Carbonate fracture-cavity prediction model training method, prediction method and device

Abstract

The invention discloses a carbonate rock fracture-cavity prediction model training method, a carbonate rock fracture-cavity prediction method and a carbonate rock fracture-cavity prediction device. The method comprises the following steps: extracting profile data in pre-obtained seismic data of a marked fracture-cavity body to obtain multiple groups of profile data; label data corresponding to the multiple groups of profile data and corresponding standardized profile data are determined according to a preset mode; dividing the multiple groups of label data and the corresponding multiple groups of standardized profile data according to a preset proportion to obtain a training set and a verification set; inputting the training set into a pre-constructed prediction model for training, and obtaining a plurality of feature maps with different resolutions through the pre-trained large model; the pre-constructed prediction model comprises a pre-trained large model and a decoding network; inputting the plurality of feature maps with different resolutions into a decoding network to obtain the position of the fracture-cavity body, a numerical prediction result and a trained prediction model; according to the prediction precision, updating the trained prediction model; and repeating the training process until a trained prediction model meeting a preset condition is obtained.

Description

Technical Field

[0001] The invention relates to a carbonate rock fracture-cavity prediction model training method, a prediction method and a device. Background Art

[0002] Carbonate fracture-cave reservoirs are one of the most important oil and gas exploration and development areas in the world. Carbonate fracture-cave reservoirs in western China account for about 2 / 3 of the proven reserves and are one of the main areas for increasing oil reserves and production. Carbonate fracture-cave reservoirs are buried deep, generally exceeding 5,000 meters, and are mainly fractures and caves. The reservoirs have strong vertical and horizontal heterogeneity, rapid lateral changes, and complex fracture-cave morphology and fluid relationships, which bring great difficulties to seismic reservoir prediction.

[0003] At present, carbonate fracture-cave reservoirs are mainly predicted through professional software. First, synthetic seismic record calibration is required to establish the connection between depth domain well data and time domain seismic data. Through well-seismic calibration matching, the development position and reflection characteristics of the target reservoir on the seismic data are determined; then, combined with the seismic interpretation layer, the seismic attribute extraction for carbonate fracture-cave reservoirs is carried out, and the reflection of carbonate fracture-cave reservoirs is carved through a variety of sensitive seismic attributes combined with actual drilling information; on this basis, inversion work is carried out, and the wave impedance inversion data body of the target area is obtained through the construction of wave impedance low-frequency model, inversion parameter selection and wave impedance inversion. The inversion data body is carved through the lower limit of the wave impedance of the effective reservoir in actual drilling, and finally the prediction results of carbonate fracture-cave bodies are obtained.

[0004] The prediction of carbonate fractures and caves is currently done manually with software. The prediction process involves many steps, and the quality control of each process is greatly affected by personal experience. At the same time, due to too many processes, the errors in each step gradually accumulate, which will eventually have a certain impact on the quality of the results. Although various commercial software provide various software modules for carbonate fracture prediction, manual participation in the entire process is still required, which is a large workload. The workload of reservoir prediction accounts for about 70% of the entire exploration target evaluation work. Summary of the invention

[0005] In order to better predict carbonate rock fracture-cavity bodies, an embodiment of the present invention provides a carbonate rock fracture-cavity body prediction model training method, a prediction method and a device.

[0006] In a first aspect, an embodiment of the present invention provides a carbonate rock fracture-cavity prediction model training method, the method comprising:

[0007] Extracting profile data from pre-obtained seismic data of annotated fracture-cavity bodies to obtain multiple groups of profile data;

[0008] Determine the label data corresponding to the plurality of groups of profile data and the corresponding standardized profile data according to a preset method;

[0009] Dividing the multiple sets of label data and the corresponding multiple sets of standardized profile data according to a preset ratio to obtain a training set and a validation set;

[0010] Input the training set into a pre-built prediction model for training, and obtain a plurality of feature maps of different resolutions through the pre-trained large model; the pre-built prediction model includes a pre-trained large model and a decoding network;

[0011] Inputting the plurality of feature maps of different resolutions into the decoding network to obtain the position and numerical prediction results of the fracture body and the trained prediction model;

[0012] updating the trained prediction model according to the position of the fracture-cavity body and the prediction accuracy of the numerical prediction result;

[0013] Repeat the above training process until a trained prediction model that meets the preset conditions is obtained.

[0014] In an optional implementation of the embodiment of the present invention, the extracting of profile data from the pre-obtained seismic data for annotating fracture-cavity bodies to obtain multiple groups of profile data includes:

[0015] Taking the fracture-cavity body as the center, profile data are extracted along different inline, crossline and tangent directions of the pre-obtained seismic data marking the fracture-cavity body to obtain multiple groups of profile data.

[0016] In an optional implementation manner of the embodiment of the present invention, determining the label data corresponding to the multiple groups of profile data and the corresponding standardized profile data in a preset manner includes:

[0017] According to the multiple groups of profile data, the corresponding fracture-cavity position labels are constructed according to the values ​​at the fracture-cavity being 1 and the values ​​at other positions being 0;

[0018] According to the multiple groups of profile data, normalization is performed according to the maximum value or the minimum value in the profile data to construct corresponding fracture-cavity body numerical regression labels;

[0019] Constructing label data corresponding to the multiple groups of profile data according to the fracture-cavity body position labels and the fracture-cavity body numerical regression labels;

[0020] The plurality of groups of profile data are standardized according to the pre-obtained mean and standard deviation of the seismic trace amplitude to obtain corresponding standardized profile data.

[0021] In an optional implementation of the embodiment of the present invention, after obtaining the trained prediction model, the method further includes:

[0022] Inputting the verification set into the trained prediction model for verification, and obtaining the position and numerical prediction results of the fracture-cavity body;

[0023] The trained prediction model is updated according to the position of the fracture body and the prediction accuracy of the numerical prediction result.

