Oil reservoir yield prediction method, equipment and product based on adaptive graph convolutional network

Abstract

The embodiment of the invention provides an oil reservoir yield prediction method based on an adaptive graph convolutional network, equipment and a product. The method comprises the steps that first dynamic data of a water injection well and second dynamic data of an oil production well are obtained; then, the dynamic data are jointly input into a pre-trained prediction model, an adjacent matrix capable of representing the space influence relation of the water injection well and the oil production well is built in a feature extraction network part of the model, and therefore the predicted oil yield considering the influence of the water injection well and the factors of the oil production well can be obtained. According to the method, the problem that the prediction accuracy is limited due to the fact that only the attributes of the oil production well are concerned in an existing oil reservoir yield prediction method is solved, and the accuracy of oil reservoir yield prediction is remarkably improved by comprehensively considering the dynamic data of the water injection well and the oil production well and the interaction relation between the water injection well and the oil production well.

Description

Technical Field

[0001] The present application relates to the field of oil production, and in particular to an oil reservoir production prediction method, equipment and product based on an adaptive graph convolutional network. Background Art

[0002] In the field of oil production, water drive oil reservoir production is a commonly used method to increase oil well production. This method injects water or other displacement fluids into the oil reservoir and uses the hydraulic pressure to push the crude oil to the production well, thereby increasing the production of the oil well.

[0003] However, in the process of implementing water-drive oil reservoir exploitation, reservoir production prediction is required in order to more effectively manage resources and optimize exploitation strategies.

[0004] Although the existing reservoir production prediction methods can provide reference for production decisions to a certain extent, they often only consider the properties of the oil wells themselves and have certain limitations, resulting in low accuracy of production prediction. Summary of the invention

[0005] The embodiments of the present application provide a method, device and product for predicting oil reservoir production based on an adaptive graph convolutional network, which is used to solve the technical problem that the existing oil reservoir production prediction only considers the properties of the oil well itself, resulting in reduced accuracy of production prediction.

[0006] In a first aspect, an embodiment of the present application provides a method for predicting oil reservoir production based on an adaptive graph convolutional network, comprising:

[0007] Acquire first dynamic data of the water injection well and second dynamic data of the oil production well, wherein the first dynamic data includes the water injection rate and / or bottom hole flow pressure of the water injection well, and the second dynamic data includes at least one of the oil production rate, water production rate, liquid production rate, and bottom hole flow pressure of the oil production well;

[0008] The first dynamic data and the second dynamic data are input into a pre-trained prediction model to obtain the predicted oil production output by the prediction model, wherein the feature extraction network in the prediction model includes an adaptive graph convolutional network, and the adjacency matrix in the adaptive graph convolutional network is used to characterize the spatial influence relationship between the water injection well and the oil production well.

[0009] In a possible implementation, the prediction model includes an encoding network, the feature extraction network and a decoding network, wherein the feature extraction network includes a plurality of cascaded convolutional layers, each of the convolutional layers includes a timing processing module and the adaptive graph convolutional network;

[0010] The step of inputting the first dynamic data and the second dynamic data into a pre-trained prediction model to obtain a predicted oil production output by the prediction model comprises:

[0011] Inputting the first dynamic data and the second dynamic data into the encoding network to obtain initial features, using the initial features as input vectors of the first convolutional layer among the multiple cascaded convolutional layers, and for any convolutional layer, passing the input vector of the convolutional layer through the timing processing module of the convolutional layer to generate a first feature, then passing the first feature through the adaptive graph convolutional network of the convolutional layer to generate a second feature, and using the second feature as the input vector of the next convolutional layer, until the first features generated by each of the multiple cascaded convolutional layers are obtained;

[0012] The first features generated by each of the multiple cascaded convolutional layers are input into the decoding network to obtain the predicted oil production.

[0013] In a possible implementation, the timing processing module includes a first timing convolutional network and a second timing convolutional network, and the step of generating a first feature by passing the input vector of the convolutional layer through the timing processing module of the convolutional layer, and then generating a second feature by passing the first feature through the adaptive graph convolutional network of the convolutional layer includes:

[0014] Inputting the input vector of the convolutional layer into the first temporal convolutional network to obtain a first sub-feature;

[0015] Inputting the input vector of the convolutional layer into the second temporal convolutional network to obtain a second sub-feature;

[0016] multiplying the first sub-feature and the second sub-feature to obtain the first feature;

[0017] The second feature is generated based on the first feature and an adjacency matrix in the adaptive graph convolutional network.

[0018] In a possible implementation, inputting the first features generated by each of the plurality of cascaded convolutional layers into the decoding network to obtain the predicted oil production includes:

[0019] Accumulating the first features generated by each of the multiple cascaded convolutional layers to obtain a fused feature;

[0020] The fused features are decoded by one or more decoding layers of the decoding network to obtain the predicted oil production.

[0021] In a possible implementation, the adjacency matrix is ​​obtained by training the prediction model based on an initial adjacency matrix, and the initial adjacency matrix is ​​determined based on first static data of the water injection well and second static data of the oil production well, the first static data including the permeability, reservoir thickness and location coordinates of the reservoir at the location of the water injection well, and the second static data including the permeability, reservoir thickness and location coordinates of the reservoir at the location of the oil production well.

[0022] In a second aspect, an embodiment of the present application provides a method for training an oil production prediction model, comprising:

[0023] Acquire sample data, wherein the sample data includes first dynamic sample data of a water injection well, second dynamic sample data of an oil production well, and an oil production label, wherein the first dynamic sample data includes the water injection rate and / or bottom hole flow pressure of the water injection well, and the second dynamic sample data includes at least one of the oil production rate, water production rate, liquid production rate, and bottom hole flow pressure of the oil production well;

[0024] Inputting the first dynamic sample data and the second dynamic sample data into an oil production prediction model to obtain a predicted oil production output by the oil production prediction model, wherein the feature extraction network in the oil production prediction model includes an adaptive graph convolution network, and the initial adjacency matrix in the adaptive graph convolution network is used to characterize the spatial influence relationship between the water injection well and the oil production well;

[0025] The parameters of the oil production prediction model and the initial adjacency matrix are adjusted based on the predicted oil production and the oil production labels.

