Reservoir classification method, device, electronic device and computer readable storage medium

Abstract

The present application provides a reservoir classification method, device, electronic device and computer-readable storage medium, wherein the method includes: obtaining a target imaging logging image and target conventional logging curve data corresponding to a target depth region in a target well; inputting the target imaging logging image into a first convolutional neural network classification model, and outputting a first reservoir classification result of a target reservoir corresponding to the target depth region; inputting the target conventional logging curve data into a second convolutional neural network classification model, and outputting a second reservoir classification result of the target reservoir; inputting the spliced ​​result obtained by splicing the first reservoir classification result and the second reservoir classification result into a target fully connected layer, and obtaining a target reservoir classification result of the target reservoir; and determining the category of the target reservoir according to the target reservoir classification result. The present solution automatically classifies reservoirs through a convolutional neural network classification model, which is conducive to improving the work efficiency and classification accuracy of reservoir classification.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a reservoir classification method, device, electronic device and computer-readable storage medium. Background Art

[0002] Reservoir refers to the rock formation that stores oil and gas, containing interconnected pores. Classifying reservoirs can objectively and generally express the oil and gas storage capacity of reservoirs, which is of great significance for accurate quantitative evaluation of oil and gas reserves, and plays an important guiding role in oil and gas exploration and development.

[0003] In the prior art, reservoir classification is usually done by manually analyzing reservoir parameters such as porosity, permeability, and lithology. However, the manual reservoir classification method mainly relies on expert experience, which is affected by subjective factors, has low classification accuracy, and has low work efficiency in the face of a large amount of well logging data. Summary of the invention

[0004] In view of this, the purpose of the present application is to provide a reservoir classification method, device, electronic device and computer-readable storage medium to improve the accuracy of reservoir classification and improve work efficiency.

[0005] In a first aspect, an embodiment of the present application provides a reservoir classification method, comprising:

[0006] Acquire a target imaging logging image and target conventional logging curve data corresponding to a target depth region in a target wellbore; the target conventional logging curve data includes target logging data corresponding to a plurality of target logging curves at the target depth region;

[0007] Inputting the target imaging logging image into a pre-trained first convolutional neural network classification model, processing the first image feature of the target imaging logging image by the first convolutional neural network classification model, and obtaining a first reservoir classification result of the target reservoir corresponding to the target depth area in the target wellbore;

[0008] Inputting the target conventional well logging curve data into a pre-trained second convolutional neural network classification model, processing each of the target well logging data contained in the target conventional well logging curve data by the second convolutional neural network classification model, and obtaining a second reservoir classification result of the target reservoir corresponding to the target depth area in the target wellbore;

[0009] Inputting a concatenated result obtained by concatenating the first reservoir classification result and the second reservoir classification result into a pre-trained target fully connected layer to obtain a target reservoir classification result of the target reservoir;

[0010] The category of the target reservoir is determined according to the target reservoir classification result.

[0011] In combination with the first aspect, the embodiment of the present application provides a first possible implementation of the first aspect, wherein before acquiring a target imaging logging image corresponding to a target depth area in a target wellbore, the method further includes:

[0012] Acquire an initial imaging logging image corresponding to the target wellbore; the initial imaging logging image contains imaging logging information corresponding to each depth in the target wellbore;

[0013] Segmenting the initial imaging logging image according to a preset depth interval to obtain a plurality of imaging logging images and the target depth region corresponding to each of the imaging logging images;

[0014] For each of the imaging logging images, grayscale processing is performed on the imaging logging image to obtain the target imaging logging image corresponding to the imaging logging image.

[0015] In combination with the first aspect, the embodiment of the present application provides a second possible implementation of the first aspect, wherein, before acquiring the target conventional logging curve data corresponding to the target depth area position in the target well, the method further includes:

[0016] Acquire initial conventional logging curve data corresponding to the target well; the initial conventional logging curve data includes initial logging curve data corresponding to a plurality of initial logging curves; the initial logging curve data includes initial curve values ​​corresponding to each depth in the target well;

[0017] For each of the initial well logging curve data, smoothing and filtering are performed on the initial well logging curve data to obtain first well logging curve data; the first well logging curve data contains first curve values ​​corresponding to each depth in the target well;

[0018] For each of the first logging curve data, normalize each of the first curve values ​​in the first logging curve data according to the maximum first curve value and the minimum first curve value in the first logging curve data to obtain the second curve value corresponding to each of the first curve values, and determine the second logging curve data and the target logging curve corresponding to the second logging curve data according to the second curve value corresponding to each depth; the value range of the second curve value is 0-1;

[0019] For the second logging curve data corresponding to each target logging curve, according to each target depth contained in the target depth area, the second curve value corresponding to each target depth is selected from the second logging curve data, and the selected second curve values ​​are averaged to obtain the target curve value corresponding to the target depth area, and the target curve value corresponding to the target depth area is used as the target logging data corresponding to the target logging curve at the target depth area.

[0020] In combination with the first aspect, the embodiment of the present application provides a third possible implementation of the first aspect, wherein the first image feature of the target imaging logging image is processed by the first convolutional neural network classification model to obtain a first reservoir classification result of the target reservoir corresponding to the target depth area in the target wellbore, including:

[0021] Taking the initial image feature matrix corresponding to the target imaging logging image as the input image feature matrix, performing a first convolution process for a first preset number of times on the input image feature matrix, performing a first nonlinear transformation after each first convolution process, and performing a first pooling process on the result obtained by each first nonlinear transformation;

[0022] The image feature matrix obtained by each first pooling process is used as the input image feature matrix for the next first convolution process;

[0023] The first convolution processing includes: calculating the dot product of each first receptive field data in the first image feature matrix and the first convolution kernel, obtaining a first eigenvalue corresponding to each first receptive field data, and constructing a second image feature matrix according to each first eigenvalue; the matrix size corresponding to the first receptive field data is the same as the matrix size corresponding to the first convolution kernel;

[0024] The first nonlinear transformation includes: for each third image feature matrix, inputting each second eigenvalue in the third image feature matrix into a preset activation function to obtain a third eigenvalue corresponding to each second eigenvalue in the third image feature matrix; in the preset activation function, for each second eigenvalue, when the second eigenvalue is greater than 0, determining the second eigenvalue as the third eigenvalue; when the second eigenvalue is not greater than 0, changing the second eigenvalue to obtain the third eigenvalue; and constructing a fourth image feature matrix using the third eigenvalue;

[0025] The first pooling process includes: extracting a maximum value from each second receptive field data in the fifth image feature matrix, and constructing a sixth image feature matrix according to the maximum value extracted from each second receptive field data;

[0026] The output image feature matrix obtained by the last first pooling process is subjected to a first full connection process to obtain the first reservoir classification result corresponding to the target imaging logging image.

[0027] In combination with the first aspect, the embodiment of the present application provides a fourth possible implementation of the first aspect, wherein the second convolutional neural network classification model is used to process each of the target logging data contained in the target conventional logging curve data to obtain a second reservoir classification result corresponding to the target depth area in the target wellbore, including:

[0028] Using each of the target logging data included in the target conventional logging curve data as input logging data, performing a second convolution process for a second preset number of times on the input logging data, performing a second nonlinear transformation after each second convolution process, and performing a second pooling process on a result obtained by each second nonlinear transformation;

[0029] The feature data obtained by each second pooling process is used as the input logging data for the next second convolution process;

[0030] The second convolution processing includes: dividing all the first well logging data to obtain at least one third receptive field data; for each of the third receptive field data, calculating the dot product of the first well logging data contained in the third receptive field data and the second convolution kernel to obtain the fourth eigenvalue corresponding to the third receptive field data, and constructing the second well logging data according to the fourth eigenvalue; the matrix size corresponding to the third receptive field data is the same as the matrix size corresponding to the second convolution kernel; the second nonlinear transformation includes: for each third well logging data, judging whether the third well logging data is greater than 0; if the third well logging data is greater than 0, determining the third well logging data as the fourth well logging data; if the third well logging data is not greater than 0, determining the product of the third well logging data and a preset multiple as the fourth well logging data;

[0031] The second pooling processing includes: dividing all the fifth logging data to obtain at least one fourth receptive field data; for each of the fourth receptive field data, extracting the maximum value from the fourth receptive field data, and constructing the sixth logging data according to the maximum value extracted from each of the fourth receptive field data; performing a second full connection processing on the output logging data obtained by the last second pooling processing to obtain the second reservoir classification result corresponding to the target conventional logging curve data.

[0032] In combination with the first aspect, the embodiment of the present application provides a fifth possible implementation of the first aspect, wherein, after determining the categories corresponding to the target reservoirs corresponding to each of the target depth areas in the target well, the method further includes:

[0033] For each of the target reservoirs, determining whether the category corresponding to the target reservoir is the same as the category corresponding to other target reservoirs adjacent to the target reservoir;

[0034] If the category corresponding to the target reservoir is the same as the category corresponding to other target reservoirs adjacent to the target reservoir, the target reservoir and the other target reservoirs adjacent to the target reservoir are merged to obtain a synthetic reservoir.

[0035] In combination with the first aspect, an embodiment of the present application provides a sixth possible implementation of the first aspect, wherein the first convolutional neural network classification model, the second convolutional neural network classification model, and the target fully connected layer are trained in the following manner:

[0036] Constructing training samples for training the first convolutional neural network classification model, the second convolutional neural network classification model and the target fully connected layer; the training samples include: sample imaging logging images and sample conventional logging curve data corresponding to the sample depth area in the sample well, and category labels of sample reservoirs corresponding to the sample depth area; the sample conventional logging curve data includes sample logging data corresponding to multiple sample logging curves at the sample depth area;

[0037] Inputting the sample imaging logging image into a first initial classification model, processing the second image feature of the sample imaging logging image by the first initial classification model, and obtaining a first prediction result of the sample reservoir;

[0038] Inputting the sample conventional well logging curve data into a second initial classification model, and processing each of the sample well logging data contained in the sample conventional well logging curve data by the second initial classification model to obtain a second prediction result of the sample reservoir;

[0039] Inputting a third prediction result obtained by concatenating the first prediction result and the second prediction result of the sample reservoir into an initial fully connected layer, and outputting a prediction classification result of the sample reservoir;

[0040] Calculating a loss value according to the predicted classification result of the sample reservoir and the category label;

[0041] The loss value is used to update the parameters in the first initial classification model, the second initial classification model, and the initial fully connected layer, and the updated first initial classification model, the second initial classification model, and the initial fully connected layer are trained again until the training reaches a preset number of training times, thereby obtaining the first convolutional neural network classification model, the second convolutional neural network classification model, and the target fully connected layer.

[0042] In combination with the first aspect, the embodiment of the present application provides a seventh possible implementation of the first aspect, wherein, after determining the category of the target reservoir according to the target reservoir classification result, the method further includes:

[0043] According to the category of the target reservoir, a storage amount of the target substance stored in the target reservoir is estimated.