[0024] In an optional implementation of the embodiment of the present invention, the pre-trained large model is obtained by the following method:

[0025] Get labeled earthquake data;

[0026] Extracting profile data from the labeled seismic data to obtain multiple groups of profile data;

[0027] Randomly selecting at least one preset method to mask the multiple groups of profile data to construct multiple groups of mask profile samples;

[0028] Based on self-supervised learning, the plurality of groups of mask profile samples are respectively input into a plurality of pre-built computer vision algorithm models for training to obtain a plurality of base models;

[0029] The multiple base models are fused based on the attention mechanism to obtain the pre-trained large model.

[0030] In a second aspect, an embodiment of the present invention provides a carbonate rock fracture-cavity prediction method, the method comprising:

[0031] Standardize the amplitude of the pre-acquired seismic data of the target work area to obtain standardized seismic data;

[0032] The standardized seismic data are input into the trained prediction model to obtain the location and numerical prediction results of the fracture-cavity body.

[0033] In an optional implementation of the embodiment of the present invention, after inputting the standardized seismic data into the trained prediction model for prediction and obtaining the location and numerical prediction results of the fracture-cavity body, the method further includes:

[0034] According to the numerical prediction results of the fracture-cavity body, the false detection parts in the fracture-cavity body position prediction results are filtered out.

[0035] In a third aspect, an embodiment of the present invention provides a carbonate rock fracture-cavity prediction model training device, the device comprising:

[0036] A profile extraction module is used to extract profile data from pre-obtained seismic data of annotated fracture-cavity bodies to obtain multiple groups of profile data;

[0037] A sample construction module, used to determine the label data corresponding to the plurality of groups of profile data and the corresponding standardized profile data according to a preset method;

[0038] A sample division module, used for dividing the multiple sets of label data and the corresponding multiple sets of standardized profile data according to a preset ratio to obtain a training set and a validation set;

[0039] A first training module is used to input the training set into a pre-built prediction model for training, and obtain a plurality of feature maps with different resolutions through the pre-trained large model; the pre-built prediction model includes a pre-trained large model and a decoding network;

[0040] A second training module is used to input the plurality of feature maps of different resolutions into the decoding network to obtain the position and numerical prediction results of the fracture body and the trained prediction model;

[0041] An updating module, used for updating the trained prediction model according to the position of the fracture body and the prediction accuracy of the numerical prediction result;

[0042] The model determination module is used to repeat the above training process until a trained prediction model that meets preset conditions is obtained.

[0043] In a fourth aspect, an embodiment of the present invention provides a carbonate rock fracture-cavity prediction device, the device comprising:

[0044] A data processing module is used to perform amplitude standardization on the pre-acquired seismic data of the target work area to obtain standardized seismic data;

[0045] The prediction module is used to input the standardized seismic data into the trained prediction model to obtain the location and numerical prediction results of the fracture-cavity body.

[0046] In a fifth aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the carbonate rock fracture-cave prediction model training method and / or the carbonate rock fracture-cave prediction method as described above.

[0047] In a sixth aspect, an embodiment of the present invention provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, it implements the carbonate rock fracture-cave prediction model training method and / or the carbonate rock fracture-cave prediction method as described above.

[0048] In a seventh aspect, an embodiment of the present invention provides a computer program product comprising instructions. When the computer program product is run on a computer device, the computer device executes the carbonate rock fracture-cave prediction model training method and / or the carbonate rock fracture-cave prediction method as described above.

[0049] In an eighth aspect, an embodiment of the present invention provides a chip, the chip comprising a processor and a communication interface, the communication interface and the processor are coupled, and the processor is used to run a computer program or instruction to implement the carbonate rock fracture-cave prediction model training method as described above, and / or the carbonate rock fracture-cave prediction method.

[0050] The beneficial effects of the above technical solution provided by the embodiment of the present invention include at least:

[0051] The carbonate rock fracture-cave prediction model training method provided by the embodiment of the present invention can quickly build a prediction model by fine-tuning a small number of samples on the basis of a pre-trained seismic large model, and construct training samples by annotating the extracted seismic profile data, train the prediction model, and update the training model according to the fracture-cave prediction results. Repeat the above training process until a trained prediction model that meets the preset conditions is obtained. The prediction model trained by this method can accurately realize the fracture-cave prediction of the target work area, improve the efficiency of carbonate rock fracture-cave reservoir research, and quickly obtain the reservoir prediction results of the study area, solving the problem of fracture-cave prediction accuracy and efficiency for a large range of work areas; providing a method with strong universality and simplified prediction process; reducing the multi-solution of prediction results, accelerating the efficiency of well site demonstration in the target area, and greatly improving the efficiency of carbonate rock fracture-cave prediction, effectively supporting the deployment of exploration wells. It also changes the existing human-computer interactive fracture-cave prediction work state, allowing reservoir prediction researchers to get rid of the time-consuming and laborious carbonate rock fracture-cave carving and inversion prediction work.

[0052] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings.

