernel function moving unit is configured to move the position of the position kernel function in the data space according to the weighted average value until the offset is less than a first preset value.

[0046] In an embodiment of the present application, the similarity between all data in the position kernel function is the highest.

[0047] In an embodiment of the present application, the division result generating module comprises:

[0048] The probability determining unit is configured to determine a probability of each data being divided into a corresponding cluster data in the K cluster data according to the similarity between the data.

[0049] The division result generating unit is configured to divide the seismic data set in the target work area into the K cluster data according to the probability to generate the division result.

[0050] In an embodiment of the present application, the division result generating module further comprises:

[0051] The data dividing unit is configured to divide the current cluster data into another cluster data if a distance between the current cluster data center and the cluster data center of the another cluster data is less than a second preset distance.

[0052] In an embodiment of the present application, the division result generating module further comprises:

[0053] The cluster category increasing unit is configured to increase a cluster data category if a distance between the current cluster data center and the cluster data center of another cluster data is not less than the second preset distance.

[0054] In an embodiment of the present application, the seismic facies classification result determining module comprises:

[0055] The seismic facies generating unit is configured to generate a seismic facies of the work area according to the classification result.

[0056] The seismic facies classification result determination unit is configured to determine a final seismic facies classification result of the seismic data set according to the rock physical parameters and / or the sedimentary facies of the seismic facies.

[0057] In a third aspect, the present application provides a computer program product, comprising computer programs / instructions, which, when executed by a processor, implement the steps of the seismic facies classification method of seismic data.

[0058] In a fourth aspect, the present application provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the seismic facies classification method of seismic data when executing the program.

[0059] In a fifth aspect, the present application provides a computer readable storage medium, which stores a computer program, wherein the computer program, when executed by a processor, implements the steps of the seismic facies classification method of seismic data.

[0060] As can be seen from the above description, the embodiments of the present application provide a seismic facies classification method and device of seismic data, and the corresponding seismic facies classification method of seismic data comprises the following steps: first, determining the similarity between each data and other data in the seismic data set in the target work area according to the kernel function of each data, so as to determine the similarity between all data in the seismic data set; then, dividing the seismic data set in the target work area into K cluster data according to the similarity between all data, so as to generate a division result; wherein K is equal to the number of seismic facies types in the target work area; finally, calibrating the division result according to the rock physical parameters and / or the sedimentary facies of the target work area, so as to determine the seismic facies classification result of the seismic data set.

[0061] The corresponding seismic facies classification device of seismic data comprises: a data similarity determination module, configured to determine the similarity between each data and other data in the seismic data set in the target work area according to the kernel function of each data, so as to determine the similarity between all data in the seismic data set; a division result generation module, configured to divide the seismic data set in the target work area into K cluster data according to the similarity between all data, so as to generate a division result; wherein K is equal to the number of seismic facies types in the target work area; and a seismic facies classification result determination module, configured to calibrate the division result according to the rock physical parameters and / or the sedimentary facies of the target work area, so as to determine the seismic facies classification result of the seismic data set.

[0062] The method and device for seismic facies classification of seismic data provided by the embodiment of the present application classify seismic trace shapes by using mean shift algorithm according to the lateral variability of seismic signals in a certain target layer section, and the classification result forms discrete "seismic facies". Then, the seismic facies are compared and calibrated with petrophysical parameters or sedimentary facies, so as to realize the description of the planar distribution rule of petrophysical parameters or sedimentary facies by using seismic data. BRIEF DESCRIPTION OF DRAWINGS

[0063] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0064] Figure 1 It is a flowchart of the method for seismic facies classification of seismic data in the embodiment of the present application;

[0065] Figure 2 It is another flowchart of the method for seismic facies classification of seismic data in the embodiment of the present application;

[0066] Figure 3 It is a flowchart of step 100 of the method for seismic facies classification of seismic data in the embodiment of the present application;

[0067] Figure 4 It is a flowchart of step 200 of the method for seismic facies classification of seismic data in the embodiment of the present application;

[0068] Figure 5 It is another flowchart of step 200 of the method for seismic facies classification of seismic data in the embodiment of the present application;

[0069] Figure 6 It is a third flowchart of step 200 of the method for seismic facies classification of seismic data in the embodiment of the present application;

[0070] Figure 7 It is a flowchart of step 300 of the method for seismic facies classification of seismic data in the embodiment of the present application;

[0071] Figure 8 It is a block diagram of the device for seismic facies classification of seismic data in the specific embodiment of the present application;

[0072] Figure 9 It is a structural schematic diagram of the electronic device in the embodiment of the present application. DETAILED DESCRIPTION

[0073] In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some but not all of the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0074] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system or a computer program product. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0075] It should be noted that the terms "comprising" and "having" and any variations thereof in the specification and claims of the present application and the above drawings are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device containing a series of steps or units does not have to be limited to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products or devices. The embodiments in the present application and the features in the embodiments can be combined with each other without conflict. The present application will be described in detail below with reference to the drawings and in combination with the embodiments.