[0044] In a second aspect, the present application also provides a reservoir training device, including:

[0045] The first acquisition module is used to acquire a target imaging logging image and target conventional logging curve data corresponding to a target depth region in a target wellbore; the target conventional logging curve data includes target logging data corresponding to a plurality of target logging curves at the target depth region;

[0046] A first input module is used to input the target imaging logging image into a pre-trained first convolutional neural network classification model, and process the first image feature of the target imaging logging image through the first convolutional neural network classification model to obtain a first reservoir classification result of the target reservoir corresponding to the target depth area in the target wellbore;

[0047] A second input module is used to input the target conventional well logging curve data into a pre-trained second convolutional neural network classification model, and process each of the target well logging data contained in the target conventional well logging curve data through the second convolutional neural network classification model to obtain a second reservoir classification result of the target reservoir corresponding to the target depth area in the target well;

[0048] A third input module is used to input a concatenated result obtained by concatenating the first reservoir classification result and the second reservoir classification result into a pre-trained target fully connected layer to obtain a target reservoir classification result of the target reservoir;

[0049] The first determination module is used to determine the category of the target reservoir according to the classification result of the target reservoir.

[0050] In combination with the second aspect, the embodiment of the present application provides a first possible implementation of the second aspect, which further includes:

[0051] A second acquisition module is used to acquire an initial imaging logging image corresponding to the target wellbore before the first acquisition module acquires the target imaging logging image corresponding to the target depth area in the target wellbore; the initial imaging logging image contains imaging logging information corresponding to each depth in the target wellbore;

[0052] A segmentation module, used to segment the initial imaging logging image according to a preset depth interval to obtain a plurality of imaging logging images and the target depth area corresponding to each of the imaging logging images;

[0053] The grayscale processing module is used to perform grayscale processing on each imaging logging image to obtain the target imaging logging image corresponding to the imaging logging image.

[0054] In combination with the second aspect, the embodiment of the present application provides a second possible implementation of the second aspect, which further includes:

[0055] A third acquisition module is used to acquire initial conventional logging curve data corresponding to the target well before the first acquisition module acquires the target conventional logging curve data corresponding to the target depth area position in the target well; the initial conventional logging curve data includes initial logging curve data corresponding to each of a plurality of initial logging curves; the initial logging curve data includes initial curve values ​​corresponding to each depth in the target well;

[0056] A filtering module, for performing smoothing filtering on each of the initial well logging curve data to obtain first well logging curve data; the first well logging curve data contains first curve values ​​corresponding to each depth in the target well;

[0057] a normalization module, for normalizing each of the first well logging curve data according to the maximum first curve value and the minimum first curve value in the first well logging curve data, to obtain a second curve value corresponding to each of the first curve values, and determining the second well logging curve data and the target well logging curve corresponding to the second well logging curve data according to the second curve value corresponding to each depth; the value range of the second curve value is 0-1;

[0058] A mean processing module is used to select the second curve value corresponding to each target depth from the second logging curve data corresponding to each target logging curve according to each target depth contained in the target depth area, and perform mean processing on the selected second curve values ​​to obtain the target curve value corresponding to the target depth area, and use the target curve value corresponding to the target depth area as the target logging data corresponding to the target logging curve at the target depth area.

[0059] In combination with the second aspect, an embodiment of the present application provides a third possible implementation of the second aspect, wherein the first input module, when used to calculate the first image feature of the target imaging logging image through the first convolutional neural network classification model to obtain a first reservoir classification result of the target reservoir corresponding to the target depth area in the target wellbore, is specifically used to:

[0060] Taking the initial image feature matrix corresponding to the target imaging logging image as the input image feature matrix, performing a first convolution process for a first preset number of times on the input image feature matrix, performing a first nonlinear transformation after each first convolution process, and performing a first pooling process on the result obtained by each first nonlinear transformation;

[0061] The image feature matrix obtained by each first pooling process is used as the input image feature matrix for the next first convolution process;

[0062] The first convolution processing includes: calculating the dot product of each first receptive field data in the first image feature matrix and the first convolution kernel, obtaining a first eigenvalue corresponding to each first receptive field data, and constructing a second image feature matrix according to each first eigenvalue; the matrix size corresponding to the first receptive field data is the same as the matrix size corresponding to the first convolution kernel;

[0063] The first nonlinear transformation includes: for each third image feature matrix, inputting each second eigenvalue in the third image feature matrix into a preset activation function to obtain a third eigenvalue corresponding to each second eigenvalue in the third image feature matrix; in the preset activation function, for each second eigenvalue, when the second eigenvalue is greater than 0, determining the second eigenvalue as the third eigenvalue; when the second eigenvalue is not greater than 0, changing the second eigenvalue to obtain the third eigenvalue; and constructing a fourth image feature matrix using the third eigenvalue;

[0064] The first pooling process includes: extracting a maximum value from each second receptive field data in the fifth image feature matrix, and constructing a sixth image feature matrix according to the maximum value extracted from each second receptive field data;

[0065] The output image feature matrix obtained by the last first pooling process is subjected to a first full connection process to obtain the first reservoir classification result corresponding to the target imaging logging image.

[0066] In combination with the second aspect, an embodiment of the present application provides a fourth possible implementation of the second aspect, wherein the second input module is used to process each of the target logging data contained in the target conventional logging curve data through the second convolutional neural network classification model to obtain a second reservoir classification result of the target reservoir corresponding to the target depth area in the target wellbore, specifically for:

[0067] Using each of the target logging data included in the target conventional logging curve data as input logging data, performing a second convolution process for a second preset number of times on the input logging data, performing a second nonlinear transformation after each second convolution process, and performing a second pooling process on a result obtained by each second nonlinear transformation;

[0068] The feature data obtained by each second pooling process is used as the input logging data for the next second convolution process;

[0069] The second convolution processing includes: dividing all the first well logging data to obtain at least one third receptive field data; for each of the third receptive field data, calculating the dot product of the first well logging data contained in the third receptive field data and the second convolution kernel to obtain the fourth eigenvalue corresponding to the third receptive field data, and constructing the second well logging data according to the fourth eigenvalue; the matrix size corresponding to the third receptive field data is the same as the matrix size corresponding to the second convolution kernel; the second nonlinear transformation includes: for each third well logging data, judging whether the third well logging data is greater than 0; if the third well logging data is greater than 0, determining the third well logging data as the fourth well logging data; if the third well logging data is not greater than 0, determining the product of the third well logging data and a preset multiple as the fourth well logging data;

[0070] The second pooling processing includes: dividing all the fifth logging data to obtain at least one fourth receptive field data; for each of the fourth receptive field data, extracting the maximum value from the fourth receptive field data, and constructing the sixth logging data according to the maximum value extracted from each of the fourth receptive field data; performing a second full connection processing on the output logging data obtained by the last second pooling processing to obtain the second reservoir classification result corresponding to the target conventional logging curve data.

[0071] In combination with the second aspect, the embodiment of the present application provides a fifth possible implementation of the second aspect, which further includes:

[0072] A judgment module is used to determine, after determining the categories corresponding to the target reservoirs corresponding to each of the target depth areas in the target well, whether the category corresponding to the target reservoir is the same as the category corresponding to other target reservoirs adjacent to the target reservoir for each of the target reservoirs;

[0073] The merging module is used to merge the target reservoir with other target reservoirs adjacent to the target reservoir to obtain a synthetic reservoir if the category corresponding to the target reservoir is the same as the category corresponding to other target reservoirs adjacent to the target reservoir.

[0074] In combination with the second aspect, the embodiment of the present application provides a sixth possible implementation of the second aspect, which further includes:

[0075] A construction module, used to construct training samples for training the first convolutional neural network classification model, the second convolutional neural network classification model and the target fully connected layer; the training samples include: sample imaging logging images and sample conventional logging curve data corresponding to the sample depth area in the sample well, and the category label of the sample reservoir corresponding to the sample depth area; the sample conventional logging curve data includes sample logging data corresponding to multiple sample logging curves at the sample depth area;

[0076] a fourth input module, used for inputting the sample imaging logging image into a first initial classification model, processing the second image feature of the sample imaging logging image by the first initial classification model, and obtaining a first prediction result of the sample reservoir;

[0077] a fifth input module, used for inputting the sample conventional logging curve data into a second initial classification model, and processing each of the sample logging data contained in the sample conventional logging curve data by the second initial classification model to obtain a second prediction result of the sample reservoir;

[0078] A sixth input module, used for inputting a third prediction result obtained by concatenating the first prediction result and the second prediction result of the sample reservoir into an initial fully connected layer, and outputting a prediction classification result of the sample reservoir;

[0079] A calculation module, used for calculating a loss value according to the predicted classification result of the sample reservoir and the category label;

[0080] An updating module is used to use the loss value to update the parameters in the first initial classification model, the second initial classification model and the initial fully connected layer, and to retrain the updated first initial classification model, the second initial classification model and the initial fully connected layer until the training reaches a preset number of training times and stops training, thereby obtaining the first convolutional neural network classification model, the second convolutional neural network classification model and the target fully connected layer.

[0081] In combination with the second aspect, the embodiment of the present application provides a seventh possible implementation of the second aspect, which further includes:

[0082] The second determination module is used to evaluate the storage amount of the target substance stored in the target reservoir according to the category of the target reservoir after the first determination module determines the category of the target reservoir according to the classification result of the target reservoir.

[0083] In a third aspect, an embodiment of the present application further provides an electronic device, comprising: a processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor and the memory communicate through the bus, and when the machine-readable instructions are executed by the processor, the steps in any possible implementation of the first aspect above are performed.

[0084] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps in any possible implementation of the first aspect described above are executed.

[0085] The embodiment of the present application provides a reservoir classification method, device, electronic device and computer-readable storage medium, which processes the first image feature of the target imaging logging image through a first convolutional neural network classification model to obtain a first reservoir classification result of the target reservoir corresponding to the target depth area in the target drilling; and processes each target logging data contained in the target conventional logging curve data through a second convolutional neural network classification model to obtain a second reservoir classification result of the target reservoir; the splicing result obtained by splicing the first reservoir classification result and the second reservoir classification result is input into the target fully connected layer to obtain the target reservoir classification result of the target reservoir; according to the target reservoir classification result, the category of the target reservoir is determined. Compared with the method of manually classifying reservoirs in the prior art, the present solution automatically classifies reservoirs through a convolutional neural network model, which is conducive to improving the work efficiency of reservoir classification and improving the accuracy of reservoir classification.