[0053] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0055] Figure 1 A schematic diagram of the steps of a carbonate rock fracture-cavity prediction model training method provided by an embodiment of the present invention;

[0056] Figure 2 A schematic diagram for comparing the results of marking the reflection boundaries of fracture-cavity bodies on the original seismic data of a certain work area provided by an embodiment of the present invention;

[0057] Figure 3 A schematic diagram for comparing the results of regressing wave impedance numerical annotation of the original seismic data of a certain work area provided by an embodiment of the present invention;

[0058] Figure 4 A schematic diagram of the structure of a unet neural network provided by an embodiment of the present invention;

[0059] Figure 5 A schematic diagram of the structure of an FCN neural network provided by an embodiment of the present invention;

[0060] Figure 6 A schematic diagram of the structure of a Deeplab neural network provided in an embodiment of the present invention;

[0061] Figure 7 A schematic diagram of a process for constructing a self-supervised learning task provided by an embodiment of the present invention;

[0062] Figure 8 A schematic diagram of the structure of a decoding network in a prediction model provided by an embodiment of the present invention;

[0063] Fig. 9 A schematic diagram of the structure of an upsampling module provided in an embodiment of the present invention;

[0064] Fig.10 A training flow chart of a prediction model provided by an embodiment of the present invention;

[0065] Fig.11 A schematic diagram of the steps of a carbonate rock fracture-cavity prediction method provided by an embodiment of the present invention;

[0066] Fig.12 A schematic diagram of a section of original seismic data of an oil field provided by an embodiment of the present invention;

[0067] Fig.13 A schematic diagram of a fracture-cavity body reflection boundary prediction result and a regression wave impedance prediction result obtained by predicting the original seismic data of a certain oil field provided in an embodiment of the present invention;

[0068] Fig.14 A schematic diagram of an original seismic data profile and a predicted fracture-cavity body reflection boundary position in a certain oil field provided by an embodiment of the present invention;

[0069] Fig.15 A schematic diagram of the structure of a carbonate rock fracture-cavity prediction model training device provided by an embodiment of the present invention;

[0070] Fig.16 A schematic diagram of the structure of a carbonate rock fracture-cavity prediction device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0071] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present invention. However, it should be clear to those skilled in the art that the present invention may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present invention.

[0072] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0073] It should also be understood that the term "and / or" used in the present description and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0074] As used in the present specification and the appended claims, the term "if" may be interpreted as "when" or "upon" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" may be interpreted as meaning "upon determination" or "in response to determining" or "upon detection of [described condition or event]" or "in response to detecting [described condition or event]", depending on the context.

[0075] In addition, in the description of the present specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0076] References to "one embodiment" or "some embodiments" etc. described in the present specification mean that one or more embodiments of the present invention include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0077] It should be understood that the order of execution of the steps in the following embodiments does not imply a precedence of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0078] In order to illustrate the technical solution of the present invention, specific embodiments are provided below for illustration.

[0079] The inventors have found that in the prior art, the prediction of carbonate fractures and caves mainly relies on software and is mainly manual, which results in a huge workload. At present, the prediction of ultra-deep carbonate fracture-cave oil reservoirs is carried out by inversion and attribute carving methods, and a variety of attributes that are sensitive to fracture-cave reservoirs are used to identify and carve the reflections of abnormal fractures and caves. The process of this method is relatively complicated, and the selected attributes vary from person to person, resulting in a certain deviation in the prediction results and a certain degree of uncertainty. At the same time, the workload is huge, and there are problems such as time-consuming work, insufficient personnel, and the accuracy of fracture-cave prediction being greatly affected by human experience. The existing carbonate fracture-cave prediction technology currently needs to be further improved in terms of accuracy and work efficiency, improve prediction efficiency, reduce prediction work nodes and processes, and achieve fast and accurate reservoir prediction. Based on this, the inventors have made the present invention after further research and development, providing a carbonate fracture-cave prediction model training method, prediction method and device.

[0080] Embodiment 1

[0081] The present invention provides a carbonate rock fracture-cavity prediction model training method, referring to Figure 1 As shown, the method includes:

[0082] S101: extracting profile data from pre-obtained seismic data of annotated fracture-cavity bodies to obtain multiple groups of profile data.

[0083] In the embodiment of the present invention, the original seismic data acquired through field exploration are collected, and the fracture-cavity bodies of the original seismic data are interpreted and annotated to prepare for subsequent sample construction.

[0084] In a specific embodiment, the original seismic data is annotated with the fracture-cavity reflection boundary. Figure 2 As shown, Figure 2 From left to right, the first and third images are the original seismic data profiles, and the second and fourth images are the annotated fracture-cavity reflection boundary results. The results of regressing the original seismic data and annotating the wave impedance numerical values ​​are shown in Figure 1. Figure 3 As shown, Figure 3 From left to right, the 1st and 3rd images are the original seismic data profiles, and the 2nd and 4th images are the regression wave impedance numerical results annotated by experts.

[0085] Preprocess the seismic data, including: segy seismic data analysis, segy seismic data effective range interception, format conversion and data normalization. Since seismic data is generally saved in segy format, it is necessary to parse the seismic data before using it. According to the header information in the segy file, define the byte number of Inline and cdp, and extract the seismic trace data within the control layer range for subsequent sample construction.

[0086] In the above step S101, the section data in the pre-obtained seismic data of the annotated fracture-cavity body is extracted to obtain multiple groups of section data, specifically including:

[0087] Taking the fracture-cavity body as the center, profile data are extracted along different inline, crossline and tangent directions of the pre-obtained seismic data marking the fracture-cavity body to obtain multiple groups of profile data.

[0088] In an embodiment of the present invention, after obtaining the preprocessed seismic data of the annotated fracture-cavity body, it is necessary to extract the profile data. Taking the fracture-cavity body position as the center, the seismic data profile is extracted along multiple directions such as inline, crossline and arbitrary line to obtain multiple groups of profile data, wherein each group of profile data has 128 sampling points in the horizontal and vertical directions respectively, and each sampling point has a numerical value.

[0089] S102: Determine label data corresponding to the multiple groups of profile data and corresponding standardized profile data according to a preset method.