[0076] The acquisition, storage, use, processing and the like of data in the technical solutions of the present application comply with the relevant provisions of laws and regulations.

[0077] Embodiment one:

[0078] The embodiments of the present application provide a specific implementation of a seismic facies classification method for seismic data, referring to Figure 1 , specifically including the following contents:

[0079] Step 100: determining the similarity between each data and other data in the target work area according to the kernel function of each data in the seismic data set in the target work area, so as to determine the similarity between all data in the seismic data set;

[0080] Step 200: dividing the seismic data set in the target work area into K cluster data according to the similarity between all data, so as to generate a division result; wherein K is equal to the number of seismic facies types in the target work area;

[0081] Step 300: Calibrating the division result according to the petrophysical parameters and / or sedimentary facies of the target work area to determine the seismic facies classification result of the seismic data set.

[0082] From the above description, the embodiment of the application provides a seismic facies classification method of seismic data, which comprises the following steps: firstly, determining the similarity between each data in the seismic data set in the target work area and other data according to the kernel function of the data, so as to determine the similarity between all data in the seismic data set; then, dividing the seismic data set in the target work area into K cluster data according to the similarity between all data, so as to generate a division result; wherein K is equal to the number of seismic facies types in the target work area; finally, calibrating the division result according to the petrophysical parameters and / or sedimentary facies of the target work area, so as to determine the seismic facies classification result of the seismic data set.

[0083] The seismic facies classification method of seismic data provided by the embodiment of the application classifies the seismic trace shape by using the mean shift algorithm according to the lateral variability of the seismic signal in a certain target layer section, and the classification result forms discrete "seismic facies". Then, the seismic facies is calibrated according to the petrophysical parameters or the sedimentary facies, so as to realize the description of the planar distribution rule of the petrophysical parameters or the sedimentary facies by using the seismic data.

[0084] Embodiment two:

[0085] The kernel function is a kind of technology for non-linear problems, which is used in machine learning algorithms such as support vector machine (SVM). In machine learning, a linear model can only process linearly separable problems, and the role of the kernel function is to map the data in the original feature space to a higher-dimensional feature space, so that the originally linearly inseparable problem becomes linearly separable. By performing linear division in the high-dimensional feature space, non-linear division in the low-dimensional feature space is indirectly realized. The definition of the kernel function is as follows:

[0086]

[0087] Wherein, K(x, y) represents the kernel function value between samples x and y, and φ(x) and φ(y) represent the mapping of samples x and y in the high-dimensional feature space, respectively.

[0088] Preferably, the kernel function in step 100 comprises a linear kernel function, a Gaussian kernel function, a polynomial kernel function and a Sigmoid kernel function, specifically:

[0089] Linear kernel function (Linear Kernel): K(x, y) = x·y, which maps the sample to the original feature space.

[0090] Gaussian Kernel: Measures the distance between two samples by the similarity of Gaussian distribution.

[0091] Polynomial Kernel: K(x,y) = (x·y+c)^d, Measures the distance between two samples by the similarity of polynomial.

[0092] Sigmoid Kernel: K(x,y) = tanh(αx·y+c), Measures the distance between two samples by the similarity of Sigmoid function

[0093] Preferably, the kernel function in step 100 is Gaussian kernel. Gaussian kernel is a kind of kernel function commonly used in machine learning and pattern recognition for nonlinear classification and regression problems. It is based on the probability density function of Gaussian distribution, which is used to measure the similarity between two samples. In this application, the definition of Gaussian kernel is as follows:

[0094] K(x,y) = exp(-||x-y||^2 / (2*σ^2))

[0095] Where K(x,y) represents the similarity between samples x and y, ||x-y|| represents the Euclidean distance between samples x and y, and σ represents the bandwidth parameter of Gaussian kernel.

[0096] The above Gaussian kernel function can map data points to a higher dimensional feature space, so that originally linearly inseparable data becomes linearly separable in the new feature space. By adjusting the bandwidth parameter, the smoothness of the Gaussian kernel function, i.e. the similarity between sample points, can be controlled.

[0097] For step 200, given a set of N seismic data D and the number of seismic facies to be distinguished K, the seismic data is organized into K(K≤N) partitions, S k represents the data set of each partition after partitioning, where a partition is called a cluster, and each cluster represents a seismic facies.

[0098] In step 300, rock physical parameters are a set of important parameters that describe the properties and behavior of rocks. These parameters are of great significance for the exploration, development and production of rocks, and can be used for interpreting seismic data, predicting reservoir properties, evaluating rock mechanical behavior, etc. The following are some common rock physical parameters:

[0099] Density: The density of rock refers to the mass of rock per unit volume, expressed in g / cm 3 or kg / m 3Density: Density is the mass per unit volume of a substance, often expressed in g / cm³ or kg / m³. It is an important parameter in geophysical exploration, rock imaging, and reservoir evaluation.

