Method and device for physical simulation volcano channel depiction recognition

Abstract

The invention provides a method and a device for depicting and identifying a physical simulation volcano channel. In order to solve the problem that the volcanic channel is difficult to identify on the profile, has no feature or only presents weak reflection, the method finds out the attribute sensitive to the volcanic channel and strengthens the volcanic channel through various attribute analysis, so that the aim of depicting the volcanic channel can be achieved by applying tracking and rendering technologies. And a basis is provided for volcanic rock facies prediction and gas-bearing prediction.

Description

Technical Field

[0001] The present invention relates to the technical field of geological exploration, and more particularly, to a method and apparatus for physical simulation of volcanic channel characterization and identification. Background Art

[0002] With the development of social economy, exploring and exploiting potential volcanic rock oil and gas resources has become an important strategic direction. Compared with conventional oil and gas reservoirs, volcanic rock oil and gas has advantages such as large resource volume and long exploitation cycle, and is defined as one of the important new clean energy sources. Therefore, accelerating the exploration and development of volcanic rock oil and gas fields has great economic and social significance.

[0003] However, in the actual exploration and development process, the overall heterogeneity of volcanic rock reservoirs is strong, the lithology identification is difficult, and a series of problems such as high drilling risk and operation cost seriously restrict the progress. In particular, the fine identification and description of the key volcanic eruption channels directly affect the development, modeling and production effectiveness, and are the weak links and difficulties of the current technology. This is because the strong noise and complex anisotropic effects inside the channels lead to poor quality of seismic data signals and it is impossible to directly obtain clear and interpretable geological information. The current conventional manual interpretation methods and technical means are difficult to effectively achieve the automatic and accurate locking and characterization of small volcanic channels. This seriously restricts the effective exploration and utilization of the entire volcanic rock oil and gas resources. Therefore, there is an urgent need to develop new algorithms and calculation methods to achieve the efficient automatic description and characterization of complex volcanic channels, which is of great significance for promoting the industrial-scale development of volcanic rock oil and gas. Summary of the Invention

[0004] In view of this, the present invention aims to provide a technical solution that can accurately characterize volcanic channels based on seismic physical simulation.

[0005] According to one aspect of the present invention, a method for physical simulation of volcanic channel characterization and identification is proposed, including the following steps:

[0006] Step 1, perform anisotropic diffusion calculation, structural texture calculation and local variance calculation on the profile obtained from the seismic physical simulation model to obtain structural-guided anisotropic characteristics, structural texture attributes and local variance attributes;

[0007] Step 2, perform waveform difference multi-channel similarity coherence body calculation based on the structural-guided anisotropic characteristics to extract coherence attributes;

[0008] Step 3, establish a grid stratigraphic body according to the contact relationship of the strata, and assign the coherence body attributes, structural texture attributes and local variance attributes to the grid stratigraphic body;

[0009] Step 4, perform normalization processing on the data of the grid stratigraphic body;

[0010] Step 5: Convert the normalized grid formation body data into seismic body data, and obtain training seed data based on the seismic body data to train a deep learning model for predicting volcanic channel features;

[0011] Step 6: Apply the trained deep learning model, and combine the horizontal slices of the seismic body to predict the spatial position and shape of the volcanic channel geological body.

[0012] In some embodiments, in Step 5, the formation slicing technique is applied to convert the normalized grid formation body data into seismic body data.

[0013] In some embodiments, in Step 5, obtaining the training seed data based on the seismic body data includes:

[0014] Trace the geological body seed points every 20 main survey lines on the seismic body data to obtain the training seed data.

[0015] In some embodiments, the method further includes:

[0016] Add a random volcanic channel to the seismic physical simulation model for dynamic simulation demonstration to filter the trained deep learning model.