[0090] In the above step S102, determining the label data corresponding to the plurality of groups of profile data and the corresponding standardized profile data according to a preset method specifically includes:

[0091] According to the multiple groups of profile data, the corresponding fracture-cavity position labels are constructed according to the values ​​at the fracture-cavity being 1 and the values ​​at other positions being 0;

[0092] According to the multiple groups of profile data, normalization is performed according to the maximum value or the minimum value in the profile data to construct corresponding fracture-cavity body numerical regression labels;

[0093] Constructing label data corresponding to the multiple groups of profile data according to the fracture-cavity body position labels and the fracture-cavity body numerical regression labels;

[0094] The plurality of groups of profile data are standardized according to the pre-obtained mean and standard deviation of the seismic trace amplitude to obtain corresponding standardized profile data.

[0095] In an embodiment of the present invention, for each group of profile data, label data corresponding to the profile group is obtained. The label data has the same size as the profile data, and the label data is divided into two parts: a fracture body position prediction label and a fracture body numerical regression label. The corresponding label is constructed in a preset manner, thereby obtaining label data corresponding to each group of profile data. Among them, the preset method for constructing the fracture body position prediction label can be based on the fracture body marked in the profile data, the value at the fracture body is recorded as 1, and the values ​​of other positions are recorded as 0; the preset method for constructing the fracture body numerical regression label can be normalized according to the maximum or minimum value in the numerical range statistically calculated in the profile data, so that the value changes continuously from 0 to 1, and the value in the invalid range is set to -1 and does not participate in the model training.

[0096] The label data obtained in the above way can be divided into two categories: one is the reflection boundary of the fracture-cavity reservoir, that is, the fracture-cavity position label; the other is the specific wave impedance value of the fracture-cavity body, that is, the fracture-cavity body numerical regression label.

[0097] When processing multiple sets of profile data, the extracted slice data should also be standardized according to the mean and standard deviation of the seismic trace amplitude.

[0098] The label data processed by the above step S102 is a 128×128 data matrix, in which the data consists of 0 and 1; the seismic profile data corresponding to the label data is a 128×128 data matrix, in which the data consists of standardized seismic amplitude values.

[0099] S103: Divide the multiple sets of label data and the corresponding multiple sets of standardized profile data according to a preset ratio to obtain a training set and a validation set.

[0100] In the embodiment of the present invention, the multiple sets of label data obtained in step S102 and the corresponding multiple sets of standardized profile data are divided into training sets and validation sets according to a preset ratio. The preset ratio is set according to the actual situation, for example, it can be set to 80%:20% for training set: validation set, and a test set can be further divided to perform a final evaluation of the model performance.

[0101] S104: Input the training set into a pre-built prediction model for training, and obtain a plurality of feature maps of different resolutions through the pre-trained large model; the pre-built prediction model includes a pre-trained large model and a decoding network.

[0102] In the embodiment of the present invention, the prediction model uses a multi-task deep convolutional neural network based on the UNet structure. The neural network consists of two parts: an encoding network (i.e., a pre-trained large model) and a decoding network. The encoding part is a multi-resolution encoding network based on a pre-trained large model, which is composed of a self-supervised pre-trained large model based on seismic data, and extracts multi-resolution high-dimensional semantic features of seismic data through the pre-trained network. Among them, the structure of the encoding part is variable and determined by the pre-trained large model.

[0103] In the embodiment of the present invention, the pre-trained large model can be obtained by the following method:

[0104] Get labeled earthquake data;

[0105] Extracting profile data from the labeled seismic data to obtain multiple groups of profile data;

[0106] Randomly selecting at least one preset method to mask the multiple groups of profile data to construct multiple groups of mask profile samples;

[0107] Based on self-supervised learning, the plurality of groups of mask profile samples are respectively input into a plurality of pre-built computer vision algorithm models for training to obtain a plurality of base models;

[0108] The multiple base models are fused based on the attention mechanism to obtain the pre-trained large model.

[0109] In the embodiment of the present invention, the specific process of constructing the above-mentioned pre-trained large model can be as follows:

[0110] 1. Get labeled earthquake data.

[0111] Pre-stack seismic data and post-stack seismic data generated during seismic exploration are collected and sorted. Pre-stack seismic data mainly refers to CRP gather data, and post-stack seismic data includes attribute data such as seismic amplitude data and wave impedance. After collecting seismic data, it is necessary to pre-process the seismic data, including abnormal data cleaning, data normalization and standardization. Among them, abnormal data cleaning is specifically to perform data statistics on multiple groups of seismic data and use the normal distribution method to clean up some abnormal data; data normalization and standardization specifically map the clear seismic data to a preset range, so that the processed seismic data can be better analyzed and model trained.

[0112] In addition to preprocessing the collected seismic data in the above manner, the seismic data can also be denoised by filtering the seismic data through low-pass filtering, median filtering and high-pass filtering to eliminate noise; after converting the signal to the frequency domain through Fourier transform, the frequency component can be adjusted to achieve the purpose of denoising.

[0113] After preprocessing the seismic data, various statistical methods and machine learning techniques are used to extract and analyze the features of the processed data. These features include the magnitude of the seismic amplitude, the magnitude of the seismic wave impedance, the shape of the seismic waveform, and the seismic amplitude frequency, which help to understand and describe the content of the seismic data.

[0114] According to the results of the feature analysis of different dimensions of the pre-processed seismic data, the inventors chose to use Unet, Fully Convolution Network (FCN) and deeplab computer vision algorithms to build corresponding base models to extract corresponding features, wherein Unet, FCN and deeplab neural network algorithms are all algorithms suitable for computer vision. In the architecture of Unet, FCN and deeplab neural networks, multiple branches and multiple convolutional layers are used to extract features of different scales, and shallow features and deep features are connected through jump connections, so that the network can use low-level features and high-level features for prediction at the same time, effectively improving the expression ability of the model.

[0115] In a specific embodiment, the network structure of the Unet neural network can be as follows: Figure 4 As shown, the network structure of the FCN neural network can be as follows Figure 5 As shown, the network structure of the deeplab neural network can be as follows Figure 6 As shown in the figure, the settings of each neural network hyperparameter can be set according to the actual situation.