[0100] Acoustic Velocity: Acoustic velocity is the speed at which sound waves travel through a medium, often expressed in m / s. It is related to the density and elastic modulus of the rock and can be used to determine rock type and predict reservoir properties.

[0101] Elastic Modulus: Elastic modulus is a parameter that describes the stiffness and deformation ability of a material. Commonly used elastic moduli include Young's modulus, shear modulus, and bulk modulus. It is used in rock mechanics studies and to predict the elastic properties of rocks.

[0102] Porosity: Porosity is the ratio of the volume of pores to the total volume of a rock, often expressed as a percentage. It is an important parameter in reservoir evaluation and fluid migration, and is used to assess the rock's ability to store fluids and its permeability.

[0103] Permeability: Permeability is the ability of a fluid to flow through a porous medium, often expressed in mD (millidarcy). It is important in reservoir evaluation and oil and gas development.

[0104] Saturation: Saturation is the proportion of pore space in a rock that is filled with a fluid. Commonly used saturations include water saturation and oil saturation.

[0105] Elastic Modulus: The elastic modulus of a rock describes the degree to which it responds to external stress. Common elastic moduli include bulk modulus and shear modulus.

[0106] Friction Angle: Friction angle is the internal friction characteristic between rock or soil particles. It is important in studying the stability and bearing capacity in geotechnical engineering.

[0107] Depositional Environment: Depositional environment refers to the geographical or geological conditions under which rocks are formed, including the physical and chemical environment of rock deposition. Depositional environment is an important concept for describing the depositional characteristics, sedimentary structures, and rock composition of rocks, and has important significance for rock classification, interpretation, and oil and gas exploration. Depositional environment is mainly influenced by the following factors:

[0108] Hydrodynamic conditions: These include water depth, current velocity, wave energy, and other factors. Different hydrodynamic conditions can lead to different types of sedimentary facies, such as river, lake, marine, estuary, and coastal facies.

[0109] Sediment source: The source of sediments includes land, sea, climate, and other factors. Different sediment sources can affect the type and composition of sediments.

[0110] Sediment transport and deposition processes: These include sediment transport, deposition rate, and deposition mode. Sediment processes can affect the grain size, sorting, and other characteristics of sediments.

[0111] Common sedimentary facies include:

[0112] River sedimentary facies: Formed by river erosion, transport, and deposition, characterized by river channels, alluvial plains, and river deltas.

[0113] Lake sedimentary facies: Formed by lake environments, characterized by lake deposits, lake shore deposits, and lake bank deposits.

[0114] Marine sedimentary facies: Formed by marine environments, characterized by marine deposits, submarine fans, and coastal deposits.

[0115] Coastal sedimentary facies: Formed by the junction of the sea and land, characterized by beach deposits, estuary deposits, and coastal deposits.

[0116] Seismic facies is a term used to describe the discontinuities or wave patterns in seismic data that reflect different geological characteristics. Seismic facies can be used to interpret subsurface geological structures, lithology, and reservoir properties, and has important implications for oil and gas exploration, geological hazard assessment, and geological engineering. Seismic facies are determined and interpreted through information such as amplitude, frequency, phase, and wave shape in seismic data. Common seismic facies include:

[0117] Reflection facies: Reflection facies refers to the discontinuities on seismic profiles that reflect the existence and properties of subsurface geological interfaces. Reflection facies can be used to identify stratigraphic interfaces, construct stratigraphic models, and interpret subsurface structures.

[0118] Hump facies: Hump facies refers to the wave patterns on seismic profiles that resemble a grass hat shape, commonly found near faults or fracture zones. Hump facies can be used to identify faults, assess geostress conditions, and interpret tectonic activity.

[0119] Flat Facies: Flat Facies refers to the flat or nearly flat waveform characteristics presented on the seismic section, which is often found in homogeneous reservoirs or sedimentary layers. Flat Facies can be used to identify reservoirs, evaluate reservoir properties and predict fluid distribution.

[0120] Noise Facies: Noise Facies refers to the high-frequency noise or random fluctuation waveform characteristics presented on the seismic section, which is often found in seismic data noise or signal interference. Noise Facies needs to be removed or filtered to improve data quality and interpretation reliability.

[0121] Through the interpretation and analysis of seismic facies, information about underground geological structure, sedimentary facies, reservoir characteristics, etc. can be obtained, which can assist in geological interpretation, reserve evaluation and well location optimization, etc. At the same time, seismic facies can also be analyzed with geological facies, lithological facies, etc. to improve the understanding and prediction ability of underground geology.

[0122] In some embodiments of the present application, referring to Figure 2 , the seismic facies classification method of seismic data further comprises:

[0123] Step 400: generating a kernel function of each data according to a position kernel function of each data in data space and a data kernel function.