[0017] According to another aspect of the present invention, a device for physical simulation of volcanic channel characterization and recognition is also proposed, including:

[0018] An attribute calculation unit, configured to perform anisotropic diffusion calculation, structural texture calculation, and local variance calculation on the profile obtained from the seismic physical simulation model to obtain structural orientation anisotropic characteristics, structural texture attributes, and local variance attributes;

[0019] A coherence extraction unit, configured to perform waveform difference multi-channel similarity coherence body calculation based on the structural orientation anisotropic characteristics to extract coherence attributes;

[0020] A grid formation body construction unit, configured to establish a grid formation body according to the contact relationship of the formations, and assign the coherence body attributes, structural texture attributes, and local variance attributes to the grid formation body;

[0021] A normalization unit, configured to perform normalization processing on the data of the grid formation body;

[0022] A deep learning model training unit, configured to convert the normalized grid formation body data into seismic body data, and obtain training seed data based on the seismic body data to train a deep learning model for predicting volcanic channel features;

[0023] A prediction unit, configured to apply the trained deep learning model, and combine the horizontal slices of the seismic body to predict the spatial position and shape of the volcanic channel geological body.

[0024] In some embodiments, in the deep learning model training unit, the normalized grid formation data is converted into seismic volume data by applying the formation slicing technique.

[0025] In some embodiments, in the deep learning model training unit, obtaining training seed data based on the seismic volume data includes:

[0026] Tracking geological body seed points every 20 main survey lines on the seismic volume data to obtain the training seed data.

[0027] In some embodiments, the device further includes:

[0028] A model filtering unit, configured to add random volcanic channels to the seismic physical simulation model for simulating dynamic schematics, and filter the trained deep learning model.

[0029] According to another aspect of the present invention, an electronic device is further provided, and the electronic device includes:

[0030] A memory storing executable instructions;

[0031] A processor, the processor running the executable instructions in the memory to implement the method for physical simulation of volcanic channel characterization and recognition described above.

[0032] According to another aspect of the present invention, a computer-readable storage medium is further provided, and the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, it implements the method for physical simulation of volcanic channel characterization and recognition described above.

[0033] The present invention has at least the following beneficial effects.

[0034] 1. Improve the accuracy of volcanic channel recognition

[0035] The present invention comprehensively applies multi-attribute calculation and deep learning technologies, which can effectively improve data resolution, suppress noise, and enhance the reflection characteristics of geological bodies, thereby greatly improving the accuracy and effect of recognizing structures such as volcanic channels.

[0036] 2. Reduce the amount of manual interpretation and improve work efficiency

[0037] By constructing an automated end-to-end prediction and recognition process, avoiding a large amount of manual survey line tracking and volume interpretation workload, the characterization speed of geological bodies such as volcanic channels can be greatly improved.

[0038] 3. Provide support for subsequent development

[0039] Improving the fine characterization and description of volcanic system channels can better determine the subsequent deployment of drilling well positions, development designs, etc., providing strong technical support for the development of volcanic rock oil and gas resources.

[0040] 4. The proposed solution has broad application prospects for expansion

[0041] In the present invention, the technical process combining multi-attribute analysis and deep learning concepts can be extended and applied to the automatic identification and prediction of other types of geological bodies.

[0042] In summary, the present invention can effectively solve the problems encountered in the current exploration and development of volcanic rocks, making important contributions to promoting technological progress in this field.

[0043] The method and apparatus of the present invention have other characteristics and advantages, which will be obvious in the accompanying drawings and subsequent specific embodiments incorporated herein, or will be described in detail in the accompanying drawings and subsequent specific embodiments incorporated herein. These drawings and specific embodiments are jointly used to explain the specific principles of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] By describing the exemplary embodiments of the present invention in more detail in conjunction with the accompanying drawings, the above and other objects, features, and advantages of the present invention will become more obvious. Among them, in the exemplary embodiments of the present invention, the same reference numerals generally represent the same components.

[0045] Figure 1 Shows a method for physically simulating the characterization and identification of volcanic channels according to an embodiment of the present invention.

[0046] Figure 2 Shows a schematic diagram of a formation established according to an exemplary embodiment of the present invention.

[0047] Figure 3 Shows a schematic diagram of a predicted and characterized volcanic channel according to an exemplary embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0048] The preferred embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although the preferred embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to make the present invention more thorough and complete, and to fully convey the scope of the present invention to those skilled in the art.

