Seismic attribute prediction method and device

Abstract

The invention discloses a seismic attribute prediction method and device, and the method comprises the steps: obtaining seismic original data, and marking the seismic original data; performing data preprocessing on the labeled seismic original data, and determining a processing result; constructing an initial earthquake prediction model, training the initial earthquake prediction model by using a processing result, and determining a target earthquake prediction model; and obtaining target block seismic data, inputting the target block seismic data into the target seismic prediction model, and determining a prediction result. According to the method, the prediction model is constructed, so that the attribute prediction of the seismic data is realized, the seismic attributes of a plurality of blocks are prevented from having certain similarity and difference in the seismic attribute analysis work, and the seismic data of the blocks are respectively learned, so that the accuracy of seismic attribute prediction is ensured.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of geophysical exploration seismic interpretation and integration, and particularly relates to a seismic attribute prediction method and device. BACKGROUND

[0002] Seismic attributes are information reflecting the geometry, kinematics, dynamics and statistics of seismic waves obtained by mathematical calculation or mathematical transformation on pre-stack or post-stack seismic data. There are currently more than one hundred seismic attributes that can be extracted from seismic data, among which commonly used seismic attributes include amplitude, frequency, waveform and other attributes, which are basic elements for reservoir prediction. Seismic attributes can be used for reservoir distribution description, reservoir thickness prediction, oil and gas detection, sedimentary facies analysis, and play an important role in oilfield exploration and development.

[0003] In the current seismic attribute analysis work, the seismic attributes of multiple blocks often have certain similarities and differences. If all block seismic data is directly mixed together for learning, it is difficult to learn the common characteristics of the blocks, and the accuracy of seismic attribute prediction cannot be guaranteed. SUMMARY

[0004] Therefore, the embodiments of the present application provide a seismic attribute prediction method and device, which solve the problem that the accuracy of seismic attribute prediction cannot be guaranteed due to the mixing of all block seismic data together by the existing method.

[0005] According to a first aspect, the embodiments of the present application provide a seismic attribute prediction method, comprising:

[0006] Obtaining seismic raw data and labeling the seismic raw data;

[0007] Data preprocessing is performed on the labeled seismic raw data to determine a processing result;

[0008] An initial seismic prediction model is constructed, and the initial seismic prediction model is trained using the processing result to determine a target seismic prediction model;

[0009] Obtaining target block seismic data, inputting the target block seismic data into the target seismic prediction model, and determining a prediction result.

[0010] In combination with the first aspect, in a first implementation manner of the first aspect, the obtaining seismic raw data and labeling the seismic raw data comprises:

[0011] Obtaining a preset feature and collecting seismic raw data with the preset feature;

[0012] The seismic original data at the same position coordinate are labeled to determine seismic samples of different blocks.

[0013] With reference to the first aspect and the first implementation manner, in a second implementation manner of the first aspect, the labeled historical data is subjected to data preprocessing, and a processing result is determined, including:

[0014] The seismic samples of each different block are subjected to an operation of removing outliers, and operation data is determined;

[0015] The operation data is subjected to denoising processing, and denoising data is determined;

[0016] The denoising data is subjected to normalization processing, and an optimized processing result is determined.

[0017] With reference to the first aspect and the first implementation manner, in a third implementation manner of the first aspect, the initial seismic prediction model is constructed, including:

[0018] The seismic samples are identified and classified by using a preset network model, and an identification result and a classification label are determined;

[0019] The preset network model parameters are determined according to the identification result and the classification label, so as to construct the initial seismic prediction model.

[0020] With reference to the third implementation manner of the first aspect, in a fourth implementation manner of the first aspect, the seismic samples are identified and classified by using the preset network model, and the identification result and the classification label are determined, including:

[0021] The preset network model includes an encoder, a decoder, a discriminator, a decoupler, a block identifier, a first classifier and a second classifier;

[0022] The seismic samples are transformed by using the encoder, and seismic main features are obtained;

[0023] The seismic main features are identified and classified by using each structure of the preset network model respectively, and corresponding parameters, an identification result and a classification label are determined.