[0124] The position kernel function is a kernel function based on the spatial position information between samples to calculate the similarity. Specifically, the similarity between samples is measured by comparing the distance or proximity between them. Commonly used position kernel functions include Gaussian kernel function and exponential kernel function. The Gaussian kernel function can calculate the similarity by the Euclidean distance between samples, while the exponential kernel function calculates the similarity by the Manhattan distance between samples. Position kernel function is commonly used in fields such as spatial data analysis, image processing and geographic information system.

[0125] A data kernel is a kernel function that measures the similarity between samples based on their data features. Specifically, the similarity between samples is measured by comparing their feature vectors. Common data kernels include the linear kernel, the polynomial kernel, and the radial basis function kernel. The linear kernel directly computes the inner product between feature vectors as the similarity, the polynomial kernel computes the similarity by mapping feature vectors to a high-dimensional space, and the radial basis function kernel computes the similarity by the Euclidean distance between samples. Data kernels are commonly used in machine learning methods such as support vector machines (SVM), kernel principal component analysis (KPCA), and clustering analysis.

[0126] In some embodiments of the present application, referring to Figure 3 , step 100 comprises:

[0127] Step 101: Perform the following iterative operations:

[0128] Step 102: On the data space, compute the weighted average of the kernel functions of all data points within the position kernel function;

[0129] Step 103: Compute the offset of all data points within the position kernel function to the weighted average;

[0130] Step 104: Move the position of the position kernel function on the data space according to the weighted average until the offset is less than a first predetermined value;

[0131] Wherein, the similarity between all data within the position kernel function is the highest.

[0132] In steps 101 to 104, suppose we want to divide L-dimensional data X on N data points into K classes. Then the data can be represented as X ij , where i = 1, 2, …, N represents the data point serial number, j = 1, 2, …, L represents the variable serial number of data on each data point (length of seismic trace). The data point position is represented by P i . When the data length (stratum thickness) of all data points is the same, the data length (stratum thickness) variable is omitted, and the data is represented as X i .

[0133] Then the position kernel function is moved in the data space using the mean shift clustering method, which is a sliding window based method and a cluster center based algorithm, which means that it aims to locate the center point of each cluster by updating the center point candidate to the mean of the points within the sliding window.

[0134] First, a position kernel function K s (P) and a data kernel function K r (X) are defined, and the position kernel function is actually a moving window (a circle in two dimensions and a sphere in three dimensions) with a radius h s . For example, a Gaussian kernel is selected The position kernel function is The data kernel function is and the two form the kernel function K(X) of the mean shift algorithm,

[0135] Then, an arbitrary data point is selected as the center of the window (the center of the circle), and the data on the data point is marked as the cth class (initially c = 1), that is, k = c, and the data on the data point is taken as the mean value of the cth class data, and all points within the window range are marked as the cth class, and the probability (frequency) that these points belong to this class is added by 1.

[0136] The weighted average value of the kernel function of all data points within the window (with a radius h s ) is calculated, and the average distance (offset) of other points within the window to the point is calculated.

[0137] The window is moved to the data point where the mean value is located, and the data point is taken as the new center of the window, and the above three steps are repeated until the offset is less than a certain threshold, and the points encountered in this iteration process are classified into the cth cluster, and the probability number of the points belonging to the cth cluster is added by 1.

[0138] In some embodiments of the present application, referring to Figure 4 , step 200 comprises:

[0139] Step 201: determining the probability of each data being divided into the corresponding cluster data in the K cluster data according to the similarity between the data;

[0140] Step 202: dividing the seismic data set in the target area into K cluster data according to the probability to generate the division result.

[0141] In some embodiments of the present application, referring to Figure 5 , step 200 further comprises:

[0142] Step 203: if the distance between the center of the current cluster data and the center of another cluster data is less than the second preset distance, the current cluster data is divided into the another cluster data.

[0143] If the distance between the center of the current cluster c and the center of another cluster c2 is less than a certain threshold, the current cluster is classified as c2.

[0144] In some embodiments of the present application, referring to Figure 6 , step 200 further comprises:

[0145] Step 204: if the distance between the center of the current cluster data and the center of another cluster data is not less than the second preset distance, a cluster data class is added.

[0146] If the distance between the center of the current cluster c and the center of another cluster c2 is less than a certain threshold, the cluster class is +1.

[0147] It should be noted that if a data point is traversed by both cluster c1 and cluster c2, it is classified as the cluster class with more marked frequencies.

[0148] In some embodiments of the present application, referring to Figure 7 , step 300 comprises:

[0149] Step 301: generating a seismic facies of the work area according to the classification result;

[0150] Step 302: calibrating the seismic facies according to the rock physical parameter and / or the sedimentary facies to determine a final seismic facies classification result of the seismic data set.