[0049] Example 1

[0050] Figure 1The flowchart of a method for physical simulation volcanic channel characterization and recognition according to an embodiment of the present invention is shown. As shown in the figure, the method includes Step 1 to Step 6.

[0051] Step 1, perform anisotropic diffusion calculation, structural texture calculation, and local variance calculation on the profile obtained from the seismic physical simulation model to obtain structural orientation anisotropic characteristics, structural texture attributes, and local variance attributes.

[0052] Performing anisotropic diffusion calculation on the profile obtained from the seismic physical simulation model can effectively implement the feature enhancement processing of post-stack seismic data and significantly improve the accuracy of structural recognition of the data volume.

[0053] In an example, the basic idea of anisotropic diffusion calculation is that by selecting the characteristic wave vectors of structures in different directions as the diffusion control factors and performing non-linear intensity coupling, the diffusion process can reflect anisotropic characteristics. By extracting features such as the dispersion characteristics and coherence changes of seismic data in different directions, anisotropic characteristics related to the dip or development direction of the structure can be obtained, which provides a basis for subsequent physical property inversion and the identification of geological bodies.

[0054] In some embodiments, structural texture calculation can be performed based on the analysis and calculation of multi-azimuth data resampling at a certain spatial point to obtain structural texture attributes.

[0055] In an example, spatial sample points can be determined, the seismic signal image can be resampled from different azimuth angles to obtain the local window information in multiple azimuths of the sample points, and then by designing an azimuth filtering and wavelet transform decomposition algorithm, the frequency energy characteristics of sub-bands in different directions, as well as the angular variation patterns of spectral phase, polarity, and coherence are analyzed. By integrating these changing parameters, the attribute characteristics such as the thickness, density, and dominant direction of the structural texture are quantitatively characterized, so as to effectively enhance and detect the structural contour to obtain structural texture attributes.

[0056] Local differential resampling calculation can be performed on seismic data for local variance calculation. Its core idea is to perform differential sampling and statistics of seismic data in the spatial domain by setting a sliding local analysis window. In some examples, each pixel sample point can be traversed, the signals in its surrounding local area are extracted, the differential results such as the gradient difference in each direction and the gray Laplacian operator are calculated, and the variance and range of these differential results are integrated, which can effectively reflect the characteristics of the signal variation rate in the local area of the sample point. Finally, the local variance attributes of the seismic data in the horizontal and vertical directions can be obtained, which are used to describe the areas where the structures and properties in the formation change rapidly, thus facilitating the identification of the boundary characteristics of geological bodies.

[0057] The effective suppression of noise components can be achieved through Step 1, thereby effectively enhancing the geological signal characteristics in seismic data.

[0058] Step 2: Based on the structure-oriented anisotropic characteristics, perform waveform-difference multi-channel similarity coherence body calculation to extract coherence attributes.

[0059] Based on the structure-oriented anisotropic characteristics obtained in Step 1, perform waveform-difference multi-channel similarity coherence body calculation to extract coherence attributes.

[0060] Through the coherence body technique, the lateral inhomogeneity of formation lithology, etc., can be described by using the variation of seismic signal coherence values, thereby studying the spatial distribution and characteristics of volcanic channels.

[0061] In one example, based on the obtained structure-oriented anisotropic characteristics, the direction of the dominant structural line can be determined. Then, sub-beams of seismic profiles can be extracted along different directions, the waveform differences between the sub-beams can be analyzed, and through multi-channel statistics such as cross-correlation and multi-analysis window stacking averaging, the coherence between the sub-beams can be obtained.

[0062] Step 3: Establish a grid formation body according to the contact relationship of the strata, and assign coherence body attributes, structural texture attributes, and local variance attributes to the grid formation body.

[0063] A reasonable three-dimensional grid-like formation body geometric model can be constructed according to the distribution range of different strata, interface contact relationships, etc.