[0024] With reference to the fourth implementation manner of the first aspect, in a fifth implementation manner of the first aspect, the seismic main features are identified and classified by using each structure of the preset network model respectively, and corresponding parameters, an identification result and a classification label are determined, including:

[0025] The seismic main features are decoded by using the decoder, and reconstruction loss data is determined;

[0026] The seismic main features are discriminated by using the discriminator, and normal discrimination loss data is determined;

[0027] classify the seismic main features by using the first classifier to determine lithology classification loss data;

[0028] decompose the seismic main features by using the decoupler to determine block features and category features;

[0029] classify the block features and the category features by using the second classifier to determine a lithology category of a corresponding feature of the input;

[0030] identify the block features and the category features by using the block identifier to determine a source block of a corresponding feature of the input.

[0031] In a sixth implementation form of the first aspect, in combination with the fifth implementation form of the first aspect, training the initial seismic prediction model by using the processing result to determine a target seismic prediction model comprises:

[0032] input seismic samples of each different block and corresponding labels thereof into the initial seismic prediction model in sequence to determine training parameters;

[0033] adjust the initial seismic prediction model by using the seismic samples of each different block corresponding to the training parameters to determine a target seismic prediction model.

[0034] The seismic attribute prediction method provided by the embodiment of the present application can realize attribute prediction of seismic data by constructing a prediction model, and can ensure the accuracy of seismic attribute prediction by learning seismic data of different blocks respectively due to the similarity and difference of seismic attributes of multiple blocks in seismic attribute analysis.

[0035] According to the second aspect, the embodiment of the present application provides a seismic attribute prediction device, comprising:

[0036] a first processing module configured to acquire seismic original data and label the seismic original data;

[0037] a second processing module configured to perform data preprocessing on the labeled seismic original data to determine a processing result;

[0038] a third processing module configured to construct an initial seismic prediction model and train the initial seismic prediction model by using the processing result to determine a target seismic prediction model;

[0039] a fourth processing module configured to acquire target block seismic data, input the target block seismic data into the target seismic prediction model, and determine a prediction result.

[0040] The seismic attribute prediction device provided by the embodiment can realize attribute prediction of seismic data by constructing a prediction model, thereby avoiding the situation that the seismic attributes of multiple blocks have certain similarities and differences in seismic attribute analysis, and ensuring the accuracy of seismic attribute prediction.

[0041] According to a third aspect, an electronic device is provided, comprising a memory and a processor, which are in communication connection with each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the seismic attribute prediction method in the first aspect or any one of the implementation manners of the first aspect.

[0042] According to a fourth aspect, a computer readable storage medium is provided, which stores computer instructions for causing the computer to perform the seismic attribute prediction method in the first aspect or any one of the implementation manners of the first aspect. BRIEF DESCRIPTION OF DRAWINGS

[0043] In order to more clearly illustrate the specific embodiments of the present application or the technical solutions in the prior art, the drawings needed in the specific embodiments or the prior art description will be briefly introduced as follows. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0044] Figure 1 is a flowchart of the seismic attribute prediction method according to the embodiment of the present application;

[0045] Figure 2 is a schematic diagram of the overall network structure of the initial seismic prediction model according to the embodiment of the present application;

[0046] Figure 3 is a schematic diagram of the seismic attribute prediction device according to the preferred embodiment of the present application;

[0047] Figure 4 is a schematic diagram of the hardware structure of the electronic device provided by the embodiment of the present application. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme of the embodiments of the present application will be described clearly and completely in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are some of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0049] Embodiment One

[0050] A seismic attribute prediction method is provided in the embodiment, which can be used in electronic devices such as computers, mobile phones, tablets and the like. Figure 1 is a flowchart of the seismic attribute prediction method according to the embodiment of the present application.

[0051] As shown in Figure 1 , the flowchart includes the following steps:

[0052] S11, obtaining seismic raw data and labeling the seismic raw data. In the embodiment, the seismic raw data is obtained by using existing techniques or data acquisition devices, and is labeled according to the geographical position coordinates of the obtained seismic raw data, for subsequent data processing and data prediction. It should be noted that the means and techniques for obtaining data in the embodiment are not limited in any way, as long as data acquisition can be achieved.

[0053] S12, data preprocessing is performed on the labeled seismic raw data, and a processing result is determined. The labeled data is processed to ensure the reliability of the data and the consistency of the data distribution. Details will be described in subsequent steps, and the embodiment will not be described again.