[0026] In a possible implementation, it further includes:

[0027] Acquire the first static sample data of the water injection well and the second static sample data of the oil production well, wherein the first static sample data includes the permeability, reservoir thickness and position coordinates of the reservoir at the location of the water injection well, and the second static sample data includes the permeability, reservoir thickness and position coordinates of the reservoir at the location of the oil production well;

[0028] The initial adjacency matrix is ​​determined according to the first static sample data and the second static sample data.

[0029] In a third aspect, an embodiment of the present application provides an oil reservoir production prediction device based on an adaptive graph convolutional network, comprising:

[0030] An acquisition module, used for acquiring first dynamic data of the water injection well and second dynamic data of the oil production well, wherein the first dynamic data includes the water injection rate and / or bottom hole flow pressure of the water injection well, and the second dynamic data includes at least one of the oil production rate, water production rate, liquid production rate, and bottom hole flow pressure of the oil production well;

[0031] An input module is used to input the first dynamic data and the second dynamic data into a pre-trained prediction model to obtain the predicted oil production output by the prediction model, wherein the feature extraction network in the prediction model includes an adaptive graph convolutional network, and the adjacency matrix in the adaptive graph convolutional network is used to characterize the spatial influence relationship between the water injection well and the oil production well.

[0032] In a possible implementation, the input module is further used to input the first dynamic data and the second dynamic data into the encoding network to obtain initial features, use the initial features as input vectors of the first convolutional layer in the multiple cascaded convolutional layers, and for any convolutional layer, generate a first feature by passing the input vector of the convolutional layer through the timing processing module of the convolutional layer, then generate a second feature by passing the first feature through the adaptive graph convolutional network of the convolutional layer, and use the second feature as the input vector of the next convolutional layer, until the first features generated by each of the multiple cascaded convolutional layers are obtained;

[0033] The input module is specifically used to input the first features generated by each of the multiple cascaded convolutional layers into the decoding network to obtain the predicted oil production.

[0034] In a possible implementation, the input module is further used to input the input vector of the convolution layer into the first temporal convolutional network to obtain a first sub-feature;

[0035] The input module is further used to input the input vector of the convolution layer into the second temporal convolutional network to obtain a second sub-feature;

[0036] The device also includes: a product module;

[0037] The product module is used to multiply the first sub-feature and the second sub-feature to obtain the first feature;

[0038] The device also includes: a generating module;

[0039] The generation module is used to generate the second feature based on the first feature and the adjacency matrix in the adaptive graph convolutional network.

[0040] In a possible implementation manner, the device further includes: an accumulation module;

[0041] The accumulation module is used to accumulate the first features generated by each of the multiple cascaded convolutional layers to obtain a fused feature;

[0042] The device also includes: a decoding module;

[0043] The decoding module is used to decode the fused features through one or more decoding layers of the decoding network to obtain the predicted oil production.

[0044] In a fourth aspect, an embodiment of the present application provides a training device for an oil production prediction model, comprising:

[0045] An acquisition module, used to acquire sample data, wherein the sample data includes first dynamic sample data of a water injection well, second dynamic sample data of an oil production well, and an oil production label, wherein the first dynamic sample data includes the water injection rate and / or bottom hole flow pressure of the water injection well, and the second dynamic sample data includes at least one of the oil production rate, water production rate, liquid production rate, and bottom hole flow pressure of the oil production well;

[0046] An input module, used for inputting the first dynamic sample data and the second dynamic sample data into an oil production prediction model to obtain a predicted oil production output by the oil production prediction model, wherein the feature extraction network in the oil production prediction model includes an adaptive graph convolution network, and the initial adjacency matrix in the adaptive graph convolution network is used to characterize the spatial influence relationship between the water injection well and the oil production well;

[0047] An adjustment module is used to adjust the parameters of the oil production prediction model and the initial adjacency matrix based on the predicted oil production and the oil production label.

[0048] In a possible implementation, the acquisition module is further used to acquire first static sample data of the water injection well and second static sample data of the oil production well, wherein the first static sample data includes the permeability, reservoir thickness and position coordinates of the reservoir at the location of the water injection well, and the second static sample data includes the permeability, reservoir thickness and position coordinates of the reservoir at the location of the oil production well;

[0049] The device further comprises: a determination module;

[0050] The determination module is used to determine the initial adjacency matrix according to the first static sample data and the second static sample data.

[0051] In a fifth aspect, an embodiment of the present application provides an oil reservoir production prediction device based on an adaptive graph convolutional network, comprising: a memory, a processor;

[0052] The memory stores computer-executable instructions;

[0053] The processor executes the computer-executable instructions stored in the memory, so that the processor executes the above first aspect and / or various possible implementations of the first aspect.

[0054] In a sixth aspect, an embodiment of the present application provides a computer-readable storage medium, in which computer-readable storage medium is stored computer execution instructions, and when the computer execution instructions are executed by a processor, they are used to implement the first aspect above and / or various possible implementation methods of the first aspect.

[0055] In a seventh aspect, an embodiment of the present application provides a computer program product, including a computer program, which, when executed by a processor, implements the above first aspect and / or various possible implementation methods of the first aspect.

[0056] The method, device and product for predicting oil reservoir production based on an adaptive graph convolutional network provided in the embodiment of the present application first obtain the first dynamic data of the water injection well and the second dynamic data of the oil production well. Subsequently, these dynamic data are inputted into a pre-trained prediction model. The feature extraction network part of the model has a built-in adjacency matrix that can characterize the spatial influence relationship between the water injection well and the oil production well, so as to obtain a predicted oil production that takes into account both the influence of the water injection well and the oil production well. This method not only solves the problem of limited prediction accuracy in the existing oil reservoir production prediction method due to only focusing on the properties of the oil production well itself, but also significantly improves the accuracy of oil reservoir production prediction by comprehensively considering the dynamic data of the water injection well and the oil production well and the interaction relationship between them. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0058] Figure 1 Schematic diagram of the process of oil reservoir production prediction method based on adaptive graph convolutional network provided in this application Figure 1 ;

[0059] Figure 2 Schematic diagram of the process of oil reservoir production prediction method based on adaptive graph convolutional network provided in this application Figure 2 ;

[0060] Figure 3 Schematic diagram of the process of oil reservoir production prediction method based on adaptive graph convolutional network provided in this application Figure 3 ;

[0061] Figure 4 Schematic diagram of the training method of the oil production prediction model provided in this application Figure 4 ;

[0062] Figure 5 Schematic diagram of the training method of the oil production prediction model provided in this application Figure 5 ;

[0063] Figure 6 A schematic diagram of the prediction structure of the reservoir production prediction method based on the adaptive graph convolutional network provided in this application;

[0064] Figure 7 A schematic diagram of the structure of an oil reservoir production prediction device based on an adaptive graph convolutional network provided in this application;

[0065] Figure 8 A schematic diagram of the structure of a training device for an oil production prediction model provided in this application;

[0066] Fig. 9 A schematic diagram of the structure of an oil reservoir production prediction device based on an adaptive graph convolutional network provided in this application.