[0064] The various attributes obtained above - coherence body attributes, structural texture attributes, local variance attributes, etc. - all reflect the structural or physical property change characteristics existing in the strata. These attribute fields of geological information can be correspondingly assigned to the matching grid formation body one by one to achieve binding with the spatial position. After the attribute field configuration, different structural formation bodies have corresponding geological characteristic parameters, providing a basis for subsequent identification of geological body boundaries and conversion into seismic attribute bodies.

[0065] Step 4: Normalize the data of the grid formation body.

[0066] The data of various attributes of the grid formation body can be normalized by taking logarithms or dividing to further enhance the effective geological signals of the target horizon.

[0067] Step 5: Convert the normalized data of the grid formation body into seismic volume data, and based on the seismic volume data, obtain training seed data to train a deep learning model for predicting volcanic channel characteristics.

[0068] In some embodiments, the formation slice technique can be applied to convert the normalized data of the grid formation body into seismic volume data.

[0069] Stratigraphic slices can visually and vividly display the horizontal distribution of stratigraphic attributes, making geological identification and interpretation more convenient. Moreover, the slice images have strong hierarchical correlations, which are more in line with people's cognitive habits and conducive to contour tracking. Therefore, according to this embodiment, using the rich pixel-level information of the slice images for modeling training has a relatively high accuracy. And based on extracting sample points from the slices for attribute mapping and deep learning, the constraint effect of the corresponding relationship is improved. Subsequently, the slices of seismic data are also aligned with the stratigraphic slices, facilitating the display of prediction results.

[0070] In one example, geological body seed points can be tracked every 20 main survey lines on the seismic volume data to obtain training seed data, which is similar to sparse interpretation. The volcanic channels of the seismic profiles are interpreted and tracked at an interval of 20 main survey lines.

[0071] Specifically, in this step, the constructed and configured grid stratigraphic body and its attributes such as anisotropy, coherence, and local variance can be converted and mapped into an equivalent seismic attribute body coordinate system and data. On the seismic attribute body, sampling points containing the target volcanic channel structure information are extracted at a certain interval (for example, every 20 main survey lines). These seed point data with position coordinate labels can be used as inputs to train a deep neural network model to learn lithology prediction, physical property prediction, etc. for volcanic channel features. By adjusting the parameters to optimize the training process, a trained deep learning model can be obtained.

[0072] Step 6: Apply the trained deep learning model to combine with the horizontal slices of the seismic volume to predict the spatial position and shape of the volcanic channel geological body.

[0073] The trained convolutional neural network deep learning model can be deployed. Its input end can access the seismic data volume to be predicted. Specifically, multiple horizontal slice two-dimensional sequences of the seismic data volume can be obtained at a certain time or depth interval. The slice images can be input into the model in sequence, and the model will output the possible existence areas of volcanic channels detected on each slice. By integrating the prediction results of different slices, detailed contour information such as the overall shape and distribution range of the target volcanic channel body structure in three-dimensional space can be obtained.

[0074] The predicted spatial position and shape of the volcanic channel geological body can be used for later interpretation work and can also be used as auxiliary information for automated drilling well location selection, etc.

[0075] In some embodiments, the method further includes:

[0076] Adding random volcanic channels to the seismic physical simulation model for simulation dynamic demonstration to filter the trained deep learning model.

[0077] Based on the initially established seismic physical simulation model, artificial false volcanic channels with random positions and shapes can be added; using the model data with the added false channels, the deep learning network is still trained according to the original process to obtain new prediction results; comparing the prediction results with the distribution of volcanic channels in the seismic physical model, the network is repeatedly adjusted and optimized until it can finally correctly identify all the added false channels.

[0078] According to this embodiment, through repeated testing and optimization, the model can more accurately predict the characteristics of volcanic channels and significantly improve the generalization adaptability of the model.

[0079] This embodiment has at least the following beneficial effects.

[0080] 1. Improve the recognition accuracy of volcanic channels

[0081] This embodiment comprehensively uses multi-attribute calculation and deep learning technologies, which can effectively improve the data resolution, suppress noise, and enhance the reflection characteristics of geological bodies, thus greatly improving the accuracy and effect of recognizing structures such as volcanic channels.