[0054] S13, an initial seismic prediction model is constructed, and the initial seismic prediction model is trained using the processing result to determine a target seismic prediction model. The initial seismic prediction model is constructed by using an existing network model, and then the model is trained using the processed data to determine the trained target seismic prediction model. Details will be described in subsequent steps, and the embodiment will not be described again.

[0055] S14, obtaining target block seismic data, inputting the target block seismic data into the target seismic prediction model, and determining a prediction result. The trained target seismic prediction model is used to predict the target block seismic data to determine the prediction result. Details will be described in subsequent steps, and the embodiment will not be described again.

[0056] Embodiment Two

[0057] The seismic attribute prediction method provided in the embodiment realizes the prediction of the attributes of seismic data by constructing a prediction model, thereby avoiding the similarity and difference of the seismic attributes of multiple blocks in the seismic attribute analysis work, learning the seismic data of each block, and ensuring the accuracy of the prediction of the seismic attributes.

[0058] In another embodiment, a seismic attribute prediction method is also provided, and another flowchart of the seismic attribute prediction method according to the embodiment of the present application includes the following steps:

[0059] S21. Obtain the raw earthquake data and annotate it.

[0060] In this embodiment, step S21 includes the following steps:

[0061] S211, Obtain preset features, collect raw seismic data with preset features; specifically, preset features are defined geological similarity features, collect raw seismic data of multiple blocks with certain geological similarity, and extract their attributes to form feature vectors.

[0062] S212, label the raw seismic data at the same location coordinates to determine the seismic samples for different blocks.

[0063] In this embodiment, seismic attributes at the same latitude and longitude location constitute a seismic feature vector, i.e., a seismic sample; well logging data is acquired, and lithological labels are assigned to the seismic samples based on their actual location coordinates; training data <X> for different blocks are constructed. i Y i >,i=1,2,3,...,K,X i This represents the seismic sample matrix of the i-th block, where each row represents a seismic sample, Y. i X represents i The corresponding lithological labels, K being a positive integer greater than 0 representing the number of training blocks, and the target block seismic sample matrix represented as X. t .

[0064] S22, perform data preprocessing on the labeled raw seismic data and determine the processing results.

[0065] Specifically, step S22 above also includes the following steps:

[0066] S221, outlier removal is performed on seismic samples from different blocks to determine the operational data. Outlier removal from the seismic sample data ensures data availability.

[0067] S222, Denoise the operation data and determine the denoised data; to ensure the accuracy of data processing, denoise the data after removing outliers to avoid noise interference.

[0068] S223, normalize the denoised data to determine the optimized processing result. Ensure the reliability and distribution consistency of the data.

[0069] S23, construct an initial earthquake prediction model, and use the processing results to train the initial earthquake prediction model to determine the target earthquake prediction model. See step S13 for details, which will not be repeated in this embodiment.

[0070] S24, obtaining the target block seismic data, inputting the target block seismic data into the target seismic prediction model, and determining the prediction result. For details, refer to step S14, which will not be repeated in the present embodiment.

[0071] In another embodiment, a seismic attribute prediction method is also provided, and the flow includes the following steps:

[0072] S31, obtaining seismic raw data and labeling the seismic raw data. For details, refer to step S21, which will not be repeated in the present embodiment.

[0073] S32, performing data preprocessing on the labeled seismic raw data, and determining a processing result. For details, refer to step S22, which will not be repeated in the present embodiment.

[0074] S33, constructing an initial seismic prediction model, and training the initial seismic prediction model by using the processing result, to determine a target seismic prediction model.

[0075] Specifically, the above step S33 further includes the following steps:

[0076] S331, identifying and classifying the seismic sample by using a preset network model, to determine an identification result and a classification label; the preset network model is an existing network model structure, and the present embodiment does not limit the specific network model structure.