[0086] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are specifically cited below and described in detail with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0087] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0088] Figure 1A flow chart of a reservoir classification method provided in an embodiment of the present application is shown;

[0089] Figure 2 A schematic diagram of an initial image feature matrix, a first convolution kernel, and a second image feature matrix provided in an embodiment of the present application is shown;

[0090] Figure 3 A schematic diagram showing a fifth image feature matrix and a sixth image feature matrix provided in an embodiment of the present application is shown;

[0091] Figure 4 A schematic structural diagram of a reservoir classification device provided in an embodiment of the present application is shown;

[0092] Figure 5 A schematic structural diagram of an electronic device provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0093] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present application.

[0094] In the prior art, the methods for reservoir classification mainly include the following three categories: (1) Classifying reservoirs by manually analyzing reservoir parameters such as porosity, permeability, and lithology. (2) Analyzing conventional logging curve data through machine learning methods such as multivariate linear regression, support vector machine, and fuzzy logic to achieve automatic classification of reservoir categories. (3) Automatic classification of reservoir categories by analyzing imaging logging data.

[0095] Although the above methods have achieved the classification of reservoirs, there are still some problems: among them, the artificial reservoir classification method mainly relies on expert experience, which is affected by subjective factors, has low classification accuracy, and has low work efficiency in the face of a large amount of logging data. The method based on the analysis of conventional logging curve data or imaging logging data has a certain degree of singleness, and the classification accuracy is not high.

[0096] In view of the above problems, based on this, the embodiments of the present application provide a reservoir classification method, device, electronic device and computer-readable storage medium to improve the accuracy and work efficiency of reservoir classification, which is described below through embodiments.

[0097] Embodiment 1:

[0098] To facilitate understanding of this embodiment, a reservoir classification method disclosed in an embodiment of the present application is first introduced in detail. Figure 1 A flow chart of a reservoir classification method provided in an embodiment of the present application is shown. Figure 1 As shown, the following steps are included:

[0099] S101: Acquire a target imaging logging image and target conventional logging curve data corresponding to a target depth region in a target wellbore; the target conventional logging curve data includes target logging data corresponding to a plurality of target logging curves at the target depth region.

[0100] The target wellbore stores substances such as oil and natural gas, wherein the target wellbore stores the oil or natural gas through a reservoir. The reservoir has interconnected pores, allowing the target wellbore to store the oil or natural gas through the reservoir.

[0101] The target well can be a newly drilled well. For a newly completed well (i.e. a newly drilled well), it is necessary to determine the reservoir division results corresponding to different depths of the entire well, that is, to classify the reservoirs in the target well and determine whether each reservoir in the target well is a high-quality reservoir or a non-high-quality reservoir. The purpose of classifying reservoirs is to analyze the potential for oil and gas development in the target well and determine the key to further increase recoverable reserves in the target well.

[0102] The target depth refers to the distance between each position in the well and the ground, and the target depth area refers to the distance between each area in the well and the ground. For example, the target depth area is between 100 meters and 102 meters, between 102 and 104 meters, etc.

[0103] The target imaging logging image corresponding to the target depth area includes the well logging image corresponding to the target depth area in the target well. The target conventional logging curve data corresponding to the target depth area includes the target logging data corresponding to each of the multiple target logging curves corresponding to the target depth area in the target well. There are multiple target logging curves, and the horizontal coordinate of each target logging curve is the distance from the ground, that is, the depth in the well, and the vertical coordinate is the curve value corresponding to each depth. Among them, the number of data included in the target conventional logging curve data is the same as the number of target logging curves.

[0104] For example, there are 7 target logging curves, each of which corresponds to a curve value in the target depth region. Therefore, the target conventional logging curve data corresponding to the target depth region includes 7 curve values. The reservoir properties of different depth regions can be reflected according to the trend of the target logging curve changing with depth.

[0105] S102: Input the target imaging logging image into a pre-trained first convolutional neural network classification model, process the first image feature of the target imaging logging image through the first convolutional neural network classification model, and obtain a first reservoir classification result of the target reservoir corresponding to the target depth area in the target wellbore.

[0106] The first convolutional neural network classification model can be a two-dimensional convolutional neural network classification model. Each target depth area corresponds to a target reservoir. For example, if the target depth area is between 100-102 meters, the target reservoir corresponding to the target depth area is a reservoir between 100-102 meters in the target wellbore. The first reservoir classification result is the probability of the target reservoir in each category predicted by the first convolutional neural network classification model. Exemplarily, when the reservoir categories are divided into non-high-quality reservoirs, first-class high-quality treatments, and second-class high-quality reservoirs, the first reservoir classification result is the probability that the target reservoir belongs to non-high-quality reservoirs, first-class high-quality treatments, and second-class high-quality reservoirs, respectively.

[0107] S103: Input the target conventional logging curve data into a pre-trained second convolutional neural network classification model, and process each target logging data contained in the target conventional logging curve data through the second convolutional neural network classification model to obtain a second reservoir classification result of the target reservoir corresponding to the target depth area in the target well.

[0108] The second convolutional neural network classification model may be a one-dimensional convolutional neural network classification model. The second reservoir classification result is the probability of the target reservoir in each category predicted by the second convolutional neural network classification model. For the same target reservoir, the first reservoir classification result and the second reservoir classification result corresponding to the target reservoir may be the same or different.

[0109] S104: Inputting a concatenated result obtained by concatenating the first reservoir classification result and the second reservoir classification result into a pre-trained target fully connected layer to obtain a target reservoir classification result of the target reservoir.

[0110] Exemplarily, the classification result of the first reservoir is [0.2, 0.5, 0.3], and the classification result of the second reservoir is [0.1, 0.8, 0.1]. The concatenated result of the two reservoir classification results is [0.2, 0.5, 0.3, 0.1, 0.8, 0.1]. This concatenated result is input into the pre-trained target fully connected layer to obtain the target reservoir classification result, that is, the probability that the target reservoir belongs to a non-high-quality reservoir, a first-class high-quality reservoir, or a second-class high-quality reservoir.

[0111] S105: Determine the category of the target reservoir according to the target reservoir classification result.

[0112] The target reservoir classification results include the probabilities that the target reservoir belongs to a non-high-quality reservoir, a first-class high-quality reservoir, and a second-class high-quality reservoir. The category with the highest probability is selected from the target reservoir classification results to be determined as the final category of the target reservoir.

[0113] For example, the target reservoir classification result includes a probability of 0.2 that the target reservoir belongs to a Class I high-quality reservoir, a probability of 0.45 that the target reservoir belongs to a Class II high-quality reservoir, and a probability of 0.1 that the target reservoir belongs to a non-high-quality reservoir. The Class II high-quality reservoir corresponding to the probability of 0.45 is determined as the category of the target reservoir, that is, the category of the target reservoir is a Class II high-quality reservoir.

[0114] Compared with the method of only analyzing conventional logging curves or only analyzing imaging logging data in the prior art, the reservoir classification method of this scheme comprehensively considers two sets of data, namely, target imaging logging images and target conventional logging curve data. By comprehensively analyzing the imaging logging images and conventional logging curve data, it is beneficial to improve the accuracy of classification.

[0115] In a feasible implementation manner, before executing step S101 to obtain a target imaging logging image corresponding to a target depth area in a target wellbore, the method further includes:

[0116] S1011: Acquire an initial imaging logging image corresponding to the target wellbore; the initial imaging logging image contains imaging logging information corresponding to each depth in the target wellbore.

[0117] The initial imaging logging image is a method of imaging the physical parameters of the well wall and objects around the well based on the observation of the geophysical field in the borehole. It can intuitively and visually reflect the formation characteristics corresponding to different depths in the target well.

[0118] The initial imaging logging image includes imaging logging images corresponding to various depth regions in the entire target wellbore, which are images taken by a logging instrument.

[0119] S1012: Segmenting the initial imaging logging image according to a preset depth interval to obtain a plurality of imaging logging images and a target depth region corresponding to each imaging logging image.

[0120] In a specific embodiment, the header and tail of the initial imaging logging image may contain annotation information (i.e., redundant information), such as the target well number, the target well name, the color system annotation, the north-south annotation, and other annotation information. Therefore, when segmenting the initial imaging logging image, the annotation information on the header and tail of the initial imaging logging image is first removed (i.e., the redundant information on the initial imaging logging image is removed) to obtain the initial imaging logging image after the annotation information is removed. Then, according to the preset depth interval, the initial imaging logging image after the annotation information is removed is segmented again to obtain multiple imaging logging images.

[0121] The size of the imaging logging image is smaller than the size of the initial imaging logging image, and the preset depth interval may be 1 meter, 1.5 meters, 2 meters, etc., and may be adjusted empirically.

[0122] For example, when the initial imaging logging image includes logging images at a depth of 100-500 meters in the target well, if the preset depth interval is 2 meters, there are 200 imaging logging images obtained after segmentation, including 1 imaging logging image corresponding to the target depth area of ​​100-102 meters, 1 imaging logging image corresponding to the target depth area of ​​102-104 meters, and 1 imaging logging image corresponding to the target depth area of ​​498-500 meters, etc. The size of each imaging logging image is the same.

[0123] S1013: For each imaging logging image, grayscale processing is performed on the imaging logging image to obtain a target imaging logging image corresponding to the imaging logging image.

[0124] Both the initial imaging logging image and the imaging logging image are color images in RGB format (RGB represents the three color channels of red R, green G, and blue B). Among them, each pixel in the color image of the RGB color space is jointly determined by the data of the three channels of R, G, and B. Since it is more complicated to process the data of the three channels, in the imaging logging image, the color has too little effect on the judgment of the reservoir category. Therefore, the imaging logging image is grayed, that is, the three RGB channels are unified into one channel of data. Since our human eyes have different sensitivities to the three RGB colors, this scheme uses a weighted method to obtain the grayed data. The specific formula is as follows:

[0125] gray(x)=0.299×red(x)+0.587×green(x)+0.114×blue(x)

[0126] Among them, for each pixel point x in the imaging logging image, red(x) represents the pixel value corresponding to the red R channel of the pixel point x, green(x) represents the pixel value corresponding to the green G channel of the pixel point x, blue(x) represents the pixel value corresponding to the blue B channel of the pixel point x, and gray(x) represents the grayscale value of the pixel point x.

[0127] In a possible implementation, before executing step S101 to obtain target conventional logging curve data corresponding to the target depth zone position in the target well, the method further includes:

[0128] S1014: Acquire initial conventional logging curve data corresponding to the target well; the initial conventional logging curve data includes initial logging curve data corresponding to a plurality of initial logging curves; the initial logging curve data includes initial curve values ​​corresponding to each depth in the target well.

[0129] There are multiple initial logging curves. For example, the initial logging curves may be natural gamma, natural potential, borehole diameter, acoustic wave, neutron, density, resistivity and other 7 logging curves. Each initial logging curve contains the initial curve values ​​corresponding to each depth in the target wellbore. The horizontal coordinate of each initial logging curve is the depth in the target wellbore, and the vertical coordinate is the corresponding curve value. The initial logging curve data corresponding to each initial logging curve contains the curve values ​​corresponding to each depth on the initial logging curve. The initial conventional logging curve data contains the curve values ​​corresponding to each initial logging curve at each depth.