[0116] For the preprocessed seismic data, part of the preprocessed seismic data is labeled by sampling to obtain labeled seismic data.

[0117] For prestack seismic data, the relationship between amplitude and offset (incident angle) in prestack seismic data is processed. The amplitude at the same position is inconsistent due to different offsets (incident angles). According to this change, the P-wave velocity (Vp), S-wave velocity (Vs), and density (ρ) solved by the simplified Zoeppritz equation can be used to calculate various elastic parameters such as the ratio of longitudinal and transverse wave velocities, Poisson's ratio, and Lame coefficient. Based on rock physics analysis, elastic parameters sensitive to reservoir fluids in this area are selected. At the same time, combined with geological knowledge such as structure and reservoir formation, high-quality reservoir location analysis is carried out to obtain labeled prestack seismic data.

[0118] For post-stack seismic data, the structural characteristics of the target layer and reservoir prediction are obtained through the horizontal and vertical changes of amplitude, and labeled post-stack seismic data are obtained. The main application features include the strength of the amplitude, the shape of the waveform, the lateral continuity and discontinuity of the waveform, etc.

[0119] 2. Extract the profile data from the labeled seismic data to obtain multiple groups of profile data.

[0120] The profiles are randomly extracted along different inline, crossline and tangent directions of the labeled pre-stack seismic data and / or the labeled post-stack seismic data, so as to ensure the diversity of the training data.

[0121] 3. Randomly select at least one preset method to mask multiple groups of profile data and construct multiple groups of mask profile samples.

[0122] The inventors borrowed the image processing method, used the seismic data as image data for masking, and set up a variety of preset methods, including:

[0123] (1) Modify the proportion of masked data in the profile data. Different masking ratios can be set according to actual needs, such as 25%, 50%, and 75%. By changing the masking ratio, data missing situations of different severity can be simulated to evaluate the tolerance for data missing.

[0124] (2) Divide the profile data into multiple regions, and mask the data in each region. According to actual needs, different masking blocks of different sizes can be randomly used to mask the data in the region, such as 8x8, 16x16, and 32x32, to simulate data missing in different ranges. This can better understand the impact of data missing in different ranges on subsequent processing and analysis.

[0125] (3) Select data with different values ​​in the profile data for masking. Randomly select different values ​​in the profile data and use the randomly selected different values ​​to randomly mask the values ​​in the profile data. Select according to actual needs. For example, you can fill it with 0 values ​​or Gaussian noise values. You can simulate different types of data missing situations, so as to evaluate the processing effect on different signal types.

[0126] In the process of sample construction, the above preset methods can be combined and adjusted according to actual needs to generate a series of samples under different circumstances and construct multiple sets of mask profile samples. These samples can be used to test and evaluate the performance of algorithms and models based on profile data, and can also provide references for subsequent data processing and analysis. By randomly masking the seismic profile in the way of image processing and repairing the image through computer vision algorithms, a large pre-trained model can be established.

[0127] 4. Based on self-supervised learning, multiple groups of mask profile samples are input into multiple pre-built computer vision algorithm models for training to obtain multiple base models.

[0128] The constructed multiple groups of mask profile samples are respectively input into the Unet, FCN and Deeplab neural networks for training through self-supervised learning. By constructing self-supervised learning tasks, the massive data accumulated in the field of seismic exploration are fully utilized to realize the training of unlabeled samples.

[0129] During the training process, the neural network will learn to extract useful features, and different neural networks will extract corresponding features. The Unet neural network is used to construct the semantic segmentation task of seismic data, and the semantic features of different scales in the seismic data are extracted; the FCN neural network is used to construct the target detection task of seismic data, and the target features of different scales in the seismic data are extracted; the deeplab neural network is used to construct the image restoration task of seismic data, and the image restoration features of different scales in the seismic data are extracted. Among them, semantic features refer to the semantics of the image expressed by different combinations of features in seismic data, such as strata, faults, etc.; target features refer to the target locations expressed by different combinations of features in seismic data, such as the location of strata; image restoration features refer to the feature combinations required for image restoration.

[0130] During the training process, multiple computer vision algorithm models will output multiple groups of reconstructed sections, compare the reconstructed sections with the corresponding original sections, obtain the reconstruction errors, adjust the weights of the multiple computer vision algorithm models according to the loss function, so that the reconstructed sections output next time are closer to the original sections, repeat the above process until the reconstruction errors meet the preset conditions, and obtain multiple base models corresponding to the multiple computer vision algorithm models. The preset conditions are set according to the actual situation.

[0131] In addition, during the training process, parameter adjustment and training optimization are performed for each computer vision algorithm model. The learning rate is adjusted by setting the initial learning rate, learning rate decay strategy and using an adaptive learning rate method. The complexity of the model is constrained by adding regularization terms. The adjustment of regularization parameters can be determined by cross-validation and other methods to balance the complexity and generalization ability of the model. The optimal training batch size is selected through experiments and the dynamic batch size method is used to adjust the batch size. The initialization of the model is adjusted by selecting an appropriate weight initialization method based on the network structure.

[0132] The specific implementation of the above parameter adjustment and training optimization can refer to the detailed description in the prior art and is not specifically limited here.

[0133] In a specific embodiment, the training process of each computer vision algorithm model, that is, the construction process of the self-supervised learning task, is as follows: Figure 7As shown in the figure, it is divided into two parts: training data construction and self-supervised model training. In the process of training data construction, multiple groups of profile data, i.e., multiple groups of original profiles, are obtained by extracting profiles from the preprocessed seismic data set, i.e., labeled seismic data. Then, the multiple groups of profile data are randomly masked to obtain multiple groups of masked profile samples. In the process of self-supervised model training, the multiple groups of masked profile data are respectively input into multiple computer vision algorithm models for training. The multiple computer vision algorithm models will output the model prediction results, i.e., output the reconstructed profiles. The reconstruction error loss of the reconstructed profiles and the corresponding original profiles is compared. The multiple computer vision algorithm models are updated according to the reconstruction error. The above training process is repeated until the reconstruction error meets the preset conditions, and multiple trained computer vision algorithm models, i.e., multiple base models, are obtained.