[0077] In the present embodiment, as shown in FIG. 3, the overall network of the initial seismic prediction model includes an encoder En, a decoder De, a discriminator Dis, a decoupler Sen, a block identifier Cd, a distribution difference measurement module, and two classifiers C1 and C2; wherein the encoder is used to transform the seismic sample to obtain a seismic main feature; then each structure of the preset network model is used to identify and classify the seismic main feature, to determine the corresponding parameters, the identification result and the classification label, and the specific implementation process is as follows: Figure 2

[0078] (1) The encoder En can transform the seismic sample by using a fully connected network to obtain a seismic main feature f m , and the specific calculation process is as follows,

[0079] f m = En(x)

[0080] wherein x e X1 U... U X k represents a seismic sample, and f m represents the feature output after the sample is processed by the encoder, i.e., the seismic main feature.

[0081] ​(2) Decoder De can be constructed by a fully connected network. In the process of decoding, we expect to minimize the reconstruction error, so that the encoded seismic principal features can retain more information. The specific calculation method is as follows,

[0082]

[0083] wherein, represents the data restored by the decoder; the mean square error loss (MSE) is used to make and x consistent, and the specific calculation method is as follows,

[0084]

[0085] wherein L r represents the reconstruction loss, and N represents the total number of samples.

[0086] (3) Discriminator Dis is used to distinguish the data features based on the prior sampling and the seismic samples, so that the seismic principal feature distribution tends to the preset prior distribution. The specific calculation method is as follows,

[0087]

[0088] wherein the standard normal distribution is used as the prior distribution, and L n represents the normal discriminant loss, m represents the number of samples in a batch, and the seismic principal feature samples f m output by the real seismic samples in a batch after being transformed by the encoder also contain pseudo samples randomly sampled from the standard normal distribution, and the number of the two is the same; d i represents the domain label of the i-th sample in the batch, and the d i of the pseudo sample is 0, and the d i of the real seismic sample is 1; represents the output of the i-th sample in the batch through Dis.

[0089] (4) Classifier C1 is a conventional classifier, such as support vector machine, extreme learning machine, fully connected network, etc.; the lithology category of the seismic sample is judged according to the seismic principal feature f m , and the lithology classification loss is calculated as follows:

[0090]

[0091] wherein L y represents the lithology classification loss, N represents the total number of samples, and P x→k represents the probability that x i belongs to the k-th class, k = 1, 2, 3,..., c, and c represents the total number of lithology categories.

[0092] (5) The decoupler Sen decomposes the seismic main feature f m into block feature f d and class feature f c . The first half of the feature vector output by Sen is f d and the second half is f c . The block feature is specific to the block, while the class feature represents the feature specific to the lithology class. To enhance the decoupling effect, an independence loss L d is added between the block feature f c and the class feature f m . The specific calculation is as follows,

[0093]

[0094] where L m represents the independence loss, represents the joint probability distribution of (f d , f c ), represent the marginal probability distribution of f d and f c , respectively.

[0095] (6) The classifier C2 has the same structure as C1, which is a conventional label classifier model, which can be an SVM or a fully connected network. C2 is used to classify the block feature f m and the class feature f d obtained after decomposition of f c , determine the lithology class of the input feature, and remove the specific block information from f c through cross-antagonistic training. The training related to f c is subject to two constraints, one is the lithology classification loss of C2:

[0096]

[0097] where f c is transformed from x, and the other is the block classification certainty:

[0098]

[0099] (7) The block recognizer Cd is used to recognize the block feature f d and the class feature f c obtained after decomposition, determine which block the input feature comes from, and remove the specific block information from f d through cross-antagonistic training. Cd is also essentially a classifier. The training related to f d is subject to two constraints, one is the block classification loss:

[0100]

[0101] Another is the lithology classification certainty:

[0102]

[0103] (8) The distribution difference measurement module calculates the data distribution difference between multiple domains by multi-domain maximum mean difference (MD-MMD). The specific calculation result is as follows,

[0104]

[0105] wherein, L d represents the distribution difference loss, K represents the total number of blocks, respectively represent the seismic main attribute features of the i-th and j-th block samples.

[0106] S332, determining the preset network model parameters according to the identification result and the classification label to construct an initial seismic prediction model.

[0107] S333, training the initial seismic prediction model by using the processing result to determine a target seismic prediction model.

[0108] Specifically, the seismic samples of each different block and the corresponding labels are sequentially input into the initial seismic prediction model to determine the training parameters; the initial seismic prediction model is adjusted in parameters by using the training parameters corresponding to the seismic samples of each different block to determine the target seismic prediction model.