[0082] 2. Reduce the amount of manual interpretation and improve work efficiency

[0083] By constructing an automated end-to-end prediction and recognition process, avoiding a large amount of manual work such as survey line tracking and volume interpretation, the depiction speed of geological bodies such as volcanic channels can be greatly improved.

[0084] 3. Provide support for subsequent development

[0085] Improving the fine depiction of the channels of the volcanic system can better determine the subsequent drilling well location deployment, development design, etc., and provide strong technical support for the development of volcanic rock oil and gas resources.

[0086] 4. The proposed solution has broad application prospects for expansion

[0087] The technical process combining multi-attribute analysis and deep learning concepts in this embodiment can be extended and applied to the automatic recognition and prediction of other types of geological bodies.

[0088] Example 2

[0089] According to an embodiment of the present invention, a device for physically simulating the depiction and recognition of volcanic channels includes:

[0090] An attribute calculation unit for performing anisotropic diffusion calculation, structural texture calculation, and local variance calculation on the profile obtained from the seismic physical simulation model to obtain structural orientation anisotropic characteristics, structural texture attributes, and local variance attributes;

[0091] A coherence extraction unit, configured to calculate a waveform-difference multi-trace similarity coherence volume based on the structure-oriented anisotropic characteristics, so as to extract coherence attributes;

[0092] A grid stratigraphic body construction unit, configured to establish a grid stratigraphic body according to the contact relationship of strata, and endow the grid stratigraphic body with coherence body attributes, structural texture attributes, and local variance attributes;

[0093] A normalization unit, configured to perform normalization processing on the data of the grid stratigraphic body;

[0094] A deep learning model training unit, configured to convert the normalized grid stratigraphic body data into seismic volume data, and obtain training seed data based on the seismic volume data, and train a deep learning model for predicting volcanic channel characteristics;

[0095] A prediction unit, configured to apply the trained deep learning model, and combine the horizontal slice of the seismic volume to predict the spatial position and shape of the volcanic channel geological body.

[0096] In some embodiments, in the deep learning model training unit, the stratum slicing technology is applied to convert the normalized grid stratigraphic body data into seismic volume data.

[0097] In some embodiments, in the deep learning model training unit, obtaining training seed data based on the seismic volume data includes:

[0098] Tracking geological body seed points every 20 main survey lines on the seismic volume data to obtain the training seed data.

[0099] In some embodiments, the apparatus further includes:

[0100] A model filtering unit, configured to add random volcanic channels to the seismic physical simulation model for simulation dynamic demonstration, and filter the trained deep learning model.

[0101] This embodiment has at least the following beneficial effects.

[0102] 1. Improve the recognition accuracy of volcanic channels

[0103] This embodiment comprehensively uses multi-attribute calculation and deep learning technologies, which can effectively improve data resolution, suppress noise, and enhance the reflection characteristics of geological bodies, thereby greatly improving the accuracy and effect of identifying structures such as volcanic channels.

[0104] 2. Reduce the amount of manual interpretation and improve work efficiency

[0105] By constructing an automated end-to-end prediction and recognition process, the workload of a large number of manual survey line tracking and volume interpretation can be avoided, and the depiction speed of geological bodies such as volcanic channels can be greatly improved.

[0106] 3. Provide support for subsequent development

[0107] Improving the fine characterization of the volcanic system channels can better determine the subsequent drilling well location deployment, development design, etc., and provide strong technical support for the development of volcanic rock oil and gas resources.

[0108] 4. The proposed solution has broad application prospects for expansion

[0109] The technical process combining multi-attribute analysis and deep learning concepts in this embodiment can be expanded and applied to the automatic identification and prediction of other types of geological bodies.

[0110] For other detailed descriptions and advantages of this embodiment, reference can be made to the corresponding descriptions in the foregoing embodiments, which will not be elaborated here.

[0111] Example 3

[0112] According to another aspect of the present invention, an electronic device is also provided. The electronic device includes:

[0113] A memory storing executable instructions:

[0114] A processor that runs the executable instructions in the memory to implement the method for physical simulation of volcanic channel characterization and identification according to the present invention.