[0067] The above drawings have shown clear embodiments of the present application, which will be described in more detail later. These drawings and text descriptions are not intended to limit the scope of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0068] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0069] First, the terms involved in this application are explained.

[0070] (1) Adaptive graph convolution network (AGCN) is a neural network model specifically designed for processing graph structured data. Its core is the ability to adaptively perform convolution operations based on the topological structure and node features of the graph, thereby effectively extracting feature information from the graph data.

[0071] (2) Temporal Convolutional Network (TCN) is a network architecture that transforms the structure of convolutional neural networks to handle sequence modeling tasks. The function of temporal convolutional networks is to extract temporal features of time series data. The characteristic of temporal convolutional networks is that causal convolution ensures that the prediction of each time point depends only on the current and previous time points. Dilated convolution is used to expand the receptive field and capture the features of long time spans without increasing computational complexity.

[0072] With the continuous advancement of oil extraction technology, water-driven oil reservoir extraction has become a widely used method to increase oil well production. The core of this method is to accurately inject water or other displacement fluids into the oil reservoir, cleverly use the hydraulic pressure as a driving force, drive the crude oil to move effectively towards the production well, and thus achieve a significant increase in oil well production.

[0073] However, in the actual operation of water-driven oil reservoir exploitation, in order to achieve refined management of oil resources and optimization of exploitation strategies, reservoir production prediction is required.

[0074] Although the existing reservoir production prediction methods can provide reference for production decisions to a certain extent, they often only consider the properties of the oil wells themselves and have certain limitations, resulting in low accuracy of production prediction.

[0075] In response to the above problems, the present application provides an oil reservoir production prediction method based on an adaptive graph convolutional network, which first obtains the first dynamic data of the water injection well and the second dynamic data of the oil production well. Subsequently, these dynamic data are input together into a pre-trained prediction model. The feature extraction network part of the model has a built-in adjacency matrix that can characterize the spatial influence relationship between the water injection well and the oil production well, so that a predicted oil production can be obtained that takes into account both the influence of the water injection well and the oil production well. This method not only solves the problem of limited prediction accuracy in existing oil reservoir production prediction methods due to only focusing on the properties of the oil production well itself, but also significantly improves the accuracy of oil reservoir production prediction by comprehensively considering the dynamic data of the water injection well and the oil production well and the interaction between them.

[0076] The technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems are described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0077] Figure 1 Schematic diagram of the process of oil reservoir production prediction method based on adaptive graph convolutional network provided in this application Figure 1 ,like Figure 1 As shown, the method includes:

[0078] S101. Acquire first dynamic data of the water injection well and second dynamic data of the oil production well, wherein the first dynamic data includes the water injection rate and / or bottom hole flow pressure of the water injection well, and the second dynamic data includes at least one of the oil production rate, water production rate, liquid production rate, and bottom hole flow pressure of the oil production well.

[0079] The injection rate of an injection well refers to the amount of water injected into the formation by the injection well per unit time. For example, the injection rate of injection well A is 100 cubic meters of water per day, which means that the well will inject 100 cubic meters of water into the formation within 24 hours.

[0080] The bottom hole pressure of an injection well refers to the fluid pressure at the bottom of the injection well. For example, the bottom hole pressure of injection well A is 20MPa, which means that the fluid injected by the injection well has exerted a pressure of 20MPa on the formation.

[0081] The bottom hole flow pressure of the water injection well is used to characterize the effect of the water injection fluid on the formation pressure. The greater the bottom hole flow pressure of the water injection well, the greater the pressure exerted by the water injection fluid on the formation. Conversely, the smaller the bottom hole flow pressure of the water injection well, the smaller the pressure exerted by the water injection fluid on the formation.

[0082] The oil production rate of an oil well refers to the amount of oil extracted from the formation by the well per unit time. For example, the oil production rate of oil well A is 5 cubic meters of oil per day, which means that the well will extract 5 cubic meters of oil from the formation within 24 hours.

[0083] The water production rate of an oil well refers to the amount of water that the well extracts from the formation per unit time. For example, if the water production rate of oil well A is 10 cubic meters of water per day, it means that the well will extract 10 cubic meters of water from the formation within 24 hours.

[0084] The liquid production rate of an oil well refers to the total volume of liquid (including crude oil and water) extracted from the formation by the oil well per unit time. For example, the liquid production rate of oil well A is 7 cubic meters per day, which means that the well will extract 7 cubic meters of liquid from the formation within 24 hours.

[0085] The bottom hole flow pressure of an oil well refers to the fluid pressure at the bottom of the oil well. For example, the bottom hole flow pressure of oil well A is 15MPa, which means that the pressure on the fluid at the bottom of the oil well is 15MPa.

[0086] The purpose of acquiring the first dynamic data of the water injection well and the second dynamic data of the oil production well in this step is to obtain the overall acquisition situation of the oil production well and the water injection well in the past period of time.

[0087] It is understandable that in the process of exploitation of water-driven oil reservoirs, water injection wells and oil production wells work together. Water injection wells continuously inject water into the formation, effectively increasing the reservoir pressure, forming a water pressure driving force, and causing the crude oil to flow to the oil production well under pressure, and finally be produced to the ground.