[0109] In actual application, the seismic samples X1∪...∪X K of multiple blocks and the corresponding lithology labels Y1∪...∪Y K are sequentially input into the network for training, and the remaining networks are fixed when training a certain network:

[0110] En and De are trained to minimize L r , and the model parameters θ En of En are obtained; En and C1 are trained to minimize L c +η d L d , wherein the training of En is fine-tuning, wherein η d ∈(0.1, 1) is a balance parameter; Dis is trained to minimize L n ; En is trained to maximize L n , until Nash equilibrium is reached.

[0111] Finally, En is trained to minimize L ccTrain Sen and C2 with the goal of minimizing L cd Train Sen with the goal of minimizing L dd Train Sen and Cd with the goal of minimizing L dc Train Sen with the goal of minimizing L Repeat the last training step until a prescribed preset number of times is reached.

[0112] S34, acquiring target block seismic data, inputting the target block seismic data into a target seismic prediction model, and determining a prediction result. That is, determining a seismic attribute corresponding to the target block seismic data.

[0113] The seismic attribute prediction method provided in the embodiment of the present application can realize the prediction of the seismic data attribute by constructing a prediction model, and can avoid the situation that the seismic attributes of multiple blocks have certain similarities and differences in the seismic attribute analysis work, and the seismic data of each block is learned respectively, so that the accuracy of the seismic attribute prediction is ensured.

[0114] The embodiment provides a seismic attribute prediction device. As used in the following, the term "module" can be a combination of software and / or hardware that realizes a predetermined function. Although the device described in the following embodiment is preferably realized in software, realization in hardware, or a combination of software and hardware is also possible and contemplated.

[0115] Embodiment three

[0116] The present application discloses a seismic attribute prediction device, as shown in the following formula: Figure 3 comprises:

[0117] The first processing module 01 is configured to acquire seismic original data and label the seismic original data.

[0118] The second processing module 02 is configured to perform data preprocessing on the labeled seismic original data and determine a processing result.

[0119] The third processing module 03 is configured to remove interference type feature data according to a preset rule and determine a stacked data body.

[0120] The fourth processing module 04 is configured to detect an angle domain fluid based on the stacked data body and obtain a detection result.

[0121] The seismic attribute prediction device provided in the embodiment of the present application can realize the prediction of the seismic data attribute by constructing a prediction model, and can avoid the situation that the seismic attributes of multiple blocks have certain similarities and differences in the seismic attribute analysis work, and the seismic data of each block is learned respectively, so that the accuracy of the seismic attribute prediction is ensured.

[0122] The embodiment of the present application further provides an electronic device, please refer toFigure 4 , Figure 4 This is a schematic diagram of the structure of an electronic device provided in an optional embodiment of the present invention, such as... Figure 4 As shown, the electronic device may include: at least one processor 601, such as a CPU (Central Processing Unit), at least one communication interface 603, a memory 604, and at least one communication bus 602. The communication bus 602 is used to enable communication between these components. The communication interface 603 may include a display screen or a keyboard; optionally, the communication interface 603 may also include a standard wired interface or a wireless interface. The memory 604 may be high-speed RAM (Random Access Memory) or non-volatile memory, such as at least one disk drive. Optionally, the memory 604 may also be at least one storage device located remotely from the processor 601. The processor 601 may be integrated with the aforementioned devices, the memory 604 stores application programs, and the processor 601 calls the program code stored in the memory 604 to execute any of the aforementioned method steps.

[0123] The communication bus 602 can be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The communication bus 602 can be divided into an address bus, a data bus, and a control bus, etc. For ease of representation, Figure 4 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0124] The memory 604 may include volatile memory, such as random-access memory (RAM); the memory may also include non-volatile memory, such as flash memory, hard disk drive (HDD) or solid-state drive (SSD); the memory 604 may also include a combination of the above types of memory.

[0125] The processor 601 can be a central processing unit (CPU), a network processor (NP), or a combination of the CPU and the NP.

[0126] The processor 601 can further include a hardware chip. The hardware chip can be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The PLD can be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL), or any combination thereof.