[0151] For step 301 and step 302, specifically, the waveform classification result is used for sedimentary facies analysis. The waveform classification result forms discrete "seismic facies" on a plane, and the rock physical parameter or the sedimentary facies is calibrated to the seismic facies map, so that the numerical value is added to the interpretation result, and the qualitative seismic facies map becomes a quantitative map showing the spatial variation of the selected rock physical parameter or sedimentary facies.

[0152] As can be seen from the above description, the embodiment of the present application provides a seismic facies classification method of seismic data, which comprises the following steps: first, determining the similarity between each data and other data in a seismic data set in a target work area according to the kernel function of each data, to determine the similarity between all data in the seismic data set; then, dividing the seismic data set in the target work area into K cluster data according to the similarity between all data, to generate a division result; wherein K is equal to the number of types of seismic facies in the target work area; finally, calibrating the division result according to the rock physical parameter and / or the sedimentary facies of the target work area, to determine the seismic facies classification result of the seismic data set.

[0153] The present application classifies seismic trace shapes by using mean shift algorithm according to the lateral variability of seismic signals in a certain target layer section, and the classification result forms discrete "seismic facies". Then, the seismic facies are calibrated with petrophysical parameters or sedimentary facies, so as to realize the use of seismic data to depict the planar distribution rule of petrophysical parameters or sedimentary facies.

[0154] Embodiment three:

[0155] In one specific embodiment, the present application also provides a specific embodiment of a seismic facies classification method for seismic data, which specifically includes the following steps.

[0156] Waveform classification methods refer to methods that classify and identify collected waveform data. Waveform classification methods have wide applications in signal processing, pattern recognition, machine learning, and other fields, and can be used to identify different signal types, fault patterns, seismic waveforms, etc. The following are some existing waveform classification methods:

[0157] Statistical feature method: This method extracts statistical features of waveform data such as mean, variance, peak value, etc., and then uses these features as input for classification. Common statistical feature methods include mean-variance method, peak-valley method, etc.

[0158] Time-frequency analysis method: This method combines time and frequency domain information by calculating the characteristics of waveform data in the time-frequency domain for classification. Common time-frequency analysis methods include Short-Time Fourier Transform (STFT), Wavelet Transform, time-frequency distribution, etc.

[0159] Machine learning method: This method uses machine learning algorithms to classify waveform data. Common machine learning methods include Support Vector Machine (SVM), Artificial Neural Network (ANN), Decision Tree, etc. Machine learning methods learn classification models through training samples and feature extraction, and use the model to classify new samples.

[0160] Deep learning method: This method is an extension of machine learning, mainly using deep neural networks (Deep Neural Network) for waveform classification. Deep learning methods have strong learning and representation capabilities, and can automatically learn features from raw waveform data and perform classification.

[0161] The above methods are all "hard" classification methods, that is, strictly limit the data on a certain data point to belong to which category, and has nothing to do with other categories. In fact, the data or sample may be similar to other categories, but not the most similar, so it means that the data or sample may belong to multiple categories, but the membership degree is different for different categories, so a "soft" classification method needs to be designed for waveform classification.

[0162] Based on this, the specific implementation of the seismic facies classification method of the seismic data provided by the application comprises the following steps:

[0163] S1: clustering the seismic data of the target work area;

[0164] Specifically, under the condition of a given N-channel seismic data set D and the number K of seismic facies to be distinguished, the seismic data is organized into K (K≤N) partitions, representing the data set of each partition after partitioning, wherein a partition is called a cluster, and each cluster represents a seismic facies.

[0165] S2: establishing a kernel function based on the mean shift algorithm;

[0166] Suppose that L-dimensional data X on N data points is divided into K categories. Then the data can be represented as X ij , wherein i=1, 2,..., N represents the data point serial number, j=1, 2,..., L represents the variable serial number (length of the seismic channel) of the data on each data point. The data point position is represented by P i . When the data length of all data points is the same, the data length variable is omitted, and the data is represented as X i .

[0167] The mean shift clustering is a sliding window-based algorithm, which is a cluster center-based algorithm, which means that its goal is to locate the center point of each cluster, and the center point candidate is updated to the mean value of the points in the sliding window.

[0168] First, define a position kernel function K s (P) and a data kernel function K r (X), and the position kernel function is actually a moving window with a radius of h s (for example, a circle in two dimensions and a sphere in three dimensions). For example, a Gaussian kernel is selected , the position kernel function is , the data kernel function is , and the two form the kernel function K(X) of the mean shift algorithm,

[0169] Then, an arbitrary data point is selected as the center of the window (the center of the circle), and the data on the data point is marked as the cth class (initially c=1), that is, k=c, the data on the data point is taken as the average value of the data of the cth class, and meanwhile, all the points in the window range are marked as the cth class, and the probability (frequency) of the points belonging to the class is added by 1.

[0170] S3: the kernel function weighted average value of the data on all the data points in the window (the radius is h s ) is calculated, and meanwhile, the average distance (offset) of other points in the window to the point is calculated.