[0088] The first dynamic data of the water injection well directly reflects the strength of the water injection capacity of the water injection well, while the second dynamic data of the oil production well reveals the production capacity of the oil production well with the assistance of the water injection well.

[0089] Therefore, by acquiring the first dynamic data of the water injection well and the second dynamic data of the oil production well, the overall production conditions of the oil production well and the water injection well in the past period of time can be more accurately evaluated.

[0090] This step can be, for example, obtaining the first dynamic data of the water injection well and the second dynamic data of the oil production well in the past period of time. For example, assuming that the past 7 days are used as the time range, the system will collect the water injection rate and bottom hole flow pressure of the water injection well, and the oil production rate, water production rate, liquid production rate and bottom hole flow pressure of the oil production well in the past 7 days.

[0091] The method of obtaining the first dynamic data of the water injection well and the second dynamic data of the oil production well in this step can be, for example, obtained from a data management library or from a data management platform of a data recorder, and this application does not impose any special restrictions on this.

[0092] S102. Input the first dynamic data and the second dynamic data into a pre-trained prediction model to obtain the predicted oil production output by the prediction model, wherein the feature extraction network in the prediction model includes an adaptive graph convolutional network, and the adjacency matrix in the adaptive graph convolutional network is used to characterize the spatial influence relationship between water injection wells and oil production wells.

[0093] The purpose of this step is to obtain a predicted oil production that takes into account both the impact of water injection wells and oil production wells.

[0094] It is understandable that the feature extraction network part of the prediction model uses an adaptive graph convolutional network that can characterize the spatial influence relationship between water injection wells and oil production wells. The core of the adaptive graph convolutional network is that it can effectively propagate and update the feature information on the nodes based on the adjacency matrix. Through the unique point-edge adjacency convolution operation, the network can deeply fuse the correlation information between the first dynamic data of the water injection well and the second dynamic data of the oil production well, thereby capturing the complex interaction between the two.

[0095] Therefore, when the system inputs the first dynamic data of the water injection well and the second dynamic data of the oil production well into this pre-trained prediction model, the model can make full use of the spatial influence relationship between the water injection well and the oil production well and the water injection capacity and production capacity information contained in the dynamic data to predict future oil production.

[0096] For example, assuming that there are 10 oil production wells and 5 water injection wells with their respective dynamic data in the past 10 days, based on the above information, the system first inputs the dynamic data of these 10 days into the prediction model. Subsequently, the prediction model can use its built-in adjacency matrix for characterizing the spatial influence relationship between water injection wells and oil production wells to conduct a comprehensive analysis of the dynamic data. Finally, the prediction model outputs the predicted oil production on the 11th day.

[0097] Optionally, the present application provides a possible implementation process of the adjacency matrix, including: the adjacency matrix is ​​obtained by training the prediction model based on the initial adjacency matrix, the initial adjacency matrix is ​​determined based on the first static data of the water injection well and the second static data of the oil production well, the first static data includes the permeability, reservoir thickness and location coordinates of the reservoir where the water injection well is located, and the second static data includes the permeability, reservoir thickness and location coordinates of the reservoir where the oil production well is located.

[0098] The oil reservoir production prediction method based on the adaptive graph convolutional network provided in the embodiment of the present application inputs the dynamic data of the water injection well and the oil production well into a pre-trained prediction model, thereby obtaining the predicted oil production that takes into account both the influence of the water injection well and the factors of the oil production well. This method successfully solves the technical problem of insufficient prediction accuracy in the existing oil reservoir production prediction method due to being limited to considering the properties of the oil production well itself, thereby significantly improving the accuracy of oil reservoir production prediction, thereby providing more reliable data support for oil production decision-making.

[0099] Figure 2 Schematic diagram of the process of oil reservoir production prediction method based on adaptive graph convolutional network provided in this application Figure 2 ,like Figure 2 As shown, in this embodiment Figure 1 Based on the embodiment, a method for predicting oil reservoir production based on an adaptive graph convolutional network is described in detail. The method includes:

[0100] S201. Acquire first dynamic data of the water injection well and second dynamic data of the oil production well, wherein the first dynamic data includes the water injection rate and / or bottom hole flow pressure of the water injection well, and the second dynamic data includes at least one of the oil production rate, water production rate, liquid production rate, and bottom hole flow pressure of the oil production well.

[0101] The explanation of step S201 is similar to that of the above step S101 and will not be repeated here.

[0102] S202. Input the first dynamic data and the second dynamic data into the encoding network to obtain initial features, and use the initial features as input vectors of the first convolutional layer in multiple cascaded convolutional layers. For any convolutional layer, pass the input vector of the convolutional layer through the timing processing module of the convolutional layer to generate a first feature, and then pass the first feature through the adaptive graph convolutional network of the convolutional layer to generate a second feature, and use the second feature as the input vector of the next convolutional layer, until the first features generated by each of the multiple cascaded convolutional layers are obtained, wherein the prediction model includes an encoding network, a feature extraction network and a decoding network, the feature extraction network includes multiple cascaded convolutional layers, and each convolutional layer includes a timing processing module and an adaptive graph convolutional network.

[0103] Among them, the encoding network is constructed by cascading multiple one-dimensional convolutional layers.

[0104] The encoding process of the encoding network is the process of linearly transforming the feature vector at each position of the input data. Through convolution operations and feature extraction, the encoding network can extract the intrinsic features of the input data and convert the input data into more abstract feature vectors. These feature vectors are aggregated in the spatial and temporal dimensions to form the output feature map of the encoding network.

[0105] The encoding network processes the input data as follows:

[0106] The first step is input data parsing: the encoding network receives a multi-dimensional input tensor, which contains multiple dimensions such as batch size, input height, input width, and number of input channels.

[0107] The second step is linear transformation: the encoding network extracts features from each position of the input data through convolution operations. The convolution operation slides the convolution kernel on the input tensor, performs weighted summation of the feature vectors at each position, and adds a bias term to obtain a new feature vector.

[0108] The third step is feature aggregation: After the convolution operation, the feature vector at each position is converted into a new, more abstract feature vector. These feature vectors are aggregated in the spatial and temporal dimensions to form the output feature map of the encoding network. The output feature map reflects the abstract representation of the input data in different feature dimensions.