[0127] Optionally, the memory 604 is further configured to store program instructions. The processor 601 can invoke the program instructions to implement the seismic attribute prediction method as shown in the embodiments of the drawings.

[0128] The embodiments of the present application further provide a non-transitory computer storage medium, and the computer storage medium stores computer executable instructions. The computer executable instructions can execute the seismic attribute prediction method in any method embodiment described above. The storage medium can be a magnetic disc, an optical disc, a read-only memory (ROM), a random access memory (RAM), a flash memory, a hard disk drive (HDD), a solid-state drive (SSD), or the like. The storage medium can further include a combination of the above-mentioned storage mediums.

[0129] Although the embodiments of the present application are described in conjunction with the drawings, various modifications and changes can be made by those skilled in the art without departing from the spirit and scope of the present application, and such modifications and changes fall within the scope defined by the appended claims.

Claims

1. A method of seismic attribute prediction, characterized by, The method comprises the following steps: acquiring seismic raw data and labeling the seismic raw data; performing data preprocessing on the labeled seismic raw data to determine a processing result; constructing an initial seismic prediction model and training the initial seismic prediction model using the processing result to determine a target seismic prediction model; acquiring seismic data of a target block, inputting the seismic data of the target block into the target seismic prediction model, and determining a prediction result.

2. The method of claim 1, wherein, The method of acquiring seismic raw data and labeling the historical data comprises the following steps: acquiring a preset feature and collecting seismic raw data with the preset feature; labeling the seismic raw data at the same position coordinates to determine seismic samples of different blocks.

3. The method of claim 2, wherein, The method of performing data preprocessing on the labeled historical data to determine a processing result comprises the following steps: performing an outlier removal operation on seismic samples of each different block to determine operation data; performing denoising processing on the operation data to determine denoising data; performing normalization processing on the denoising data to determine an optimized processing result.

4. The method of claim 2, wherein, The method of constructing an initial seismic prediction model comprises the following steps: using a preset network model to identify and classify the seismic samples to determine an identification result and a classification label; determining the preset network model parameters according to the identification result and the classification label to construct the initial seismic prediction model.

5. The method of claim 4, wherein, The method of using a preset network model to identify and classify the seismic samples to determine an identification result and a classification label comprises the following steps: The preset network model comprises an encoder, a decoder, a discriminator, a decoupler, a block identifier, a first classifier, and a second classifier; using the encoder to transform the seismic samples to obtain seismic main features; using each structure of the preset network model to identify and classify the seismic main features to determine corresponding parameters, an identification result, and a classification label.

6. The method of claim 5, wherein, The method of using each structure of the preset network model to identify and classify the seismic main features to determine corresponding parameters, an identification result, and a classification label comprises the following steps: using the decoder to decode the seismic main features to determine reconstruction loss data; using the discriminator to discriminate the seismic main features to determine normal discrimination loss data; using the first classifier to classify the seismic main features to determine lithology classification loss data; using the decoupler to decompose the seismic main features to determine block features and category features; using the second classifier to classify the block features and the category features to determine the lithology category of the corresponding features inputted; using the block identifier to identify the block features and the category features to determine the source block of the corresponding features inputted.

7. The method of claim 6, wherein, The method of training the initial seismic prediction model using the processing result to determine a target seismic prediction model comprises the following steps: inputting seismic samples of each different block and their corresponding labels into the initial seismic prediction model in sequence to determine training parameters; performing parameter adjustment on the initial seismic prediction model using the seismic samples of each different block corresponding to the training parameters to determine a target seismic prediction model.

8. A seismic attribute prediction apparatus characterized by comprising: The method comprises the following steps: A first processing module is configured to acquire seismic raw data and label the seismic raw data; A second processing module is configured to perform data preprocessing on the labeled seismic raw data and determine a processing result; A third processing module is configured to construct an initial seismic prediction model, train the initial seismic prediction model using the processing result, and determine a target seismic prediction model; A fourth processing module is configured to acquire target block seismic data, input the target block seismic data into the target seismic prediction model, and determine a prediction result.

9. An electronic device, comprising: The method comprises the following steps: A memory and a processor are communicatively connected, the memory stores computer instructions, and the processor executes the computer instructions to perform the method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer instructions for making a computer execute the method according to any one of claims 1-7.

Images