[0134] 5. Based on the attention mechanism, multiple base models are integrated to build a large pre-trained model.

[0135] After obtaining multiple base models, namely, the Unet base model, the FCN base model and the deeplab base model, we also obtained the semantic features of different scales in the seismic data extracted by the Unet base model, the target features of different scales in the seismic data extracted by the FCN base model, and the image restoration features of different scales in the seismic data extracted by the deeplab base model. The extracted features of different dimensions are fused, and the fusion layer is constructed using the attention mechanism. The features of different dimensions are weighted summed and fused to obtain a more comprehensive and accurate feature representation.

[0136] In an embodiment of the present invention, a pre-trained large model, i.e., an earthquake large model, is constructed through the above steps 1 to 5. When the training set is input into the pre-constructed prediction model for training, the sample data in the training set will first pass through the pre-trained large model part in the prediction model for feature extraction to obtain multiple feature maps of different resolutions.

[0137] S105: Input the multiple feature maps of different resolutions into the decoding network to obtain the position and numerical prediction results of the fracture body and the trained prediction model.

[0138] In the embodiment of the present invention, during the prediction model training process, multiple feature maps of different resolutions output from the pre-trained large model are input into the decoding network, and the decoding network predicts the position and value of the fracture body through the nonlinear combination of the features of each part. The structure of the decoding network is as follows: Figure 8 As shown in Figure 1, firstly, an atrous spatial pyramid pooling (ASPP) structure is used to increase the receptive field of the feature map. In the remaining structure, one part is composed of multiple cascaded upsampling modules. The structure of the upsampling module is as follows: Fig. 9As shown in the figure, it includes upsampling layer, convolution layer, normalization layer and activation layer, which contains bilinear interpolation and residual structure, the convolution kernel size is 3×3, and the activation function is GELU function; one part is the prediction end, including a 3×3 convolution module, two 1×1 convolution layers and sigmoid activation layer. The convolution kernel size of 1×1 and sigmoid activation layer are used in the prediction of the fracture position, and only one convolution kernel size of 1×1 is used for the fracture numerical regression. Through the decoding network, the semantic features are decoded into the spatial position of the fracture in the data and the corresponding wave impedance value.

[0139] In the embodiment of the present invention, Fig.10 As shown in the figure, the training process of the prediction model is that the training set constructed by seismic data slices, i.e., seismic profile data, is input into the prediction model, feature extraction is performed through the pre-trained large model, and then the fracture-cavity prediction is performed through the decoding network to obtain the fracture-cavity prediction result.

[0140] S106: updating the trained prediction model according to the position of the fracture body and the prediction accuracy of the numerical prediction result.

[0141] In an embodiment of the present invention, the model training process is implemented by fitting the model output and annotation on the training set, and the validation set data is used to evaluate the generalization performance of the model on unknown data, and the model hyperparameters are optimized.

[0142] After the training set is input into the prediction model for training, the position and numerical prediction results of the fracture body and the trained prediction model are obtained, the position and numerical prediction results of the fracture body are compared with the actual results, the prediction accuracy of the trained prediction model is determined, and the trained prediction model is updated.

[0143] In the embodiment of the present invention, after obtaining the trained prediction model, the method further includes:

[0144] Inputting the verification set into the trained prediction model for verification, and obtaining the position and numerical prediction results of the fracture-cavity body;

[0145] The trained prediction model is updated according to the position of the fracture body and the prediction accuracy of the numerical prediction result.

[0146] In the embodiment of the present invention, the prediction model is continuously iteratively trained using a training set and a validation set. After each training, the trained model is evaluated for model accuracy on the validation set, and appropriate model hyperparameters are selected for iterative training based on the prediction accuracy. The above training process is repeated until a trained prediction model that meets the preset conditions is obtained. After the model training converges, the test set can also be used to perform a final evaluation of the model performance. By performing model prediction on the test set and comparing it with the labeled data, the prediction accuracy and generalization performance of the model can be quantitatively evaluated.

[0147] S107: Obtain a trained prediction model that meets preset conditions.

[0148] The carbonate fracture-cave prediction model training method provided by the embodiment of the present invention can quickly build a prediction model by fine-tuning a small number of samples on the basis of a pre-trained seismic large model, and construct training samples by annotating the extracted seismic profile data, train the prediction model, and update the training model according to the fracture-cave prediction results. Repeat the above training process until a trained prediction model that meets the preset conditions is obtained. The prediction model trained by this method can accurately realize the fracture-cave prediction of the target work area, improve the efficiency of carbonate fracture-cave reservoir research, quickly obtain the reservoir prediction results of the study area, and solve the problems of fracture-cave prediction accuracy and efficiency for a large range of work areas; provide a method with strong universality and simplified prediction process; reduce the multi-solution of prediction results, speed up the efficiency of well location demonstration in the target area, and greatly improve the efficiency of carbonate fracture-cave prediction, effectively supporting the deployment of exploration wells. Compared with conventional prediction, the prediction is transformed from full manual prediction to full intelligent prediction. It also changes the existing human-computer interactive fracture-cave prediction working state, allowing reservoir prediction researchers to get rid of the time-consuming and laborious carbonate fracture-cave carving and inversion prediction work.