[0171] S4: the window is moved to the data point where the data average value is located, the data point is taken as the new center of the window, and the above three steps are repeated until the offset is less than a certain threshold value, and the points encountered in the iteration process are classified into the cluster class c, and the frequency of the points belonging to the cluster class c is added by 1.

[0172] In addition, if the distance between the center of the current cluster class c and the center of another cluster class c2 is less than a certain threshold value, the current cluster is classified as c2, otherwise the class of the cluster is added by 1.

[0173] S5: steps S2 to S4 are repeated until all the data points have been traversed.

[0174] In particular, if a point is traversed by both the cluster c1 and the cluster c2, the point is classified into the cluster class with the higher frequency.

[0175] S6: according to the access frequency of each point in each class, the class with the maximum access frequency is taken as the class to which the current point set belongs.

[0176] Finally, the waveform classification result is used for sedimentary facies analysis. The waveform classification result forms discrete'seismic facies' on the plane, and the seismic facies map is calibrated with rock physical parameters or sedimentary facies, so that the numerical value is added to the interpretation result, so that the qualitative seismic facies map becomes a quantitative map showing the spatial variation of the selected rock physical parameter or sedimentary facies.

[0177] As can be seen from the above description, the embodiment of the present application provides a seismic facies classification method of seismic data, which comprises the following steps: firstly, the similarity between each data in a seismic data set in a target work area and other data is determined according to the kernel function of the data, so as to determine the similarity between all the data in the seismic data set; then, the seismic data set in the target work area is divided into K cluster data according to the similarity between all the data, so as to generate a division result; wherein K is equal to the number of seismic facies types in the target work area; finally, the division result is calibrated according to the rock physical parameters and / or sedimentary facies of the target work area, so as to determine the seismic facies classification result of the seismic data set.

[0178] The method for classifying seismic facies of seismic data provided by the embodiment of the present application classifies seismic trace shapes by using mean shift algorithm according to the lateral variability of seismic signals in a certain target layer section, and the classification result forms discrete "seismic facies". Then, the seismic facies are calibrated with respect to petrophysical parameters or sedimentary facies, so as to realize the description of the planar distribution rule of petrophysical parameters or sedimentary facies by using seismic data.

[0179] Embodiment four:

[0180] Based on the same inventive concept, the embodiment of the present application further provides a device for classifying seismic facies of seismic data, which can be used to realize the method described in the above embodiments. Since the principle of solving problems of the device for classifying seismic facies of seismic data is similar to that of the method for classifying seismic facies of seismic data, the implementation of the device for classifying seismic facies of seismic data can be referred to the implementation of the method for classifying seismic facies of seismic data, and the repeated parts will not be described herein. The term "unit" or "module" used below can be a combination of software and / or hardware that realizes a predetermined function. Although the system described in the following embodiments is preferably realized in software, the realization of hardware or a combination of software and hardware is also possible and is conceived.

[0181] The embodiment of the present application provides a specific implementation of a device for classifying seismic facies of seismic data, which can realize the method for classifying seismic facies of seismic data, referring to Figure 8 , the device for classifying seismic facies of seismic data comprises:

[0182] a data similarity determination module 10, configured to determine the similarity between each data and other data in the seismic data set in the target work area according to the kernel function of each data, so as to determine the similarity between all data in the seismic data set;

[0183] a division result generation module 20, configured to divide the seismic data set in the target work area into K cluster data according to the similarity between all data, so as to generate a division result; wherein, the K is equal to the number of seismic facies types in the target work area;

[0184] a seismic facies classification result determination module 30, configured to calibrate the division result with respect to the petrophysical parameters and / or sedimentary facies of the target work area, so as to determine the seismic facies classification result of the seismic data set.

[0185] In an embodiment of the present application, the device for classifying seismic facies of seismic data further comprises:

[0186] a kernel function generation module, configured to generate the kernel function of each data according to the position kernel function of each data in the data space and the data kernel function.

[0187] In an embodiment of the present application, the data similarity determining module comprises:

[0188] an iteration operation unit configured to perform the following iteration operation:

[0189] a weighted average value calculating unit configured to calculate a weighted average value of the kernel functions of all data points within the position kernel function on the data space;

[0190] an offset calculating unit configured to calculate an offset of all data points within the position kernel function to the weighted average value;

[0191] a position kernel function moving unit configured to move the position of the position kernel function on the data space according to the weighted average value until the offset is less than a first preset value;

[0192] wherein the similarity between all data within the position kernel function is the highest.

[0193] In an embodiment of the present application, the division result generating module comprises:

[0194] a probability determining unit configured to determine a probability of each data division into a corresponding cluster data in the K cluster data according to the similarity between the data;

[0195] a division result generating unit configured to divide the seismic data set in the target work area into the K cluster data according to the probability to generate the division result.