[0109] Step 4: Feature vector output: The encoding network converts the output feature map into a series of feature vectors, which can be further processed and analyzed by the subsequent feature extraction network.

[0110] The input dimensions of the encoding network include: input tensor, convolution kernel size, and initial features of the input data. Among them, (1) the format of the input tensor is: ,in, represents the input tensor; Indicates the batch size, which is determined based on human experience, for example, it can be 1; Indicates input height, input height Used to indicate the length of the historical time range for which the performance data of each of the water injection well and the oil production well are used; Represents input width, and input width N is used to indicate the total number of water injection wells and oil production wells; Indicates the number of input channels. It is used to indicate the data types that constitute the first dynamic data of the water injection well and the second dynamic data of the oil production well. (2) The format of the convolution kernel is: ,in, represents the convolution kernel, represents the number of output channels, that is, the dimension of the feature vector. (3) The format of the initial feature is ,in, The vector format representing the initial features.

[0111] The purpose of using the encoding network to process the first dynamic data and the second dynamic data in this step is to convert the first dynamic data and the second dynamic data into their respective corresponding feature vectors. These feature vectors not only represent a higher level and more abstract data representation, but also can accurately capture the key information in the original dynamic data.

[0112] The purpose of using the initial features as the input vector of the first convolutional layer in multiple cascaded convolutional layers is to gradually extract and fuse the feature information of the data at different levels to form a richer, more comprehensive and abstract feature representation.

[0113] It can be understood that after the dynamic data of the water injection well and the oil production well are converted into corresponding initial features by the encoding network, these initial features are sent to the feature extraction network for further feature extraction. The feature extraction network consists of multiple cascaded convolutional layers, each of which contains a time series processing module and an adaptive graph convolutional network. First, the initial features are input into the first convolutional layer, and the time series processing module of this layer can capture the dependency and dynamic changes of the data in the time series dimension, thereby generating the first feature. Subsequently, the first feature is sent to the adaptive graph convolutional network, which can adaptively use the adjacency matrix to capture the correlation and feature interaction of the first feature in the spatial dimension, generating a more abstract and rich second feature. This process is continuously iterated in multiple cascaded convolutional layers, and each convolutional layer takes the second feature generated by the previous convolutional layer as input and outputs its own first feature. In this way, the feature extraction network can gradually extract and fuse the feature information of the data at different levels to form a richer, more comprehensive and abstract feature representation.

[0114] Optionally, the present application provides an implementation process of inputting the first dynamic data and the second dynamic data into the encoding network to obtain the initial features, which specifically includes:

[0115] In the first step, the first dynamic data of the water injection well and the second dynamic data of the oil production well are merged to form a unified input data set.

[0116] The second step is to construct a multi-dimensional input tensor based on the merged input data set. The format of the input tensor includes: batch size B, input height T, input width N, and number of input channels. .

[0117] The third step is to determine the convolution kernel size.

[0118] The fourth step is to use the prepared convolution kernel to perform a convolution operation on the input tensor, that is, to perform a weighted summation of the feature vectors at each position and add a bias term to obtain a new feature vector.

[0119] The fifth step is to output the new feature vector to the feature extraction network part of the prediction model so that the feature extraction network can further process and analyze it.

[0120] For example, assuming the batch size is 1, there is 1 water injection well and 2 oil production wells, and three days of dynamic data are used to predict the future. The water injection well and the oil production well each have two types of data. The data of the water injection well is the water injection rate and the bottom hole flow pressure, and the data of the oil production well is the oil production and the bottom hole flow pressure. Based on the above information, the batch size can be determined first. is 1, enter the height is 3, the input width N is 3 and the number of input channels is 2, then, we can determine the input tensor is , the convolution kernel size is Finally, the format of the initial features is determined as .

[0121] S203, inputting the first features generated by each of the multiple cascaded convolutional layers into a decoding network to obtain predicted oil production.

[0122] The purpose of this step is to enable the decoding network to make full use of the first features at different levels and output a more accurate oil production value.

[0123] It is understandable that each convolutional layer can learn different feature representations from the input data, and as the convolutional layer goes deeper, these features gradually become more abstract and complex in terms of hierarchy. For example, shallow convolutional layers can usually capture the basic features and edge information in the input data, while deep convolutional layers can further extract more advanced and complex features.

[0124] Therefore, inputting the first features generated by each of the multiple cascaded convolutional layers into the decoding network can enable the decoding network to make full use of the feature information at different levels, thereby outputting a more accurate prediction of oil production.

[0125] Optionally, the present application provides a possible implementation method, which specifically includes:

[0126] In the first step, the first features generated by multiple cascaded convolutional layers are accumulated to obtain fused features.

[0127] It is understandable that each convolutional layer has its own unique perspective and focus when extracting features. Shallow convolutional layers can often capture basic and local features in the input data, while deep convolutional layers can extract more abstract and global features. Therefore, accumulating these different levels of features allows the decoding network of the prediction model to simultaneously utilize these different levels of feature information, thereby enhancing the expressiveness and predictive ability of the prediction model.

[0128] Optionally, the present application provides a possible implementation method, which specifically includes: using matrix addition operations to accumulate the first features generated by each of multiple cascaded convolutional layers to obtain a fused feature.

[0129] The matrix addition algorithm operation refers to two matrices with the same dimension generating another matrix with the same dimension, wherein each element is the sum of the elements (i, j) of the original two matrices.

[0130] In the first step, the zero matrix is ​​initialized and used as the initial value of the fused feature to ensure that the dimension of the zero matrix is ​​the same as that of each first feature.

[0131] In the second step, each first feature matrix is ​​traversed and added to the fused feature matrix one by one using matrix addition operation until the fused features are obtained.

[0132] In the second step, the fused features are decoded through one or more decoding layers of the decoding network to obtain the predicted oil production.

[0133] In each decoding layer, there is a ReLU activation function and a convolution operation. The ReLU activation function (Rectified Linear Unit) is a commonly used artificial neural network activation function. The calculation method of the ReLU activation function is:

[0134]

[0135] in, represents the fusion feature. This means that if the input If it is greater than 0, the output is itself; if the input If it is less than or equal to 0, the output is 0.