[0130] S1015: For each initial well logging curve data, perform smoothing and filtering processing on the initial well logging curve data to obtain first well logging curve data; the first well logging curve data includes first curve values ​​corresponding to each depth in the target well.

[0131] In logging operations, due to random effects such as nuclear decay, extranuclear electrons and gamma quanta, or random collisions of logging probes, multiple refractions and reflections of sound waves, etc., the initial logging curve will contain a lot of interference information. If we directly use these signals that are not related to geological features for logging analysis, it may have a great impact on the analysis results. Therefore, we use curve smoothing filtering to smooth and filter each initial logging curve, retaining the useful signals related to geological special effects to the greatest extent, and effectively suppressing the interference of short-period burr signals.

[0132] The smoothing filter process adopts Savitzky-Golay smoothing filter (a filtering method based on least squares fitting), which eliminates noise while keeping the original curve morphology unchanged.

[0133] In a specific embodiment, when the Savitzky-Golay smoothing filter method is used for smoothing filtering, specifically, a k-order polynomial fitting is performed on the data points in a window of a certain length through a weighted average algorithm of a moving sliding window, thereby obtaining a fitting effect.

[0134] For example, for each initial well logging curve, if there are 20 curve values ​​corresponding to depths on the initial well logging curve, the 20 depths are divided into groups of 5 depths each, and 4 groups of data are obtained, that is, the width of the moving sliding window is 5. For each group of data (that is, for each data in the moving sliding window), a k-order polynomial fitting is performed using the group of data to obtain the parameter a in the following formula: i (i=1...k), then input the curve values ​​corresponding to each depth contained in this set of data into the following formula:

[0135] Y=a0+a1x+a2x 2 +…+a k x k

[0136] Wherein, x represents the curve value on the initial logging curve corresponding to a certain depth, and Y represents the first curve value on the first logging curve data corresponding to the depth after smoothing.

[0137] S1016: For each first well logging curve data, each first curve value in the first well logging curve data is normalized according to the maximum first curve value and the minimum first curve value in the first well logging curve data to obtain the second curve value corresponding to each first curve value, and the second well logging curve data and the target well logging curve corresponding to the second well logging curve data are determined according to the second curve value corresponding to each depth; the value range of the second curve value is 0-1. Since different first well logging curves have different fluctuation ranges, their different dimension ranges (i.e., fluctuation ranges) will cause the first well logging curve with a smaller value to be swallowed by the first well logging curve with a larger value. Therefore, each first well logging curve data is normalized to convert the dimensional expression into a dimensionless expression, thereby improving the convergence speed and classification accuracy of the model.

[0138] For example, the ordinate of some first well logging curves may fluctuate between 1 and 5, while the ordinate of another first well logging curve may fluctuate between 40 and 50. In order to avoid the influence of different fluctuation ranges on the classification result, this solution performs normalization processing on each first curve value in the first well logging curve data of each first well logging curve. Specifically, for each first well logging curve, each first curve value in the first well logging curve data of the first well logging curve is input into the following formula:

[0139]

[0140] Among them, P max represents the maximum value of all first curve values ​​contained in the first logging curve data, P min represents the minimum value of all first curve values ​​contained in the first logging curve data, P i represents the first curve values ​​contained in the first logging curve data, Q i Indicates the normalized second curve value corresponding to each first curve value included in the first logging curve data. Since the second curve value included in each second logging curve data is between 0 and 1, the fluctuation range of the ordinate of each second logging curve is between 0 and 1.

[0141] S1017: For the second logging curve data corresponding to each target logging curve, according to each target depth contained in the target depth area, the second curve value corresponding to each target depth is selected from the second logging curve data, and the selected second curve values ​​are averaged to obtain the target curve value corresponding to the target depth area, and the target curve value corresponding to the target depth area is used as the target logging data corresponding to the target logging curve in the target depth area.

[0142] For each target logging curve, the second logging curve data corresponding to the target logging curve includes the second curve values ​​corresponding to each depth, for example, the second logging curve data includes: the second curve value corresponding to 100 meters is 40.22, the second curve value corresponding to 101 meters is 43.932, the second curve value corresponding to 102 meters is 56.33, the second curve value corresponding to 103 meters is 67.9, etc. If the target depth area is 100-102, the target depths included in the target depth area are 100 meters, 101 meters, and 102 meters, and the second curve values ​​corresponding to the target depth area 100-102 are 40.22, 43.932, and 56.33. The average value of the second curve values ​​corresponding to the target depth area 100-102 is calculated to be 48.827, that is, the target curve value corresponding to the target depth area 100-102 is 48.827. This method can be used to calculate the target curve value corresponding to each target depth area.

[0143] In one possible implementation, when executing step S102 to process the first image feature of the target imaging logging image by the first convolutional neural network classification model to obtain a first reservoir classification result of the target reservoir corresponding to the target depth area in the target wellbore, the following steps may be specifically performed:

[0144] An initial image feature matrix corresponding to the target imaging logging image is used as an input image feature matrix, a first convolution process is performed on the input image feature matrix for a first preset number of times, a first nonlinear transformation is performed after each first convolution process, and a first pooling process is performed on a result obtained by each first nonlinear transformation.

[0145] The image feature matrix obtained by each first pooling process is used as the input image feature matrix for the next first convolution process.

[0146] The first convolution processing includes: calculating the dot product of each first receptive field data in the first image feature matrix with the first convolution kernel, obtaining the first eigenvalue corresponding to each first receptive field data, and constructing the second image feature matrix according to each first eigenvalue; the matrix size corresponding to the first receptive field data is the same as the matrix size corresponding to the first convolution kernel.

[0147] The first nonlinear transformation includes: for each third image feature matrix, inputting each second eigenvalue in the third image feature matrix into a preset activation function to obtain a third eigenvalue corresponding to each second eigenvalue in the third image feature matrix; in the preset activation function, for each second eigenvalue, when the second eigenvalue is greater than 0, determining the second eigenvalue as the third eigenvalue; when the second eigenvalue is not greater than 0, changing the second eigenvalue to obtain the third eigenvalue; and constructing a fourth image feature matrix using the third eigenvalue;

[0148] The first pooling process includes: extracting a maximum value from each second receptive field data in the fifth image feature matrix, and constructing a sixth image feature matrix according to the maximum value extracted from each second receptive field data.

[0149] The output image feature matrix obtained by the last first pooling process is subjected to a first full connection process to obtain a first reservoir classification result corresponding to the target imaging logging image.

[0150] The first convolutional neural network classification model takes the target imaging logging image itself as input. It does not need to stretch or deform the target imaging logging image, and retains the spatial structure of the target imaging logging image, as well as the shape, texture and other information in the image.

[0151] The first convolutional neural network classification model mainly includes: a first convolutional layer of a first preset number of times, a first pooling layer of a first preset number of times, a first nonlinear transformation of a first preset number of times, and a first fully connected layer. A first convolution process is performed in the first convolutional layer, specifically for extracting features of the input characteristic features; a first pooling process is performed in the first pooling layer, and a first fully connected process is performed in the first fully connected layer.

[0152] In a specific embodiment, when there are two first convolutional layers, namely the first convolutional layer J1 and the first convolutional layer J2, and there are two first pooling layers, namely the first pooling layer C1 and the first pooling layer C2, the output of the first convolutional layer J1 is used as the input of the first pooling layer C1, the output of the first pooling layer C1 is used as the input of the first convolutional layer J2, the output of the first convolutional layer J2 is used as the input of the first pooling layer C2, and the output of the first pooling layer C2 is used as the input of the first fully connected layer.

[0153] The convolution layer is the core part of the convolutional neural network model, which is used to extract features from input features. This block imitates the working principle of the human visual cortex, assuming that adjacent pixels in the image are closely related, and there is no obvious connection between pixels that are far away. The convolution layer retains the spatial structure of the image and uses a local perception convolution method instead of a full connection method. This local area is also called the receptive field, where the size of the receptive field depends on the size of the convolution kernel.

[0154] First, according to the target imaging logging image, the initial image feature matrix of the target imaging logging image is determined. The initial image feature matrix of the target imaging logging image is input into the first convolutional layer J1. Figure 2 FIG. 4 shows a schematic diagram of an initial image feature matrix, a first convolution kernel, and a second image feature matrix provided in an embodiment of the present application. Figure 2 As shown, in the first convolution layer J1, if the initial image feature matrix is ​​a 5×5 matrix (i.e., the input image feature is a 5×5 matrix), and the first convolution kernel is a 3×3 matrix, then according to the matrix size of the first convolution kernel, a plurality of first receptive field data are determined from the initial image feature matrix. Among them, since the first convolution kernel is a 3×3 matrix, the first receptive field data is also a 3×3 matrix.

[0155] like Figure 2 As shown, the shaded 3×3 matrix (i.e., from the first row and the first column to the third row and the third column) is used as the first first receptive field data Z1, from the first row and the second column to the third row and the fourth column is used as the second first receptive field data Z2, from the first row and the third column to the third row and the fifth column is used as the third first receptive field data Z3, from the second row and the first column to the fourth row and the third column is used as the fourth first receptive field data Z4, and in this order, 9 first receptive field data can be obtained. That is, there are repeated parts between the first receptive field data.

[0156] The dot product of the first receptive field data Z1 and the first convolution kernel is calculated, and the first eigenvalue corresponding to the first receptive field data Z1 is 9; the dot product of the first receptive field data Z2 and the first convolution kernel is calculated, and the first eigenvalue corresponding to the first receptive field data Z2 is 3; the dot product of the first receptive field data Z3 and the first convolution kernel is calculated, and the first eigenvalue corresponding to the first receptive field data Z3 is 9. In this order, 9 first eigenvalues ​​can be obtained.

[0157] When calculating the dot product of the first receptive field data and the first convolution kernel, the matrix corresponding to the first receptive field data and the values ​​at the corresponding positions in the matrix corresponding to the first convolution kernel are specifically multiplied, and the sum of the dot products is calculated. Taking the dot product of the first receptive field data Z1 and the first convolution kernel as an example, the specific calculation process is:

[0158] 1×1+1×2+0×0+1×2+0×1+1×1+2×1+0×0+1×1=9

[0159] The second image feature matrix is ​​constructed according to the obtained 9 first eigenvalues, that is, the position of each first receptive field data on the first feature matrix is ​​determined to determine the position of the first eigenvalue corresponding to each first receptive field data on the second image feature matrix.

[0160] In this embodiment, a first convolution kernel of size 3×3 is used to extract the features of the imaging logging image. It is believed that the superposition of several consecutive 3×3 small convolution kernels not only has the same receptive field as the large convolution kernel, but also introduces more first nonlinear transformations, which can better extract the deep features of the input data.