[0115] Specifically, the memory may include one or more computer program products, and the computer program products may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory, etc. The non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.

[0116] The processor may be a central processing unit (CPU) or other forms of processing units with data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device to perform desired functions. In one embodiment of the present invention, the processor is used to run the computer-readable instructions stored in the memory.

[0117] The method for physical simulation of volcanic channel characterization and identification includes the following steps:

[0118] Step 1, perform anisotropic diffusion calculation, texture construction calculation, and local variance calculation on the profile obtained from the seismic physical simulation model to obtain construction-oriented anisotropic characteristics, construction texture attributes, and local variance attributes;

[0119] Step 2: Perform waveform difference multi-channel similarity coherence body calculation based on the constructed orientation anisotropic characteristics to extract coherence attributes;

[0120] Step 3: Establish a grid stratigraphic body according to the contact relationship of the strata, and assign coherence body attributes, tectonic texture attributes, and local variance attributes to the grid stratigraphic body;

[0121] Step 4: Normalize the data of the grid stratigraphic body;

[0122] Step 5: Convert the normalized grid stratigraphic body data into seismic volume data, and obtain training seed data based on the seismic volume data to train a deep learning model for predicting volcanic channel characteristics;

[0123] Step 6: Apply the trained deep learning model and combine with the horizontal slice of the seismic volume to predict the spatial position and morphology of the volcanic channel geological body.

[0124] In some embodiments, in Step 5, the formation slice technology is applied to convert the normalized grid stratigraphic body data into seismic volume data.

[0125] In some embodiments, in Step 5, obtaining training seed data based on the seismic volume data includes:

[0126] Tracking geological body seed points every 20 main survey lines on the seismic volume data to obtain the training seed data.

[0127] In some embodiments, the method further includes:

[0128] Adding a random volcanic channel to the seismic physical simulation model for simulation dynamic demonstration to filter the trained deep learning model.

[0129] This embodiment has at least the following beneficial effects.

[0130] 1. Improve the recognition accuracy of volcanic channels

[0131] This embodiment comprehensively uses multi-attribute calculation and deep learning technologies, which can effectively improve data resolution, suppress noise, and enhance the reflection characteristics of geological bodies, thereby greatly improving the accuracy and effect of identifying structures such as volcanic channels.

[0132] 2. Reduce the amount of manual interpretation and improve work efficiency

[0133] By constructing an automated end-to-end prediction and recognition process, avoiding a large amount of manual survey line tracking and volume interpretation workload, the characterization speed of geological bodies such as volcanic channels can be greatly improved.

[0134] 3. Provide support for subsequent development

[0135] Improving the fine characterization of volcanic system channels can better determine the subsequent well location deployment, development design, etc., providing strong technical support for the development of volcanic rock oil and gas resources.

[0136] 4. The proposed solution has broad application prospects for expansion

[0137] In this embodiment, the technical process combining multi-attribute analysis and deep learning concepts can be extended and applied to the automatic identification and prediction of other types of geological bodies.

[0138] For the detailed description of this embodiment, reference can be made to the corresponding descriptions in the foregoing embodiments, and details will not be repeated here.

[0139] Example 4

[0140] According to another aspect of the present invention, there is also provided a computer-readable storage medium storing a computer program which, when executed by a processor, implements the method for physical simulation of volcanic channel characterization and identification according to the present invention.

[0141] The computer-readable storage medium according to an embodiment of the present invention stores non-temporary computer-readable instructions. When the non-temporary computer-readable instructions are run by a processor, all or part of the steps of the methods of the various embodiments of the present invention described above are executed.

[0142] The above computer-readable storage medium includes but is not limited to: optical storage media (e.g., CD-ROM and DVD), magneto-optical storage media (e.g., MO), magnetic storage media (e.g., magnetic tape or removable hard drive), media with built-in rewritable non-volatile memory (e.g., memory card), and media with built-in ROM (e.g., ROM cartridge).

[0143] Those skilled in the art should understand that, in order to solve the technical problem of how to obtain good user experience effects, this embodiment may also include well-known structures such as communication buses and interfaces, and these well-known structures should also be included in the protection scope of the present invention.