[0149] Embodiment 2

[0150] Based on the same inventive concept, the embodiment of the present invention also provides a carbonate rock fracture cave prediction method, referring to Fig.11 As shown, the method includes:

[0151] S201: performing amplitude standardization on the pre-acquired seismic data of the target work area to obtain standardized seismic data;

[0152] S202: Input the standardized seismic data into the trained prediction model to obtain the location of the fracture-cavity body and the numerical prediction result.

[0153] In the embodiment of the present invention, after inputting the standardized seismic data into the trained prediction model for prediction and obtaining the position and numerical prediction results of the fracture-cavity body, the method further includes:

[0154] According to the numerical prediction results of the fracture-cavity body, the false detection parts in the fracture-cavity body position prediction results are filtered out.

[0155] In the embodiment of the present invention, the specific training method of the prediction model is shown in Embodiment 1 and will not be described in detail here.

[0156] For the unlabeled seismic data of the target work area, the trained prediction model is used to predict the fracture body of the amplitude-normalized seismic data, and the numerical prediction results of the fracture body and the fracture body position prediction results are obtained. The position and numerical results at the output end are combined to obtain the final fracture body prediction. Among them, the numerical prediction is the regression wave impedance prediction, and the position prediction is the fracture reflection boundary prediction. The two are related to each other and can be verified with each other to improve the accuracy. For example, according to the numerical prediction results of the fracture body, the false detection part in the fracture body position prediction results can be filtered.

[0157] In a specific embodiment, the trained prediction model is applied to the identification process of reservoir fractures and cavities in an oil field. Fig. 9 The original seismic data shown in the figure is input into the trained prediction model after amplitude normalization to obtain the position of the fracture-cavity body and the numerical prediction results, that is, the fracture-cavity body reflection boundary prediction results and the regression wave impedance prediction results as shown in Fig.10 shown.

[0158] The fracture-cavity prediction results are overlapped with the original seismic data, such as Fig.11 As shown in the figure, the background part is the original seismic data, and the superposition range is the predicted reflection boundary position of the fracture-cavity body.

[0159] Depend on Fig.11 It can be seen that the overall results obtained through the trained prediction model are consistent with the expert's explanation by 92.4%, which meets the business application needs, and the timeliness of intelligent recognition is significantly improved.

[0160] Through the above experiments, it can be found that by training the prediction model constructed based on the large seismic model, the trained prediction model is obtained for the prediction of fracture-cavity bodies. Under the premise of not reducing or even improving the interpretation quality, compared with the conventional fracture-cavity prediction model in the prior art, the number of samples required for training is reduced by more than 8 times, and the production efficiency is increased by more than 30 times.

[0161] Embodiment 3

[0162] Based on the same inventive concept, the embodiment of the present invention also provides a carbonate rock fracture cave prediction model training device, referring to Fig.15 As shown, the device comprises:

[0163] The profile extraction module 101 is used to extract profile data from the pre-acquired seismic data of the annotated fracture-cavity body to obtain multiple groups of profile data;

[0164] The sample construction module 102 is used to determine the label data corresponding to the plurality of groups of profile data and the corresponding standardized profile data according to a preset method;

[0165] The sample division module 103 is used to divide the multiple sets of label data and the corresponding multiple sets of standardized profile data according to a preset ratio to obtain a training set and a validation set;

[0166] A first training module 104 is used to input the training set into a pre-built prediction model for training, and obtain a plurality of feature maps with different resolutions through the pre-trained large model; the pre-built prediction model includes a pre-trained large model and a decoding network;

[0167] The second training module 105 is used to input the plurality of feature maps of different resolutions into the decoding network to obtain the position and numerical prediction results of the fracture body and the trained prediction model;

[0168] An updating module 106, configured to update the trained prediction model according to the position of the fracture body and the prediction accuracy of the numerical prediction result;

[0169] The model determination module 107 is used to repeat the above training process until a trained prediction model that meets preset conditions is obtained.

[0170] Embodiment 4

[0171] Based on the same inventive concept, the embodiment of the present invention also provides a carbonate rock fracture cave prediction device, referring to Fig.16 As shown, the device comprises:

[0172] The data processing module 201 is used to perform amplitude standardization on the pre-acquired seismic data of the target work area to obtain standardized seismic data;

[0173] The prediction module 202 is used to input the standardized seismic data into the trained prediction model to obtain the location of the fracture-cavity body and the numerical prediction result.

[0174] Embodiment 5

[0175] Based on the same inventive concept, an embodiment of the present invention also provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, it implements the carbonate rock fracture-cave prediction model training method described in the above embodiment one, and / or the carbonate rock fracture-cave prediction method described in the above embodiment two.

[0176] Embodiment 6

[0177] Based on the same inventive concept, an embodiment of the present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the carbonate rock fracture-vuggy body prediction model training method described in the above-mentioned embodiment 1 and / or the carbonate rock fracture-vuggy body prediction method described in the above-mentioned embodiment 2 are implemented.

[0178] Embodiment 7

[0179] Based on the same inventive concept, an embodiment of the present invention also provides a computer program product comprising instructions. When the computer program product is run on a computer device, the computer device executes the carbonate rock fracture-vuggy body prediction model training method described in the above-mentioned embodiment 1, and / or the carbonate rock fracture-vuggy body prediction method described in the above-mentioned embodiment 2.

[0180] Embodiment 8

[0181] Based on the same inventive concept, an embodiment of the present invention further provides a chip, the chip comprising a processor and a communication interface, the communication interface and the processor are coupled, and the processor is used to run a computer program or instruction to implement the carbonate rock fracture-cave prediction model training method as described in the above embodiment 1, and / or the carbonate rock fracture-cave prediction method as described in the above embodiment 2.

[0182] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) containing computer-usable program code.