[0161] The second image feature matrix output by the first convolutional layer J1 is used as the third image feature matrix, and the third image feature matrix is ​​subjected to a first nonlinear transformation. Specifically, for each third image feature matrix, each value in the matrix is ​​used as a second feature value in the third image feature matrix. Each second feature value in the third image feature matrix is ​​input into a preset activation function (ie, the ELU activation function):

[0162]

[0163] Where y represents each second eigenvalue, ELU(y) represents the third eigenvalue obtained after the first nonlinear transformation. is a fixed constant, and e is a fixed constant.

[0164] In this embodiment, the activation function of the network adopts the ELU activation function instead of the ReLU activation function to avoid the neuron necrosis problem and improve the robustness of the model to noise.

[0165] The fourth image feature matrix output by the first nonlinear transformation is input into the first pooling layer C1 as the fifth image feature matrix. Figure 3 FIG. 4 shows a schematic diagram of a fifth image feature matrix and a sixth image feature matrix provided in an embodiment of the present application. Figure 3 As shown, in the first pooling layer C1, the fifth image feature matrix is ​​divided to obtain at least one second receptive field data, and for each second receptive field data, the maximum value is extracted from the second receptive field data to obtain the maximum value corresponding to each second receptive field data, and according to the position of each second receptive field data on the fifth image feature matrix, the sixth image feature matrix is ​​constructed using the maximum value corresponding to each second receptive field data. There is no repeated part between any two second receptive field data.

[0166] For example, Figure 3 As shown, the shaded 2×2 matrix (i.e., from the first row and the first column to the second row and the second column) is used as the first second receptive field data Y1; from the first row and the third column to the second row and the fourth column is used as the second second receptive field data Y2; from the third row and the first column to the fourth row and the second column is used as the third second receptive field data Y3; from the third row and the third column to the fourth row and the fourth column is used as the fourth second receptive field data Y4. In this order, 4 second receptive field data can be obtained. Among them, the maximum value in the second receptive field data Y1 is 8, the maximum value in the second receptive field data Y2 is 9, the maximum value in the second receptive field data Y3 is 7, and the maximum value in the second receptive field data Y4 is 5.

[0167] In a specific embodiment, if the second receptive field data is a 2×2 matrix, and the fifth image feature matrix is ​​a 3×3 matrix, it is necessary to fill the fifth image feature matrix with 0, that is, fill the fourth column and the fourth row with 0, and divide the fifth image feature matrix after filling with 0 to determine multiple second receptive field data.

[0168] The output image feature matrix of the first pooling layer C2 is used as the input of the first fully connected layer. In the first fully connected layer, the output image feature matrix of the first pooling layer C2 is subjected to the first fully connected processing to obtain the first reservoir classification result corresponding to the target imaging logging image.

[0169] In a possible embodiment, when executing step S103 to process each target logging data contained in the target conventional logging curve data by the second convolutional neural network classification model to obtain a second reservoir classification result corresponding to the target depth area in the target wellbore, the following steps may be specifically performed:

[0170] Using each target logging data included in the target conventional logging curve data as input logging data, performing a second convolution process for a second preset number of times on the input logging data, performing a second nonlinear transformation after each second convolution process, and performing a second pooling process on a result obtained by each second nonlinear transformation;

[0171] The feature data obtained by each second pooling process is used as the input logging data for the next second convolution process;

[0172] The second convolution processing includes: dividing all the first well logging data to obtain at least one third receptive field data; for each third receptive field data, calculating the dot product of the first well logging data contained in the third receptive field data and the second convolution kernel to obtain the fourth eigenvalue corresponding to the first receptive field data, and constructing the second well logging data according to the fourth eigenvalue; the matrix size corresponding to the third receptive field data is the same as the matrix size corresponding to the second convolution kernel; the second nonlinear transformation includes: for each third well logging data, judging whether the third well logging data is greater than 0; if the third well logging data is greater than 0, determining the third well logging data as the fourth well logging data; if the third well logging data is not greater than 0, determining the product of the third well logging data and a preset multiple as the fourth well logging data;

[0173] The second pooling process includes: dividing all the fifth well logging data to obtain at least one fourth receptive field data; for each fourth receptive field data, extracting a maximum value from the fourth receptive field data, and constructing sixth well logging data according to the maximum value extracted from each fourth receptive field data;

[0174] The output logging data obtained by the last second pooling process is subjected to a second full connection process to obtain a second reservoir classification result corresponding to the target conventional logging curve data.

[0175] The second convolutional neural network classification model mainly includes: a second convolutional layer of a second preset number of times, a second pooling layer of a second preset number of times, a Leaky ReLU activation function, and a second fully connected layer. A second convolution process is performed in the second convolutional layer, a second pooling process is performed in the second pooling layer, a second nonlinear transformation is performed in the Leaky ReLU activation function, and a second fully connected process is performed in the second fully connected layer.

[0176] In a specific embodiment, if there are three second convolutional layers, namely, the second convolutional layer E1, the second convolutional layer E2, and the second convolutional layer E3; and there are three second pooling layers, namely, the second pooling layer R1, the second pooling layer R2, and the second pooling layer R3. Among them, the output of the second convolutional layer E1 is used as the first input of the Leaky ReLU activation function, the first output of the Leaky ReLU activation function is used as the input of the second pooling layer R1, the output of the second pooling layer R1 is used as the input of the second convolutional layer E2, the output of the second convolutional layer E2 is used as the second input of the Leaky ReLU activation function, the second output of the Leaky ReLU activation function is used as the input of the second pooling layer R2, the output of the second pooling layer R2 is used as the input of the second convolutional layer E3, the output of the second convolutional layer E3 is used as the third input of the Leaky ReLU activation function, the third output of the Leaky ReLU activation function is used as the input of the second pooling layer R3, and the output of the second pooling layer R3 is used as the input of the second fully connected layer.

[0177] Each target logging data contained in the target conventional logging curve data corresponding to the target depth area is input into the second convolutional layer E1. For example, when the target depth area is 106-108 meters and there are 9 target logging curves, the target conventional logging curve data corresponding to the target depth area 106-108 meters contains 9 target logging data, which are 0.144, 0.15, 0.73, 0.66, 0.137, 0.475, 0.48, 0.118, and 0.504 respectively.

[0178] In the second convolutional layer E1, the target logging data is used as the first logging data, and all the target logging data are divided to obtain three third receptive field data, which are [0.144, 0.15, 0.73], [0.66, 0.137, 0.475], and [0.48, 0.118, 0.504] respectively.

[0179] For each third receptive field data, calculate the dot product of the first logging data contained in the third receptive field data and the second convolution kernel to obtain the fourth eigenvalue corresponding to the third receptive field data. Specifically, calculate the dot product of the third receptive field data U1[0.144, 0.15, 0.73] and the second convolution kernel to obtain a fourth eigenvalue. Calculate the dot product of the third receptive field data U2[0.66, 0.137, 0.475] and the second convolution kernel to obtain the second fourth eigenvalue. Calculate the dot product of the third receptive field data U3[0.48, 0.118, 0.504] and the second convolution kernel to obtain the third fourth eigenvalue.

[0180] The second well logging data is constructed according to the fourth eigenvalue, that is, the three fourth eigenvalues ​​are used as the second well logging data. That is, the output of the second convolutional layer E1 is the second well logging data containing the three fourth eigenvalues.

[0181] The second well logging data output by the second convolutional layer E1 and containing three fourth eigenvalues ​​is input as the third well logging data into the Leaky ReLU (Rectified linear unit, ReLU) activation function. Specifically, the Leaky ReLU activation function is as follows:

[0182]

[0183] Among them, z is the third logging data input into the Leaky ReLU activation function, Leaky ReLU(z) is the fourth logging data output by the LeakyReLU activation function, and β is the preset parameter in the Leaky ReLU activation function, that is, the preset multiple. The output of the Leaky ReLU activation function is the fourth logging data. The value range of the fourth logging data is between 0 and 1, and the decimal is usually rounded to two decimal places.

[0184] The fourth logging data output by the Leaky ReLU activation function is input into the second pooling layer R1 as the fifth logging data. By way of example, in another specific embodiment, if the number of fifth logging data input into the second pooling layer R1 is 6, the 6 fifth logging data are divided to obtain 3 fourth receptive field data, which are (0.86, 0.44), (0.62, 0.46), and (0.51, 0.65), respectively.

[0185] The maximum value 0.86 is extracted from the fourth receptive field data (0.86, 0.44). The maximum value 0.62 is extracted from the fourth receptive field data (0.62, 0.46). The maximum value 0.65 is extracted from the fourth receptive field data (0.51, 0.65). The extracted 0.86, 0.62 and 0.65 are used to construct the sixth well logging data. That is, the output of the second pooling layer R1 is the sixth well logging data.

[0186] The logging data output by the second pooling layer R3 is input to the second fully connected layer. In the second fully connected layer, the logging data output by the second pooling layer R3 is subjected to a second fully connected processing, and a second reservoir classification result corresponding to the target conventional logging curve data is output.

[0187] In a possible embodiment, after executing step S105 to determine the categories of the target reservoirs corresponding to each target depth area in the target well, the following steps may be specifically performed:

[0188] S1051: For each target reservoir, determine whether the category corresponding to the target reservoir is the same as the category corresponding to other target reservoirs adjacent to the target reservoir.

[0189] For the target reservoir corresponding to each target depth area, there are 1 or 2 other target reservoirs adjacent to the target reservoir. Exemplarily, there is 1 other target reservoir adjacent to the target reservoir corresponding to the target depth areas 100-102, specifically the target reservoir corresponding to the target depth areas 102-104. There are 2 other target reservoirs adjacent to the target reservoir corresponding to the target depth areas 120-122, namely the target reservoir corresponding to the target depth areas 118-120 and the target reservoir corresponding to the target depth areas 122-124.

[0190] S1052: If the category corresponding to the target reservoir is the same as the category corresponding to other target reservoirs adjacent to the target reservoir, the target reservoir and the other target reservoirs adjacent to the target reservoir are merged to obtain a synthetic reservoir.

[0191] In a specific embodiment, the target reservoir corresponding to the target depth area 120-122 is classified as a first-class high-quality reservoir, the target reservoir corresponding to the target depth area 118-120 is classified as a first-class high-quality reservoir, and the target reservoir corresponding to the target depth area 122-124 is classified as a non-high-quality reservoir. Then, the target reservoir corresponding to the target depth area 120-122 is merged with the target reservoir corresponding to the target depth area 118-120 to obtain a synthetic reservoir corresponding to the target depth area 118-122, and the synthetic reservoir is classified as a first-class high-quality reservoir.