[0144] The method for physical simulation of volcanic channel characterization and identification includes the following steps:

[0145] Step 1, perform anisotropic diffusion calculation, structural texture calculation, and local variance calculation on the profile obtained from the seismic physical simulation model to obtain structural orientation anisotropic characteristics, structural texture attributes, and local variance attributes;

[0146] Step 2, perform waveform difference multi-channel similarity coherence body calculation based on the structural orientation anisotropic characteristics to extract coherence attributes;

[0147] Step 3: Establish a grid formation body based on the contact relationship of the strata, and assign coherent body attributes, tectonic texture attributes, and local variance attributes to the grid formation body;

[0148] Step 4: Normalize the data of the grid formation body;

[0149] Step 5: Convert the normalized grid formation body data into seismic volume data, and obtain training seed data based on the seismic volume data to train a deep learning model for predicting volcanic channel characteristics;

[0150] Step 6: Apply the trained deep learning model, and combine with the horizontal slices of the seismic volume to predict the spatial position and morphology of the volcanic channel geological body.

[0151] In some embodiments, in Step 5, the formation slicing technique is applied to convert the normalized grid formation body data into seismic volume data.

[0152] In some embodiments, in Step 5, obtaining training seed data based on the seismic volume data includes:

[0153] Tracking geological body seed points every 20 main survey lines on the seismic volume data to obtain the training seed data.

[0154] In some embodiments, the method further includes:

[0155] Adding a random volcanic channel to the seismic physical simulation model for simulation dynamic demonstration to filter the trained deep learning model.

[0156] This embodiment has at least the following beneficial effects.

[0157] 1. Improve the recognition accuracy of volcanic channels

[0158] This embodiment comprehensively uses multi-attribute calculation and deep learning technologies, which can effectively improve the data resolution, suppress noise, and enhance the reflection characteristics of geological bodies, thereby greatly improving the accuracy and effect of identifying structures such as volcanic channels.

[0159] 2. Reduce the amount of manual interpretation and improve work efficiency

[0160] By constructing an automated end-to-end prediction and recognition process, avoiding a large amount of manual survey line tracking and volume interpretation workload, the depiction speed of geological bodies such as volcanic channels can be greatly improved.

[0161] 3. Provide support for subsequent development

[0162] Improving the fine depiction of the volcanic system channel can better determine the subsequent drilling well location deployment, development design, etc., and provide strong technical support for the development of volcanic rock oil and gas resources.

[0163] 4. The proposed solution has broad application prospects for expansion

[0164] In this embodiment, the technical process combining multi-attribute analysis and deep learning concepts can be extended and applied to the automatic identification and prediction of other types of geological bodies.

[0165] For the detailed description of this embodiment, reference can be made to the corresponding descriptions in the foregoing embodiments, which will not be elaborated here.

[0166] Example 5

[0167] Figure 2 The figure shows a schematic diagram of a stratigraphic body established according to an exemplary embodiment of the present invention.

[0168] Figure 3 The figure shows a schematic diagram of a volcanic conduit predicted and depicted according to an exemplary embodiment of the present invention.

[0169] It can be seen that according to the embodiments of the present invention, the spatial position and shape of the volcanic conduit geological body can be accurately depicted.

[0170] It can be understood that the above-mentioned embodiments mentioned in the present disclosure can be combined with each other to form a combined embodiment without violating the principle logic. Due to space limitations, the present disclosure will not elaborate further. Those skilled in the art can understand that in the above method of the specific implementation manner, the specific execution order of each step should be determined according to its function and possible internal logic.

[0171] Note that unless otherwise directly stated, all features disclosed in this specification (including any appended claims, abstract, and drawings) can be replaced by alternative features for achieving the same, equivalent, or similar purposes. Therefore, unless otherwise clearly stated, each feature disclosed is only an example of a group of equivalent or similar features. When used, further, preferably, furthermore, and more preferably are simple introductions for elaborating another embodiment based on the foregoing embodiments. The content following the further, preferably, furthermore, or more preferably in combination with the foregoing embodiments constitutes a complete composition of another embodiment. The components formed by any combination among several further, preferably, furthermore, or more preferably settings following the same embodiment can form another embodiment.