[0196] In an embodiment of the present application, the division result generating module further comprises:

[0197] a data division unit configured to divide the current cluster data into another cluster data if the distance between the current cluster data center and the cluster data center of the another cluster data is less than a second preset distance.

[0198] In an embodiment of the present application, the division result generating module further comprises:

[0199] a cluster category increasing unit configured to increase a cluster data category if the distance between the current cluster data center and the cluster data center of another cluster data is not less than the second preset distance.

[0200] In an embodiment of the present application, the seismic facies classification result determining module comprises:

[0201] a seismic facies generating unit configured to generate the seismic facies of the work area according to the classification result;

[0202] a seismic facies classification result determining unit configured to determine the final seismic facies classification result of the seismic data set according to the rock physical parameters and / or the sedimentary facies calibration to the seismic facies.

[0203] From the above description, the embodiment of the application provides a seismic facies classification device of seismic data, which comprises a data similarity determination module, a division result generation module and a seismic facies classification result determination module.

[0204] The seismic facies classification device of seismic data provided by the embodiment of the application classifies seismic trace shapes by using a mean shift algorithm according to the lateral variability of seismic signals in a certain target layer section, and the classification result forms discrete "seismic facies". Then, the seismic facies are calibrated by rock physical parameters or sedimentary facies, so that the distribution of rock physical parameters or sedimentary facies is described by using seismic data.

[0205] Embodiment five:

[0206] The embodiment of the application further provides a specific implementation of an electronic device capable of realizing all steps in the seismic facies classification method of seismic data in the above embodiments, which is described with reference to Figure 9 The electronic device specifically comprises the following contents:

[0207] A processor 1201, a memory 1202, a communications interface 1203 and a bus 1204;

[0208] The processor 1201, the memory 1202 and the communications interface 1203 are connected to each other through the bus 1204; the communications interface 1203 is used to realize information transmission between the server-side device, the client-side device and other related devices;

[0209] The processor 1201 is used to call a computer program in the memory 1202, and when the processor executes the computer program, all steps in the seismic facies classification method of seismic data in the above embodiments are realized, for example, when the processor executes the computer program, the following steps are realized:

[0210] According to the kernel function of each data in the seismic data set in the target work area, the similarity of the data and other data is determined, so as to determine the similarity between all data in the seismic data set;

[0211] partition the seismic data set in the target work area into K cluster data according to the similarity between all data, to generate a partition result; wherein the K is equal to the number of seismic facies types in the target work area;

[0212] calibrate the partition result according to the petrophysical parameters and / or sedimentary facies of the target work area, to determine the seismic facies classification result of the seismic data set.

[0213] Embodiment six:

[0214] The embodiments of the present application also provide a computer readable storage medium capable of implementing all steps of the seismic facies classification method of seismic data in the above embodiments. The computer readable storage medium stores a computer program. When the computer program is executed by a processor, all steps of the seismic facies classification method of seismic data in the above embodiments are implemented. For example, when the processor executes the computer program, the following steps are implemented:

[0215] determine the similarity between all data in the seismic data set in the target work area according to the kernel function of each data in the seismic data set;

[0216] partition the seismic data set in the target work area into K cluster data according to the similarity between all data, to generate a partition result; wherein the K is equal to the number of seismic facies types in the target work area;

[0217] calibrate the partition result according to the petrophysical parameters and / or sedimentary facies of the target work area, to determine the seismic facies classification result of the seismic data set.

[0218] Each of the embodiments in the specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other. Each embodiment mainly describes the difference from other embodiments. Especially, for the hardware + program type embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the related parts can be referred to the part of the method embodiment.

[0219] The above describes specific embodiments of the present specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be executed in an order different than the order in the embodiments and still achieve the desired result. In addition, the processes depicted in the figures do not necessarily require the particular order shown, or sequential order, to achieve the desired results. In certain implementations, multitasking and parallel processing can be advantageous.

[0220] Although the method operations can be implemented as software, the operations can be implemented in hardware, or a combination of software and / or firmware and hardware. For example, various functions performed in the method(s) can be implemented in hardware within a single integrated circuit, multiple integrated circuits, or a combination thereof. Alternatively, the various functions performed in the method(s) can be implemented in a plurality of hardware components.

[0221] For ease of description, the above apparatus is described as various modules for performing respective functions. Of course, the functions of the modules can be implemented in one or more software and / or hardware components, and the modules can be combined or integrated into another system, or some features can be ignored or not performed. In addition, the coupling or direct coupling or communication connection between the modules shown or discussed can be indirect coupling or communication connection through some interface, device or unit, and can be electrical, mechanical or other form.

[0222] Those skilled in the art will also appreciate that, in addition to being implemented in purely computer readable program code means, the controller can be implemented using logic programming to cause the controller to perform the same functions in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers and embedded microcontrollers. Therefore, the controller can be considered as a hardware component, and the means for implementing various functions included therein can also be considered as structures within the hardware component. Alternatively, the means for implementing various functions can be considered as both software modules implementing the method and structures within the hardware component.