[0192] Due to the limitation of reservoir thickness, the thickness of each type of high-quality reservoir or non-high-quality reservoir will not be too thin. In order to accurately establish the boundary between each high-quality reservoir and non-high-quality reservoir, the regional merging method is adopted to further merge the target reservoir division results. First, the depth difference corresponding to each divided reservoir is detected globally. For the results of fine reservoir division, the reservoirs before and after the layer are compared and merged with the surrounding reservoir categories. By using this technology, finer divisions are merged, abnormal classification results are eliminated, and reservoir boundaries are re-established.

[0193] In one possible embodiment, the first convolutional neural network classification model, the second convolutional neural network classification model, and the target fully connected layer are trained in the following manner:

[0194] S1001: Constructing training samples for training the first convolutional neural network classification model, the second convolutional neural network classification model and the target fully connected layer; the training samples include: sample imaging logging images and sample conventional logging curve data corresponding to the sample depth area in the sample drilling, and category labels of sample reservoirs corresponding to the sample depth area; the sample conventional logging curve data includes sample logging data corresponding to multiple sample logging curves at the sample depth area.

[0195] When constructing the sample imaging logging image, firstly, an initial sample imaging logging image corresponding to the sample drilling is obtained. The initial sample imaging logging image includes sample logging images corresponding to each sample depth in the sample drilling.

[0196] The annotated information in the initial sample imaging logging image is removed, and the initial sample imaging logging image after the annotated information is removed is segmented according to a preset depth interval to obtain multiple intermediate sample imaging logging images and a sample depth area corresponding to each intermediate sample imaging logging image.

[0197] For each intermediate sample imaging logging image, grayscale processing is performed on the intermediate sample imaging logging image to obtain a sample imaging logging image corresponding to the intermediate sample imaging logging image.

[0198] In this application, since the pixels of the initial sample imaging logging images are too large and contain redundant information for annotation, if the initial sample imaging logging images are used directly to train the first initial classification model, it will lead to difficulties in model training. Too much redundant information will bring too much noise, affecting the final accuracy of the first convolutional neural network classification model. In order to improve the efficiency and accuracy of model training, this application adopts image segmentation technology to segment the initial sample imaging logging images, remove the redundant information for annotation of the initial sample imaging logging images, and cut them into small-size sample imaging logging images that are conducive to the training of the first convolutional neural network classification model, thereby increasing the amount of data while improving the training efficiency and accuracy of the first convolutional neural network classification model.

[0199] When constructing sample conventional logging curve data, the following steps can be followed:

[0200] Acquire initial sample conventional logging curve data corresponding to the sample well; the initial sample conventional logging curve data includes initial sample logging curve data corresponding to a plurality of initial sample logging curves; the initial sample logging curve data includes initial sample curve values ​​corresponding to each depth in the sample well;

[0201] For the initial sample logging curve corresponding to each initial sample logging curve data, smoothing and filtering are performed on the initial sample logging curve to obtain a smoothed sample logging curve and first sample logging curve data corresponding to the smoothed sample logging curve; the first sample logging curve data includes first sample curve values ​​corresponding to each depth in the sample wellbore;

[0202] For each first sample logging curve data, normalize each first sample curve value in the first sample logging curve data according to the maximum first sample curve value and the minimum first sample curve value in the first sample logging curve data to obtain the second sample curve value corresponding to each first sample curve value, and determine the second sample logging curve data and the sample logging curve corresponding to the second sample curve data according to the second sample curve value corresponding to each depth; the value range of the second sample curve value is 0-1;

[0203] For the second sample logging curve data corresponding to each sample logging curve, according to each sample depth contained in the sample depth area, the second sample curve value corresponding to each sample depth is selected from the second sample logging curve data, and the selected second sample curve values ​​are averaged to obtain the sample curve value corresponding to the sample depth area, and the sample curve value corresponding to the sample depth area is used as the sample logging data corresponding to the sample logging curve in the sample depth area.

[0204] Among them, in the sample drilling, each sample depth area corresponds to a sample imaging logging image and a sample conventional logging curve data.

[0205] S1002: Inputting the sample imaging logging image into the first initial classification model, processing the second image feature of the sample imaging logging image by the first initial classification model, and obtaining a first prediction result of the sample reservoir.

[0206] The first initial classification model includes: a first initial convolutional layer of a first preset number of times, a first initial pooling layer of a first preset number of times, an ELU activation function and a first initial fully connected layer, performing a first sample convolution processing in the first initial convolutional layer, performing a first sample pooling processing in the first initial pooling layer, performing a first initial nonlinear transformation in the ELU activation function, and performing a first sample fully connected processing in the first initial fully connected layer.

[0207] Using the initial sample imaging logging image features corresponding to the sample imaging logging image as an input sample image feature matrix, performing a first sample convolution process for a first preset number of times on the input sample image feature matrix, performing a first initial nonlinear transformation after each first sample convolution process, and performing a first sample pooling process on a result obtained by each first initial nonlinear transformation;

[0208] The sample image feature matrix obtained by each first sample pooling process is used as the input sample image feature matrix for the next first sample convolution process;

[0209] The first sample convolution processing includes: in the first initial convolution layer, calculating the dot product of each fifth receptive field data in the first sample image feature matrix and the first sample convolution kernel respectively, obtaining the fifth eigenvalue corresponding to each fifth receptive field data, and constructing the second sample image feature matrix according to each fifth eigenvalue; the matrix size corresponding to the fifth receptive field data is the same as the matrix size corresponding to the first sample convolution kernel;

[0210] The first initial nonlinear transformation includes: taking the second sample image feature matrix output by the first initial convolution layer as the third sample image feature matrix, and performing the first initial nonlinear transformation on the third sample image feature matrix; specifically, for each third sample image feature matrix, inputting each sixth eigenvalue in the third sample image feature matrix into the ELU activation function, and outputting the value corresponding to each sixth eigenvalue; in the ELU activation function, for each sixth eigenvalue, when the sixth eigenvalue is greater than 0, determining the sixth eigenvalue as the seventh eigenvalue; when the sixth eigenvalue is not greater than 0, transforming the sixth eigenvalue to obtain the seventh eigenvalue; and constructing the fourth sample image feature matrix using the seventh eigenvalue;

[0211] The first sample pooling process includes: in the first initial pooling layer, extracting the maximum value from each sixth receptive field data in the fifth sample image feature matrix, and constructing a sixth sample image feature matrix according to the maximum value extracted from each sixth receptive field data;

[0212] In the first initial fully connected layer, the output sample image feature matrix obtained by the last first sample pooling process is subjected to the first sample fully connected process to obtain the first prediction result corresponding to the sample imaging logging image. The first prediction result includes the probability of the sample reservoir corresponding to the sample depth area predicted by the first initial classification model in each category.

[0213] S1003: Input the sample conventional logging curve data into the second initial classification model, and process each sample logging data contained in the sample conventional logging curve data through the second initial classification model to obtain a second prediction result of the sample reservoir.

[0214] The second initial classification model includes: a second preset number of second initial convolutional layers, a second preset number of second initial pooling layers, a Leaky ReLU activation function, and a second initial fully connected layer. A second sample convolution process is performed in the second initial convolutional layer, a second sample pooling process is performed in the second initial pooling layer, a second initial nonlinear transformation is performed in the Leaky ReLU activation function, and a second sample fully connected process is performed in the second initial fully connected layer.

[0215] Using each sample logging data included in the sample conventional logging curve data as input sample logging data, performing a second sample convolution process for a second preset number of times on the input sample logging data, performing a sample nonlinear transformation after each second sample convolution process, and performing a second sample pooling process on the result obtained by each sample nonlinear transformation;

[0216] The sample feature data obtained by each second sample pooling process is used as the input sample logging data for the next second sample convolution process;

[0217] The second sample convolution processing includes: in the second initial convolution layer, all the first sample logging data are divided to obtain at least one seventh receptive field data; for each seventh receptive field data, the dot product of the first sample logging data contained in the seventh receptive field data and the second convolution kernel is calculated to obtain the eighth characteristic value corresponding to the seventh receptive field data, and the second sample logging data is constructed according to the eighth characteristic value; the second sample nonlinear transformation includes: in the Leaky ReLU activation function, for each third sample logging data, judging whether the third sample logging data is greater than 0; if the third sample logging data is greater than 0, determining the third sample logging data as the fourth sample logging data; if the third sample logging data is not greater than 0, determining the product of the third sample logging data and the preset multiple as the fourth sample logging data;

[0218] The second sample pooling process includes: dividing all the fifth sample logging data to obtain at least one eighth receptive field data; for each eighth receptive field data, extracting a maximum value from the eighth receptive field data, and constructing the sixth sample logging data according to the maximum value extracted from each eighth receptive field data;

[0219] In the second initial fully connected layer, the output sample logging data obtained by the last second pooling process is subjected to the second sample fully connected process to obtain the second prediction result corresponding to the sample conventional logging curve data. The second prediction result includes the probability of the sample reservoir corresponding to the sample depth area predicted by the second initial classification model in each category.

[0220] S1004: Input a third prediction result obtained by concatenating the first prediction result of the sample reservoir with the second prediction result into the initial fully connected layer, and output a prediction classification result of the sample reservoir.

[0221] The prediction classification result includes the probability of the sample reservoir corresponding to the sample depth area in each category predicted by comprehensively considering the sample imaging logging image and the sample conventional logging curve data.

[0222] S1005: Calculate the loss value according to the predicted classification result and category label of the sample reservoir.

[0223] According to the prediction classification results of the sample reservoir, the prediction category of the sample reservoir is determined; according to the prediction category and category label of the sample reservoir, the loss value is calculated through the cross entropy loss function.

[0224] In a specific embodiment, in the case of binary classification, we assume that the predicted class probabilities are p (e.g., the probability of a high-quality reservoir) and 1-p (e.g., the probability of a non-high-quality reservoir), then the loss function expression is:

[0225]

[0226] Among them, U is the loss value, M is the number of training samples, and U j is the loss result corresponding to the jth training sample, t j is the category label of the jth training sample, q j The probability of predicting the jth training sample as a high-quality reservoir.

[0227] In another specific embodiment, in the case of multi-classification, that is, when there are multiple reservoir categories, the loss function expression is:

[0228]

[0229] Among them, L is the loss value, N is the number of training samples, and L i is the loss result corresponding to the i-th training sample, G is the number of categories, y ic is the category label of the i-th training sample, P ic The probability of predicting the i-th training sample as class C.

[0230] S1006: Use the loss value to update the parameters in the first initial classification model, the second initial classification model, and the initial fully connected layer, and re-train the updated first initial classification model, the second initial classification model, and the initial fully connected layer until the training reaches a preset number of training times and stops training, thereby obtaining the first convolutional neural network classification model, the second convolutional neural network classification model, and the target fully connected layer.