[0172] Those skilled in the art should understand that the embodiments of the present invention described above and shown in the drawings are only examples and do not limit the present invention. The object of the present invention has been fully and effectively achieved. The function and structural principle of the present invention have been demonstrated and explained in the embodiments. Without departing from the said principle, the embodiments of the present invention can be deformed or modified in any way.

[0173] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present disclosure, rather than to limit them; although the present disclosure has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present disclosure.

Claims

1. A method for physically simulating volcanic channel characterization and identification, characterized in that, it includes the following steps: Step 1, perform anisotropic diffusion calculation, structural texture calculation, and local variance calculation on the profile obtained from the seismic physical simulation model to obtain structural orientation anisotropic characteristics, structural texture attributes, and local variance attributes; Step 2, perform waveform difference multi-channel similarity coherence body calculation based on the structural orientation anisotropic characteristics to extract coherence attributes; Step 3, establish a grid stratigraphic body according to the contact relationship of the strata, and assign coherence body attributes, structural texture attributes, and local variance attributes to the grid stratigraphic body; Step 4, perform normalization processing on the data of the grid stratigraphic body; Step 5, convert the normalized grid stratigraphic body data into seismic volume data, and obtain training seed data based on the seismic volume data to train a deep learning model for predicting volcanic channel characteristics; Step 6, apply the trained deep learning model, and combine the horizontal slice of the seismic volume to predict the spatial position and morphology of the volcanic channel geological body.

2. The method according to claim 1, characterized in that, in the step 5, apply the formation slicing technology to convert the normalized grid stratigraphic body data into seismic volume data.

3. The method according to claim 1, characterized in that, in step 5, obtaining the training seed data based on the seismic volume data includes: Track the geological body seed points every 20 main survey lines on the seismic volume data to obtain the training seed data.

4. The method according to claim 1, characterized in that, the method further includes: Add a random volcanic channel to the seismic physical simulation model for simulation dynamic demonstration, and filter the trained deep learning model.

5. A device for physically simulating volcanic channel characterization and identification, characterized in that, it includes: An attribute calculation unit for performing anisotropic diffusion calculation, structural texture calculation, and local variance calculation on the profile obtained from the seismic physical simulation model to obtain structural orientation anisotropic characteristics, structural texture attributes, and local variance attributes; A coherence extraction unit for performing waveform difference multi-channel similarity coherence body calculation based on the structural orientation anisotropic characteristics to extract coherence attributes; A grid stratigraphic body construction unit for establishing a grid stratigraphic body according to the contact relationship of the strata, and assigning coherence body attributes, structural texture attributes, and local variance attributes to the grid stratigraphic body; A normalization unit for performing normalization processing on the data of the grid stratigraphic body; A deep learning model training unit for converting the normalized grid stratigraphic body data into seismic volume data, and obtaining training seed data based on the seismic volume data to train a deep learning model for predicting volcanic channel characteristics; A prediction unit for applying the trained deep learning model, and combining the horizontal slice of the seismic volume to predict the spatial position and morphology of the volcanic channel geological body.

6. The device according to claim 5, characterized in that, in the deep learning model training unit, apply the formation slicing technology to convert the normalized grid stratigraphic body data into seismic volume data.

7. The device according to claim 5, characterized in that, In the deep learning model training unit, obtaining training seed data based on seismic volume data includes: Tracking geological body seed points every 20 main survey lines on the seismic volume data to obtain the training seed data.

8. The apparatus according to claim 5, wherein, the apparatus further comprises: a model filtering unit, configured to add a random volcanic channel to the seismic physical simulation model for dynamic simulation demonstration, and filter the trained deep learning model.

9. An electronic device, wherein, the electronic device comprises: a memory storing executable instructions; a processor, the processor running the executable instructions in the memory to implement the method according to any one of claims 1-4.

10. A computer-readable storage medium storing a computer program, which when executed by a processor implements the method according to any one of claims 1-4.

Images