[0136] The purpose of decoding the fused features through one or more decoding layers of the decoding network is to gradually transform the high-level abstract information contained in the fused features into specific and explainable prediction results, that is, predicting oil production.

[0137] It is understandable that the role of the decoding network is to restore the compact, high-dimensional feature representation obtained in the encoding stage back to the original, easy-to-understand dimensional space, and use these features to generate the final prediction value. This structured decoding process helps the prediction model to more accurately capture and utilize the information in the fused features, thereby improving the accuracy of the prediction.

[0138] Optionally, in the case where the decoding network has two decoding layers, the present application provides a possible implementation method, including:

[0139] In the first step, the fused features are input into the first decoding layer, and the ReLU activation function in the first decoding layer is used to perform nonlinear activation processing on the fused features to obtain activated features after nonlinear transformation.

[0140] The purpose of this step is to increase the nonlinear expression ability of the prediction model so that the prediction model can learn more complex feature relationships.

[0141] In the second step, a convolution operation is used to perform convolution processing on the nonlinearly transformed activation features obtained in the first step to obtain the first decoding vector output by the first decoding layer.

[0142] In the third step, the first decoding vector output by the first decoding layer is input into the second decoding layer, and the ReLU activation function is used again to perform nonlinear activation processing on the first decoding vector output by the first decoding layer to obtain the activated features after the nonlinear transformation again.

[0143] The purpose of this step is to further increase the nonlinear expression ability of the model and make the decoded vector more in line with the requirements of the prediction target.

[0144] In the fourth step, the convolution operation is used again to convolve the activated features of the third step with the nonlinear transformation again to obtain the predicted oil production output by the first decoding layer.

[0145] The method for predicting oil reservoir production based on an adaptive graph convolutional network provided in an embodiment of the present application first obtains the first dynamic data of the water injection well and the second dynamic data of the oil production well. Subsequently, these dynamic data are jointly input into a pre-trained prediction model, which consists of three parts: an encoding network, a feature extraction network, and a decoding network. In the feature extraction stage, the initial features are generated by the encoding network and used as the input vector of the first convolutional layer in multiple cascaded convolutional layers. Each convolutional layer is embedded with a timing processing module and an adaptive graph convolutional network. The former is responsible for extracting the first feature from the input vector, and the latter further uses graph convolution technology to process these features and generate a second feature containing a spatial influence relationship. This processing process is passed step by step until all cascaded convolutional layers each generate their corresponding first features. Finally, these first features are sent to the decoding network, and after decoding processing, an accurate predicted oil production is output. This method not only effectively solves the problem of limited prediction accuracy in existing reservoir production predictions due to only considering the properties of the oil wells themselves, but also significantly improves the accuracy and reliability of the prediction by integrating the dynamic data of injection wells and oil wells and their spatial influence relationship, thereby providing reliable data support for oil extraction.

[0146] Figure 3 Schematic diagram of the process of oil reservoir production prediction method based on adaptive graph convolutional network provided in this application Figure 3 ,like Figure 3 As shown, in this embodiment Figure 2 Based on the embodiment, the process of generating a first feature from the input vector of the convolution layer through the time series processing module of the convolution layer, and then generating a second feature from the first feature through the adaptive graph convolution network of the convolution layer is described in detail. The method includes:

[0147] S301. Input the input vector of the convolution layer into the first temporal convolutional network to obtain a first sub-feature, wherein the temporal processing module includes the first temporal convolutional network and the second temporal convolutional network.

[0148] Among them, the calculation process of the first temporal convolutional network can be expressed by the following formula:

[0149]

[0150] in, represents the value of the input time series (encoded vector) at time t, represents the weight (value) of the convolution kernel, represents the size of the convolution kernel, represents the output vector after the convolution operation, Represents the coefficient of expansion.

[0151] The purpose of this step is to extract preliminary feature representation from the input vector of the convolutional layer.

[0152] It is understandable that since the input vectors of the convolutional layer are converted by the encoding network, these input vectors already contain the encoding information of the original data (i.e., the first dynamic data of the water injection well and the second dynamic data of the oil production well). The first temporal convolutional network is good at processing data with time or sequence characteristics, and it can effectively capture the temporal dependencies in the data.

[0153] Therefore, by inputting the input vector of the convolutional layer into the first temporal convolutional network, the network can make full use of its temporal characteristics to further mine and extract richer and more detailed feature information from the input vector.

[0154] Optionally, after the input vector of the convolution layer is input into the first temporal convolutional network to obtain the first sub-feature, in order to further improve the expression ability of the first sub-feature, the present application also needs to use a sigmoid activation function to perform a nonlinear transformation on the first sub-feature to obtain a more accurate first sub-feature.

[0155] Among them, the Sigmoid activation function is a mathematical function that maps any real number input to the (0,1) interval.

[0156] The calculation method of Sigmoid activation function is:

[0157]

[0158] in, represents the output of the first temporal convolutional network, that is, the first sub-feature, Represents the output of the Sigmoid activation function. The sigmoid activation function has a smooth S-shaped curve. When the input value is small, the output value is close to 0; when the input value is large, the output value is close to 1.

[0159] It is understandable that in the feature extraction network of the prediction model, if only linear transformation is used, then no matter how many layers the feature extraction network has, its output can be simplified to a combination of linear functions, which limits the feature extraction network's ability to model complex nonlinear relationships. The sigmoid activation function can introduce nonlinear factors into the feature extraction network of the prediction model, which can enhance the expression ability of the first sub-feature.

[0160] Therefore, after the input vector of the convolution layer is input into the first temporal convolutional network to obtain the first sub-feature, the first sub-feature needs to be processed using the Sigmoid activation function to improve the accuracy of the first sub-feature.

[0161] S302: Input the input of the convolutional layer into a second temporal convolutional network to obtain a second sub-feature.

[0162] Among them, the calculation process of the second temporal convolutional network can be expressed by the following formula:

[0163]

[0164] in, Represents the input time series (encoded vector) at time The value of represents the weight (value) of the convolution kernel, represents the size of the convolution kernel, represents the output vector after the convolution operation, Represents the coefficient of expansion.