[0223] In one typical arrangement, the computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0224] The memory can include non-persistent memory and / or volatile memory, such as random access memory (RAM) and / or cache memory, non-volatile memory, such as read-only memory (ROM), EPROM, and / or flash memory. The memory is an example of computer readable media.

[0225] Each of the embodiments in the specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other. Each of the embodiments focuses on the difference from other embodiments. In particular, the system embodiments are described simply because they are basically similar to the method embodiments, and the relevant parts can be referred to the description of the method embodiments. In the description of the specification, the description of the terms "one embodiment", "some embodiments", "an example", "a specific 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 the embodiments of the specification. The illustrative description of the above terms in the specification does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any suitable manner in any one or more embodiments or examples. In addition, the person skilled in the art can combine and combine the different embodiments or examples described in the specification and the features of the different embodiments or examples without contradiction.

[0226] The above only describes the embodiments of the embodiments of the specification and does not limit the embodiments of the specification. The embodiments of the specification can have various changes and variations for those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the embodiments of the specification shall be included in the scope of claims of the embodiments of the specification.

Claims

1. A method of seismic facies classification of seismic data, characterized in that, The method comprises the following steps: determining the similarity between each data in the seismic data set in the target work area and other data according to the kernel function of the data, to determine the similarity between all data in the seismic data set; dividing the seismic data set in the target work area into K cluster data according to the similarity between all data, to generate a division result; wherein K is equal to the number of seismic facies types in the target work area; calibrating the division result according to the rock physical parameters and / or sedimentary facies of the target work area, to determine the seismic facies classification result of the seismic data set. The step of determining the similarity between each data in the seismic data set in the target work area and other data according to the kernel function of the data comprises the following iterative operations: calculating the weighted average value of the kernel function of all data points in the position kernel function on the data space; calculating the offset of all data points in the position kernel function to the weighted average value; moving the position of the position kernel function on the data space according to the weighted average value until the offset is less than a first preset value; wherein the similarity between all data in the position kernel function is the highest.

2. The seismic facies classification method of claim 1, wherein, The method further comprises the following steps: generating the kernel function of each data according to the position kernel function and the data kernel function of the data on the data space.

3. The seismic facies classification method of claim 1, wherein, The step of dividing the seismic data set in the target work area into K cluster data according to the similarity between all data, to generate a division result, comprises the following steps: determining the probability of each data being divided into the corresponding cluster data in the K cluster data according to the similarity between the data; dividing the seismic data set in the target work area into K cluster data according to the probability, to generate the division result.

4. The seismic facies classification method of claim 1, wherein, The step of dividing the seismic data set in the target work area into K cluster data according to the similarity between all data, to generate a division result, further comprises the following steps: if the distance between the current cluster data center and another cluster data center is less than a second preset distance, dividing the current cluster data into the other cluster data.

5. The seismic facies classification method of claim 1, wherein, The step of dividing the seismic data set in the target work area into K cluster data according to the similarity between all data, to generate a division result, further comprises the following steps: if the distance between the current cluster data center and another cluster data center is not less than a second preset distance, increasing a cluster data category.

6. The seismic facies classification method according to any one of claims 1 to 5, characterized in that, The step of calibrating the division result according to the rock physical parameters and / or sedimentary facies of the target work area, to determine the seismic facies classification result of the seismic data set, comprises the following steps: generating the seismic facies of the work area according to the classification result; calibrating the seismic facies according to the rock physical parameters and / or sedimentary facies, to determine the final seismic facies classification result of the seismic data set.

7. An apparatus for seismic facies classification of seismic data, characterized by The method comprises the following steps: a data similarity determination module is configured to determine the similarity between each data in the seismic data set in the target work area and other data according to the kernel function of the data, to determine the similarity between all data in the seismic data set; a division result generation module is configured to divide the seismic data set in the target work area into K cluster data according to the similarity between all data, to generate a division result; wherein K is equal to the number of seismic facies types in the target work area. The seismic facies classification result determination module is configured to calibrate the division result according to the petrophysical parameters and / or the sedimentary facies of the target work area, so as to determine a seismic facies classification result of the seismic data set. The data similarity determination module comprises an iterative operation unit configured to perform the following iterative operations: a weighted average value calculation unit configured to calculate a weighted average value of kernel functions of all data points within a position kernel function on a data space; an offset calculation unit configured to calculate offsets of all data points within the position kernel function to the weighted average value; and a position kernel function moving unit configured to move the position of the position kernel function on the data space according to the weighted average value until the offset is less than a first preset value; wherein the similarity between all data within the position kernel function is the highest.

8. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the program to implement the steps of the seismic facies classification method of the seismic data according to any one of claims 1 to 6.

9. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the seismic facies classification method of the seismic data according to any one of claims 1 to 6.

Images