[0231] After the training is completed, the first initial convolutional layer, the first initial pooling layer, and the first initial fully connected layer in the first initial classification model are used as the first convolutional layer, the first pooling layer, and the first fully connected layer in the first convolutional neural network classification model. The second initial convolutional layer, the second initial pooling layer, the initial activation function, and the second initial fully connected layer in the second initial classification model are used as the second convolutional layer, the second pooling layer, the activation function, and the second fully connected layer in the second convolutional neural network classification model.

[0232] In a possible embodiment, after executing step S105 to determine the category of the target reservoir according to the target reservoir classification result, the following steps may be specifically performed:

[0233] According to the category of the target reservoir, the storage amount of the target substance stored in the target reservoir is evaluated.

[0234] Different types of target reservoirs result in different storage amounts of target substances stored in the target reservoirs. Target substances include oil, natural gas, etc. In a specific embodiment, the types of target reservoirs are divided into first-class high-quality reservoirs, second-class high-quality reservoirs, and non-high-quality reservoirs. Among them, the storage amount of target substances stored in the first-class high-quality reservoirs is the largest, and the storage amount of target substances stored in the non-high-quality reservoirs is the smallest.

[0235] In a possible embodiment, when executing step S1052, if the category corresponding to the target reservoir is the same as the category corresponding to other target reservoirs adjacent to the target reservoir, then the target reservoir is merged with the other target reservoirs adjacent to the target reservoir to obtain a synthetic reservoir, and the following steps may be specifically performed:

[0236] The storage amount of the target substance stored in the synthetic composition is evaluated based on the number and type of the target reservoirs included in the synthetic reservoir.

[0237] Embodiment 2:

[0238] Based on the same technical concept, the embodiment of the present application also provides a reservoir classification device. Figure 4 A schematic diagram of the structure of a reservoir classification device provided in an embodiment of the present application is shown. Figure 4 As shown, the device comprises:

[0239] The first acquisition module 401 is used to acquire a target imaging logging image and target conventional logging curve data corresponding to a target depth region in a target wellbore; the target conventional logging curve data includes target logging data corresponding to a plurality of target logging curves at the target depth region;

[0240] A first input module 402 is used to input the target imaging logging image into a pre-trained first convolutional neural network classification model, and process the first image feature of the target imaging logging image through the first convolutional neural network classification model to obtain a first reservoir classification result of the target reservoir corresponding to the target depth area in the target wellbore;

[0241] A second input module 403 is used to input the target conventional well logging curve data into a pre-trained second convolutional neural network classification model, and process each of the target well logging data contained in the target conventional well logging curve data through the second convolutional neural network classification model to obtain a second reservoir classification result of the target reservoir corresponding to the target depth area in the target well;

[0242] A third input module 404 is used to input a concatenated result obtained by concatenating the first reservoir classification result and the second reservoir classification result into a pre-trained target fully connected layer to obtain a target reservoir classification result of the target reservoir;

[0243] The first determination module 405 is used to determine the category of the target reservoir according to the classification result of the target reservoir.

[0244] Optionally, a second acquisition module is used to acquire an initial imaging logging image corresponding to the target wellbore before the first acquisition module 401 acquires the target imaging logging image corresponding to the target depth area in the target wellbore; the initial imaging logging image contains imaging logging information corresponding to each depth in the target wellbore;

[0245] A segmentation module, used to segment the initial imaging logging image according to a preset depth interval to obtain a plurality of imaging logging images and the target depth area corresponding to each of the imaging logging images;

[0246] The grayscale processing module is used to perform grayscale processing on each imaging logging image to obtain the target imaging logging image corresponding to the imaging logging image.

[0247] Optionally, a third acquisition module is used to acquire initial conventional logging curve data corresponding to the target well before the first acquisition module 401 acquires the target conventional logging curve data corresponding to the target depth area position in the target well; the initial conventional logging curve data includes initial logging curve data corresponding to each of a plurality of initial logging curves; the initial logging curve data includes initial curve values ​​corresponding to each depth in the target well;

[0248] A filtering module, for performing smoothing filtering on each of the initial well logging curve data to obtain first well logging curve data; the first well logging curve data contains first curve values ​​corresponding to each depth in the target well;

[0249] a normalization module, for normalizing each of the first well logging curve data according to the maximum first curve value and the minimum first curve value in the first well logging curve data, to obtain a second curve value corresponding to each of the first curve values, and determining the second well logging curve data and the target well logging curve corresponding to the second well logging curve data according to the second curve value corresponding to each depth; the value range of the second curve value is 0-1;

[0250] A mean processing module is used to select the second curve value corresponding to each target depth from the second logging curve data corresponding to each target logging curve according to each target depth contained in the target depth area, and perform mean processing on the selected second curve values ​​to obtain the target curve value corresponding to the target depth area, and use the target curve value corresponding to the target depth area as the target logging data corresponding to the target logging curve at the target depth area.

[0251] Optionally, when the first input module 402 is used to process the first image feature of the target imaging logging image through the first convolutional neural network classification model to obtain a first reservoir classification result of the target reservoir corresponding to the target depth area in the target wellbore, it is specifically used to:

[0252] Taking the initial image feature matrix corresponding to the target imaging logging image as the input image feature matrix, performing a first convolution process for a first preset number of times on the input image feature matrix, performing a first nonlinear transformation after each first convolution process, and performing a first pooling process on the result obtained by each first nonlinear transformation;

[0253] The image feature matrix obtained by each first pooling process is used as the input image feature matrix for the next first convolution process;

[0254] The first convolution processing includes: calculating the dot product of each first receptive field data in the first image feature matrix and the first convolution kernel, obtaining a first eigenvalue corresponding to each first receptive field data, and constructing a second image feature matrix according to each first eigenvalue; the matrix size corresponding to the first receptive field data is the same as the matrix size corresponding to the first convolution kernel;

[0255] The first nonlinear transformation includes: for each third image feature matrix, inputting each second eigenvalue in the third image feature matrix into a preset activation function to obtain a third eigenvalue corresponding to each second eigenvalue in the third image feature matrix; in the preset activation function, for each second eigenvalue, when the second eigenvalue is greater than 0, determining the second eigenvalue as the third eigenvalue; when the second eigenvalue is not greater than 0, changing the second eigenvalue to obtain the third eigenvalue; and constructing a fourth image feature matrix using the third eigenvalue;

[0256] The first pooling process includes: extracting a maximum value from each second receptive field data in the fifth image feature matrix, and constructing a sixth image feature matrix according to the maximum value extracted from each second receptive field data;

[0257] The output image feature matrix obtained by the last first pooling process is subjected to a first full connection process to obtain the first reservoir classification result corresponding to the target imaging logging image.

[0258] Optionally, when the second input module 403 is used to process each of the target logging data contained in the target conventional logging curve data through the second convolutional neural network classification model to obtain a second reservoir classification result of the target reservoir corresponding to the target depth area in the target well, it is specifically used to:

[0259] Using each of the target logging data included in the target conventional logging curve data as input logging data, performing a second convolution process for a second preset number of times on the input logging data, performing a nonlinear transformation after each second convolution process, and performing a second pooling process on a result obtained by each nonlinear transformation;

[0260] The feature data obtained by each second pooling process is used as the input logging data for the next second convolution process;

[0261] The second convolution processing includes: dividing all the first well logging data to obtain at least one third receptive field data; for each of the third receptive field data, calculating the dot product of the first well logging data contained in the third receptive field data and the second convolution kernel to obtain the fourth eigenvalue corresponding to the third receptive field data, and constructing the second well logging data according to the fourth eigenvalue; the matrix size corresponding to the third receptive field data is the same as the matrix size corresponding to the second convolution kernel; the second nonlinear transformation includes: for each third well logging data, judging whether the third well logging data is greater than 0; if the third well logging data is greater than 0, determining the third well logging data as the fourth well logging data; if the third well logging data is not greater than 0, determining the product of the third well logging data and a preset multiple as the fourth well logging data;

[0262] The second pooling processing includes: dividing all the fifth logging data to obtain at least one fourth receptive field data; for each of the fourth receptive field data, extracting the maximum value from the fourth receptive field data, and constructing the sixth logging data according to the maximum value extracted from each of the fourth receptive field data; performing a second full connection processing on the output logging data obtained by the last second pooling processing to obtain the second reservoir classification result corresponding to the target conventional logging curve data.

[0263] Optionally, also include:

[0264] A judgment module is used to determine, after determining the categories corresponding to the target reservoirs corresponding to each of the target depth areas in the target well, whether the category corresponding to the target reservoir is the same as the category corresponding to other target reservoirs adjacent to the target reservoir for each of the target reservoirs;

[0265] The merging module is used to merge the target reservoir with other target reservoirs adjacent to the target reservoir to obtain a synthetic reservoir if the category corresponding to the target reservoir is the same as the category corresponding to other target reservoirs adjacent to the target reservoir.

[0266] Optionally, also include:

[0267] A construction module, used to construct training samples for training the first convolutional neural network classification model, the second convolutional neural network classification model and the target fully connected layer; the training samples include: sample imaging logging images and sample conventional logging curve data corresponding to the sample depth area in the sample well, and the category label of the sample reservoir corresponding to the sample depth area; the sample conventional logging curve data includes sample logging data corresponding to multiple sample logging curves at the sample depth area;

[0268] A fourth input module, used for inputting the sample imaging logging image into a first initial classification model, calculating the second image feature of the sample imaging logging image through the first initial classification model, and obtaining a first prediction result of the sample reservoir;

[0269] a fifth input module, used for inputting the sample conventional logging curve data into a second initial classification model, and performing weighted calculation on each of the sample logging data contained in the sample conventional logging curve data through the second initial classification model to obtain a second prediction result of the sample reservoir;

[0270] A sixth input module, used for inputting a third prediction result obtained by concatenating the first prediction result and the second prediction result of the sample reservoir into an initial fully connected layer, and outputting a prediction classification result of the sample reservoir;

[0271] A calculation module, used for calculating a loss value according to the predicted classification result of the sample reservoir and the category label;

[0272] An updating module is used to use the loss value to update the parameters in the first initial classification model, the second initial classification model and the initial fully connected layer, and to retrain the updated first initial classification model, the second initial classification model and the initial fully connected layer until the training reaches a preset number of training times and stops training, thereby obtaining the first convolutional neural network classification model, the second convolutional neural network classification model and the target fully connected layer.

[0273] Optionally, also include:

[0274] The second determination module is used to evaluate the storage amount of the target substance stored in the target reservoir according to the category of the target reservoir after the first determination module determines the category of the target reservoir according to the classification result of the target reservoir.

[0275] For the specific implementation method steps and principles, please refer to the description of Example 1, which will not be repeated here in detail.