[0165] The purpose of this step is to extract different feature representations in the input vector through different temporal convolutional networks (i.e., the second temporal convolutional network), thereby enhancing the diversity and accuracy of the prediction model.

[0166] It is understandable that although the first temporal convolutional network can initially extract feature representations from the input vector, given the complexity and diversity of the input vector, a single network often cannot fully capture all key information. Therefore, by introducing the second temporal convolutional network, it is possible to go deep into different levels and capture more detailed and rich temporal dependencies and features, thereby mining more potential feature information from the input vector.

[0167] Optionally, after the input of the convolutional layer is input into the second temporal convolutional network to obtain the second sub-feature, in order to further improve the expression ability of the second sub-feature, the present application also needs to use the tanh activation function to perform a nonlinear transformation on the first sub-feature to obtain a more accurate first sub-feature.

[0168] Among them, the tanh activation function is a mathematical function that maps any real number input to the (-1, 1) interval.

[0169] The calculation method of the tanh activation function is:

[0170]

[0171] in, represents the output of the second temporal convolutional network, that is, the second sub-feature, Represents the output of the tanh activation function. The tanh activation function has a smooth curve shape. When the input value x is small, the output value is close to -1; when the input value x is large, the output value is close to 1; and when the input value x is 0, the output value is exactly 0.

[0172] It can be understood that, first of all, the tanh activation function can introduce stronger nonlinearity, which can enable the feature extraction network of the prediction model to capture and represent more complex data relationships. Secondly, the output of the tanh activation function is symmetric about the origin, which helps the feature extraction network of the prediction model to converge faster during training and may reduce the problem of gradient disappearance or explosion, thereby improving the stability and performance of the prediction model. Therefore, after the input of the convolutional layer is input into the second temporal convolutional network to obtain the second sub-feature, the tanh activation function can be used to process the second sub-feature to enhance the expressive power of the second sub-feature, thereby making the second sub-feature more accurate.

[0173] S303 , multiply the first sub-feature and the second sub-feature to obtain a first feature.

[0174] The purpose of this step is to fuse the information contained in the first sub-feature and the second sub-feature and generate a new feature.

[0175] It can be understood that the first sub-feature and the second sub-feature represent different aspects or attributes of the input vector respectively. Therefore, by multiplying the two sub-features, a new feature representation is actually created. This new feature not only contains the information of the first sub-feature, but also the information of the second sub-feature, thereby realizing the fusion of the two sub-feature information.

[0176] Optionally, the present application provides a possible implementation method, including: using a Hadamard product algorithm to multiply the first sub-feature and the second sub-feature to obtain the first feature.

[0177] Among them, the Hadamard product is a type of matrix operation that operates on matrices of the same shape and produces a third matrix of the same dimension. In mathematics, the Hadamard product is a binary operation that uses two matrices of the same dimension to produce another matrix of the same dimension, where each element (i, j) is the product of the elements (i, j) of the original two matrices.

[0178] By using the Hadamard product algorithm, the product of the first sub-feature and the second sub-feature can be more refined and effectively fused in the fusion process, thereby helping to obtain a richer and more expressive first feature.

[0179] It can be understood that in the process of using the Hadamard product algorithm, first, it is determined that the first sub-feature and the second sub-feature have the same dimension; then, the elements at corresponding positions in the first sub-feature and the second sub-feature are multiplied element by element; then, the multiplication result is used as a new feature matrix, that is, the first feature; finally, the first feature is input into the feature extraction network of the prediction model so that the prediction model can be further processed.

[0180] S304: Generate a second feature based on the first feature and the adjacency matrix in the adaptive graph convolutional network.

[0181] The purpose of this step is to utilize the adaptive characteristics of the adjacency matrix in the adaptive graph convolutional network to dynamically adapt and process the first feature.

[0182] It can be understood that the adjacency matrix in the adaptive graph convolutional network not only represents the spatial influence relationship between the water injection well and the oil production well, but also allows it to be adaptively adjusted according to the first feature. Therefore, by combining the first feature and the adjacency matrix, the adjacency matrix can capture the dependency relationship between the water injection well, the oil production well, and the first feature, so that these relationships can be encoded into the second feature. In this way, the second feature not only contains the information of the first feature, but also incorporates the spatial influence relationship between the water injection well and the oil production well.

[0183] The method for predicting oil reservoir production based on an adaptive graph convolutional network provided in an embodiment of the present application, in the processing flow of the convolutional layer, firstly sends the input vector to the first temporal convolutional network and the second temporal convolutional network respectively. The first temporal convolutional network is responsible for extracting the preliminary temporal features of the input vector and generating the first sub-feature; while the second temporal convolutional network further analyzes the temporal information of the input vector and generates the second sub-feature. Subsequently, the two sub-features are multiplied to fuse the temporal characteristics captured by each of them, thereby obtaining a richer and more comprehensive first feature. Next, based on this first feature and the adjacency matrix in the adaptive graph convolutional network, a graph convolution operation is performed, fully considering the spatial influence relationship between the injection well and the oil production well, and finally generating a second feature containing temporal and spatial information. This method not only effectively integrates temporal and spatial features, but also significantly improves the accuracy and efficiency of feature extraction, providing more reliable and rich feature inputs for subsequent prediction tasks.

[0184] Figure 4 Schematic diagram of the training method of the oil production prediction model provided in this application Figure 4 ,like Figure 4 As shown, the method includes:

[0185] S401. Obtain sample data, wherein the sample data includes first dynamic sample data of a water injection well, second dynamic sample data of an oil production well, and an oil production label, wherein the first dynamic sample data includes the water injection rate and / or bottom hole flow pressure of the water injection well, and the second dynamic sample data includes at least one of the oil production rate, water production rate, liquid production rate, and bottom hole flow pressure of the oil production well.

[0186] The purpose of this step is to collect and prepare sample data required for training and optimizing the oil production prediction model.

[0187] It can be understood that the first dynamic sample data of the water injection well is used to describe the working state of the water injection well in the actual mining environment and reflect the pressure changes of the underground fluid. The second dynamic sample data of the oil production well is used to characterize the production performance of the oil production well in the actual mining environment and reveal the flow state of the underground fluid. The oil production label represents the specific value of the oil production in the actual mining environment. It is used as the target value of the oil production prediction model to guide the training and evaluation of the model.