[0183] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0184] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0185] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0186] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A carbonate rock fracture-cavity prediction model training method, characterized in that: include: Extracting profile data from pre-obtained seismic data of annotated fracture-cavity bodies to obtain multiple groups of profile data; Determine the label data corresponding to the plurality of groups of profile data and the corresponding standardized profile data according to a preset method; Dividing the multiple sets of label data and the corresponding multiple sets of standardized profile data according to a preset ratio to obtain a training set and a validation set; Input the training set into a pre-built prediction model for training, and obtain a plurality of feature maps of different resolutions through the pre-trained large model; the pre-built prediction model includes a pre-trained large model and a decoding network; Inputting the plurality of feature maps of different resolutions into the decoding network to obtain the position and numerical prediction results of the fracture body and the trained prediction model; updating the trained prediction model according to the position of the fracture-cavity body and the prediction accuracy of the numerical prediction result; Repeat the above training process until a trained prediction model that meets the preset conditions is obtained.

2. The method according to claim 1, characterized in that The section data is extracted from the pre-obtained seismic data of the annotated fracture-cavity body to obtain multiple groups of section data, including: Taking the fracture-cavity body as the center, profile data are extracted along different inline, crossline and tangent directions of the pre-obtained seismic data marking the fracture-cavity body to obtain multiple groups of profile data.

3. The method according to claim 1, characterized in that The step of determining the label data corresponding to the plurality of groups of profile data and the corresponding standardized profile data according to a preset manner includes: According to the multiple groups of profile data, the corresponding fracture-cavity position labels are constructed according to the values ​​at the fracture-cavity being 1 and the values ​​at other positions being 0; According to the multiple groups of profile data, normalization is performed according to the maximum value or the minimum value in the profile data to construct corresponding fracture-cavity body numerical regression labels; Constructing label data corresponding to the multiple groups of profile data according to the fracture-cavity body position labels and the fracture-cavity body numerical regression labels; The plurality of groups of profile data are standardized according to the pre-obtained mean and standard deviation of the seismic trace amplitude to obtain corresponding standardized profile data.

4. The method according to claim 1, characterized in that After obtaining the trained prediction model, it also includes: Inputting the verification set into the trained prediction model for verification, and obtaining the position and numerical prediction results of the fracture-cavity body; The trained prediction model is updated according to the position of the fracture body and the prediction accuracy of the numerical prediction result.

5. The method according to claim 1, characterized in that The pre-trained large model is obtained by the following method: Get labeled earthquake data; Extracting profile data from the labeled seismic data to obtain multiple groups of profile data; Randomly selecting at least one preset method to mask the multiple groups of profile data to construct multiple groups of mask profile samples; Based on self-supervised learning, the plurality of groups of mask profile samples are respectively input into a plurality of pre-built computer vision algorithm models for training to obtain a plurality of base models; The multiple base models are fused based on the attention mechanism to obtain the pre-trained large model.

6. A carbonate rock fracture-cavity prediction method, characterized in that: include: Standardize the amplitude of the pre-acquired seismic data of the target work area to obtain standardized seismic data; The standardized seismic data are input into the trained prediction model to obtain the location and numerical prediction results of the fracture-cavity body.

7. The method according to claim 6, characterized in that After inputting the standardized seismic data into the trained prediction model for prediction and obtaining the location of the fracture-cavity body and the numerical prediction result, the method further includes: According to the numerical prediction results of the fracture-cavity body, the false detection parts in the fracture-cavity body position prediction results are filtered out.

8. A carbonate rock fracture-cavity prediction model training device, characterized in that: include: A profile extraction module is used to extract profile data from pre-obtained seismic data of annotated fracture-cavity bodies to obtain multiple groups of profile data; A sample construction module, used to determine the label data corresponding to the plurality of groups of profile data and the corresponding standardized profile data according to a preset method; A sample division module, used for dividing the multiple sets of label data and the corresponding multiple sets of standardized profile data according to a preset ratio to obtain a training set and a validation set; A first training module is used to input the training set into a pre-built prediction model for training, and obtain a plurality of feature maps with different resolutions through the pre-trained large model; The pre-built prediction model includes a pre-trained large model and a decoding network; A second training module is used to input the plurality of feature maps of different resolutions into the decoding network to obtain the position and numerical prediction results of the fracture body and the trained prediction model; An updating module, used for updating the trained prediction model according to the position of the fracture body and the prediction accuracy of the numerical prediction result; The model determination module is used to repeat the above training process until a trained prediction model that meets preset conditions is obtained.

9. A carbonate rock fracture-cavity prediction device, characterized in that: include: A data processing module is used to perform amplitude standardization on the pre-acquired seismic data of the target work area to obtain standardized seismic data; The prediction module is used to input the standardized seismic data into the trained prediction model to obtain the location and numerical prediction results of the fracture-cavity body.

10. A computer-readable storage medium, in which instructions are stored. When the instructions are executed on a terminal, the terminal executes the carbonate rock fracture-vuggy body prediction model training method as described in any one of claims 1 to 5, and / or the carbonate rock fracture-vuggy body prediction method as described in any one of claims 6 to 7.

11. A computer device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method for training a carbonate rock fracture-vuggy body prediction model as described in any one of claims 1 to 5 and / or the method for predicting a carbonate rock fracture-vuggy body as described in any one of claims 6 to 7 is implemented.

12. A computer program product comprising instructions, which, when executed on a computer device, enables the computer device to execute the carbonate rock fracture-vuggy body prediction model training method as described in any one of claims 1 to 5, and / or the carbonate rock fracture-vuggy body prediction method as described in any one of claims 6 to 7.

13. A chip, comprising a processor and a communication interface, wherein the communication interface is coupled to the processor, and the processor is used to run a computer program or instruction to implement the carbonate rock fracture-cavity prediction model training method as described in any one of claims 1 to 5, and / or the carbonate rock fracture-cavity prediction method as described in any one of claims 6 to 7.

Images