[0276] Embodiment three:

[0277] Based on the same technical concept, the embodiment of the present application also provides an electronic device, Figure 5 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application is shown. Figure 5 As shown, the electronic device 500 includes: a processor 501, a memory 502 and a bus 503, the memory stores machine-readable instructions executable by the processor, when the electronic device is running, the processor 501 communicates with the memory 502 through the bus 503, and the processor 501 executes the machine-readable instructions to execute the method steps described in Example 1. For the specific execution method steps and principles, please refer to the description of Example 1, which will not be repeated in detail here.

[0278] Embodiment 4:

[0279] Based on the same technical concept, the fourth embodiment of the present application also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the method steps described in the first embodiment are executed. For the specific method steps and principles of execution, please refer to the description of the first embodiment, which will not be repeated in detail here.

[0280] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0281] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, and the indirect coupling or communication connection of devices or units can be electrical, mechanical or other forms.

[0282] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0283] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0284] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a non-volatile computer-readable storage medium that is executable by a processor. Based on this understanding, the technical solution of the present application can essentially be embodied in the form of a software product, or in other words, the part that contributes to the prior art or the part of the technical solution. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0285] Finally, it should be noted that the above-described embodiments are only specific implementation methods of the present application, which are used to illustrate the technical solutions of the present application, rather than to limit them. The protection scope of the present application is not limited thereto. Although the present application is described in detail with reference to the above-described embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-described embodiments within the technical scope disclosed in the present application, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A reservoir classification method, characterized in that: include: Acquire target imaging logging images and target conventional logging curve data corresponding to the target depth area in the target wellbore; The target conventional logging curve data includes target logging data corresponding to a plurality of target logging curves at the target depth region; Inputting the target imaging logging image into a pre-trained first convolutional neural network classification model, processing the first image feature of the target imaging logging image by the first convolutional neural network classification model, and obtaining a first reservoir classification result of the target reservoir corresponding to the target depth area in the target wellbore; Inputting the target conventional well logging curve data into a pre-trained second convolutional neural network classification model, processing each of the target well logging data contained in the target conventional well logging curve data by the second convolutional neural network classification model, and obtaining a second reservoir classification result of the target reservoir corresponding to the target depth area in the target wellbore; Inputting a concatenated result obtained by concatenating the first reservoir classification result and the second reservoir classification result into a pre-trained target fully connected layer to obtain a target reservoir classification result of the target reservoir; The category of the target reservoir is determined according to the target reservoir classification result.

2. The reservoir classification method according to claim 1, characterized in that: Before obtaining the target imaging logging image corresponding to the target depth area in the target well, the method further includes: Acquire an initial imaging logging image corresponding to the target wellbore; the initial imaging logging image contains imaging logging information corresponding to each depth in the target wellbore; Segmenting the initial imaging logging image according to a preset depth interval to obtain a plurality of imaging logging images and the target depth region corresponding to each of the imaging logging images; For each of the imaging logging images, grayscale processing is performed on the imaging logging image to obtain the target imaging logging image corresponding to the imaging logging image.

3. The reservoir classification method according to claim 1, characterized in that: Before acquiring the target conventional logging curve data corresponding to the target depth zone position in the target well, the method further includes: Acquire initial conventional logging curve data corresponding to the target well; the initial conventional logging curve data includes initial logging curve data corresponding to a plurality of initial logging curves; the initial logging curve data includes initial curve values ​​corresponding to each depth in the target well; For each of the initial well logging curve data, smoothing and filtering are performed on the initial well logging curve data to obtain first well logging curve data; the first well logging curve data contains first curve values ​​corresponding to each depth in the target well; For each of the first logging curve data, normalize each of the first curve values ​​in the first logging curve data according to the maximum first curve value and the minimum first curve value in the first logging curve data to obtain the second curve value corresponding to each of the first curve values, and determine the second logging curve data and the target logging curve corresponding to the second logging curve data according to the second curve value corresponding to each depth; the value range of the second curve value is 0-1; For the second logging curve data corresponding to each target logging curve, according to each target depth contained in the target depth area, the second curve value corresponding to each target depth is selected from the second logging curve data, and the selected second curve values ​​are averaged to obtain the target curve value corresponding to the target depth area, and the target curve value corresponding to the target depth area is used as the target logging data corresponding to the target logging curve at the target depth area.

4. The reservoir classification method according to claim 1, characterized in that: The first image feature of the target imaging logging image is processed by the first convolutional neural network classification model to obtain a first reservoir classification result of the target reservoir corresponding to the target depth area in the target wellbore, including: Taking the initial image feature matrix corresponding to the target imaging logging image as the input image feature matrix, performing a first convolution process for a first preset number of times on the input image feature matrix, performing a first nonlinear transformation after each first convolution process, and performing a first pooling process on the result obtained by each first nonlinear transformation; The image feature matrix obtained by each first pooling process is used as the input image feature matrix for the next first convolution process; The first convolution processing includes: calculating the dot product of each first receptive field data in the first image feature matrix and the first convolution kernel, obtaining a first eigenvalue corresponding to each first receptive field data, and constructing a second image feature matrix according to each first eigenvalue; the matrix size corresponding to the first receptive field data is the same as the matrix size corresponding to the first convolution kernel; The first nonlinear transformation includes: for each third image feature matrix, inputting each second eigenvalue in the third image feature matrix into a preset activation function to obtain a third eigenvalue corresponding to each second eigenvalue in the third image feature matrix; in the preset activation function, for each second eigenvalue, when the second eigenvalue is greater than 0, determining the second eigenvalue as the third eigenvalue; when the second eigenvalue is not greater than 0, changing the second eigenvalue to obtain the third eigenvalue; and constructing a fourth image feature matrix using the third eigenvalue; The first pooling process includes: extracting a maximum value from each second receptive field data in the fifth image feature matrix, and constructing a sixth image feature matrix according to the maximum value extracted from each second receptive field data; The output image feature matrix obtained by the last first pooling process is subjected to a first full connection process to obtain the first reservoir classification result corresponding to the target imaging logging image.

5. The reservoir classification method according to claim 1, characterized in that: The step of processing each of the target logging data contained in the target conventional logging curve data by the second convolutional neural network classification model to obtain a second reservoir classification result corresponding to the target depth area in the target wellbore includes: Using each of the target logging data included in the target conventional logging curve data as input logging data, performing a second convolution process for a second preset number of times on the input logging data, performing a second nonlinear transformation after each second convolution process, and performing a second pooling process on a result obtained by each second nonlinear transformation; The feature data obtained by each second pooling process is used as the input logging data for the next second convolution process; The second convolution processing includes: dividing all the first well logging data to obtain at least one third receptive field data; for each of the third receptive field data, calculating the dot product of the first well logging data contained in the third receptive field data and the second convolution kernel to obtain the fourth eigenvalue corresponding to the third receptive field data, and constructing the second well logging data according to the fourth eigenvalue; the matrix size corresponding to the third receptive field data is the same as the matrix size corresponding to the second convolution kernel; the second nonlinear transformation includes: for each third well logging data, judging whether the third well logging data is greater than 0; if the third well logging data is greater than 0, determining the third well logging data as the fourth well logging data; if the third well logging data is not greater than 0, determining the product of the third well logging data and a preset multiple as the fourth well logging data; The second pooling processing includes: dividing all the fifth logging data to obtain at least one fourth receptive field data; for each of the fourth receptive field data, extracting the maximum value from the fourth receptive field data, and constructing the sixth logging data according to the maximum value extracted from each of the fourth receptive field data; performing a second full connection processing on the output logging data obtained by the last second pooling processing to obtain the second reservoir classification result corresponding to the target conventional logging curve data.

6. The reservoir classification method according to claim 1, characterized in that: After determining the categories of the target reservoirs corresponding to each of the target depth areas in the target well, the method further includes: For each of the target reservoirs, determining whether the category corresponding to the target reservoir is the same as the category corresponding to other target reservoirs adjacent to the target reservoir; If the category corresponding to the target reservoir is the same as the category corresponding to other target reservoirs adjacent to the target reservoir, the target reservoir and the other target reservoirs adjacent to the target reservoir are merged to obtain a synthetic reservoir.

7. The reservoir classification method according to claim 1, characterized in that: The first convolutional neural network classification model, the second convolutional neural network classification model and the target fully connected layer are trained in the following manner: Constructing training samples for training the first convolutional neural network classification model, the second convolutional neural network classification model and the target fully connected layer; the training samples include: sample imaging logging images and sample conventional logging curve data corresponding to the sample depth area in the sample well, and category labels of sample reservoirs corresponding to the sample depth area; the sample conventional logging curve data includes sample logging data corresponding to multiple sample logging curves at the sample depth area; Inputting the sample imaging logging image into a first initial classification model, processing the second image feature of the sample imaging logging image by the first initial classification model, and obtaining a first prediction result of the sample reservoir; Inputting the sample conventional well logging curve data into a second initial classification model, and processing each of the sample well logging data contained in the sample conventional well logging curve data by the second initial classification model to obtain a second prediction result of the sample reservoir; Inputting a third prediction result obtained by concatenating the first prediction result and the second prediction result of the sample reservoir into an initial fully connected layer, and outputting a prediction classification result of the sample reservoir; Calculating a loss value according to the predicted classification result of the sample reservoir and the category label; The loss value is used to update the parameters in the first initial classification model, the second initial classification model, and the initial fully connected layer, and the updated first initial classification model, the second initial classification model, and the initial fully connected layer are trained again until the training reaches a preset number of training times, thereby obtaining the first convolutional neural network classification model, the second convolutional neural network classification model, and the target fully connected layer.

8. The reservoir classification method according to claim 1, characterized in that: After determining the category of the target reservoir according to the target reservoir classification result, the method further includes: According to the category of the target reservoir, a storage amount of the target substance stored in the target reservoir is estimated.

9. A reservoir classification device, characterized in that: include: The first acquisition module is used to acquire target imaging logging images and target conventional logging curve data corresponding to a target depth area in a target wellbore; The target conventional logging curve data includes target logging data corresponding to a plurality of target logging curves at the target depth region; A first input module is used to input the target imaging logging image into a pre-trained first convolutional neural network classification model, and process the first image feature of the target imaging logging image through the first convolutional neural network classification model to obtain a first reservoir classification result of the target reservoir corresponding to the target depth area in the target wellbore; A second input module is used to input the target conventional well logging curve data into a pre-trained second convolutional neural network classification model, and process each of the target well logging data contained in the target conventional well logging curve data through the second convolutional neural network classification model to obtain a second reservoir classification result of the target reservoir corresponding to the target depth area in the target well; A third input module is used to input a concatenated result obtained by concatenating the first reservoir classification result and the second reservoir classification result into a pre-trained target fully connected layer to obtain a target reservoir classification result of the target reservoir; The first determination module is used to determine the category of the target reservoir according to the classification result of the target reservoir.

10. An electronic device, characterized in that: include: A processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor and the memory communicate via the bus, and when the machine-readable instructions are executed by the processor, the steps of the method as claimed in any one of claims 1 to 8 are performed.

11. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are executed.

Images