[0188] Therefore, by obtaining sample data, it is possible to provide rich and comprehensive training samples for the oil production prediction model.

[0189] S402. Input the first dynamic sample data and the second dynamic sample data into the oil production prediction model to obtain the predicted oil production output by the oil production prediction model, wherein the feature extraction network in the oil production prediction model includes an adaptive graph convolutional network, and the initial adjacency matrix in the adaptive graph convolutional network is used to characterize the spatial influence relationship between water injection wells and oil production wells.

[0190] The purpose of this step is to use the information contained in the first dynamic sample data and the second dynamic sample data to train and optimize the oil production prediction model, so that it can learn the relationship between the dynamic characteristics of water injection wells and oil production wells and oil production.

[0191] It can be understood that the first dynamic sample data and the second dynamic sample data not only reflect the working status and production performance of the water injection well and the oil production well in the actual production environment, but also reveal the pressure change and flow status of the underground fluid.

[0192] The feature extraction network in the oil production prediction model has a built-in initial adjacency matrix, which can characterize the spatial influence relationship between injection wells and oil production wells. It has adaptive characteristics and can dynamically adjust itself according to other data.

[0193] Therefore, by using the first dynamic sample data and the second dynamic sample data as input factors and inputting them into the oil production prediction model, it means that the oil production prediction model can make full use of the adaptive characteristics of the feature extraction network, and combine the first dynamic sample data and the second dynamic sample data to train and optimize the oil production prediction model, thereby generating a predicted oil production that is more in line with the actual mining environment.

[0194] S403: Adjust the parameters and initial adjacency matrix of the oil production prediction model based on the predicted oil production and the oil production label.

[0195] The purpose of this step is to evaluate the accuracy of the oil production prediction model by comparing the predicted oil production with the actual oil production labels, and adjust the parameters and initial adjacency matrix of the oil production prediction model accordingly.

[0196] It can be understood that the predicted oil production output by the oil production prediction model is first compared with the actual oil production label to evaluate the prediction accuracy of the oil production prediction model. This comparison process can reveal the deviations and errors in the prediction of the oil production prediction model. Subsequently, the parameters and initial adjacency matrix of the oil production prediction model are adjusted according to these deviations and errors. Parameter adjustment includes but is not limited to increasing the weights, biases and other internal parameters of the model to optimize the prediction performance of the oil production prediction model. The adjustment of the initial adjacency matrix may involve changing the representation of the spatial influence relationship between wells to more accurately reflect the interaction between wells in the actual production environment. Through this series of adjustments and optimizations, the prediction accuracy of the oil production prediction model can be gradually improved to make it more in line with the needs of the actual production environment.

[0197] The training method of the oil production prediction model provided by the embodiment of the present application first obtains sample data, which includes the first dynamic sample data of the water injection well and the second dynamic sample data of the oil production well, as well as the corresponding oil production label. Then, these dynamic sample data are input into the oil production prediction model, and the feature extraction network of the model adopts an adaptive graph convolutional network, and its initial adjacency matrix is ​​used to characterize the spatial influence relationship between the water injection well and the oil production well. Through the operation of the model, the predicted oil production can be obtained. In order to continuously improve the prediction accuracy of the model, it is also necessary to adjust and optimize the parameters and initial adjacency matrix of the oil production prediction model based on the predicted oil production and the actual oil production label. This method effectively solves the technical problem that only the oil production well's own attributes are considered in the existing oil reservoir production prediction, while ignoring the spatial influence relationship between the water injection well and the oil production well, thereby reducing the accuracy of the production prediction. In addition, by introducing the initial adjacency matrix of the adaptive graph convolutional network, the method can more comprehensively capture the dynamic changes in the oil reservoir, thereby improving the accuracy of the oil production prediction.

[0198] Figure 5 Schematic diagram of the training method of the oil production prediction model provided in this application Figure 5 ,like Figure 5 As shown, in this embodiment Figure 4 Based on the embodiment, a training method for an oil production prediction model is described in detail, and the method includes:

[0199] S501. Obtain sample data, wherein the sample data includes first dynamic sample data of a water injection well, second dynamic sample data of an oil production well, and an oil production label, wherein the first dynamic sample data includes the water injection rate and / or bottom hole flow pressure of the water injection well, and the second dynamic sample data includes at least one of the oil production rate, water production rate, liquid production rate, and bottom hole flow pressure of the oil production well.

[0200] The explanation of step S501 is similar to that of step S401, and will not be repeated here.

[0201] S502. Input the first dynamic sample data and the second dynamic sample data into the oil production prediction model to obtain the predicted oil production output by the oil production prediction model, wherein the feature extraction network in the oil production prediction model includes an adaptive graph convolutional network, and the initial adjacency matrix in the adaptive graph convolutional network is used to characterize the spatial influence relationship between water injection wells and oil production wells.

[0202] The explanation of step S502 is similar to that of step S402, and will not be repeated here.

[0203] S503. Obtain first static sample data of the water injection well and second static sample data of the oil production well, wherein the first static sample data includes the permeability, reservoir thickness and position coordinates of the reservoir at the location of the water injection well, and the second static sample data includes the permeability, reservoir thickness and position coordinates of the reservoir at the location of the oil production well.

[0204] Among them, the permeability of the reservoir is a physical quantity that describes the ability of rock to allow fluid to pass through its pores. The unit is usually Darcy (D) or millidarcy (mD). The larger the value, the smaller the resistance to fluid flow in the rock.

[0205] The permeability of the reservoirs where the water injection wells and oil production wells are located at different locations is different. For example, when the water injection well is at position A, the permeability of the reservoir at its location is 45mD, and when the water injection well is at position B, the permeability of the reservoir at its location is 23mD. Similarly, when the oil production well is at position C, the permeability of the reservoir at its location is 14mD, and when the oil production well is at position D, the permeability of the reservoir at its location is 25mD.

[0206] Reservoir thickness refers to the thickness of reservoir rock in the vertical direction, which reflects the scale of reservoir rock and its ability to store fluids, and is usually expressed in meters (m) or feet (ft).

[0207] The reservoir thickness of the reservoirs where the water injection wells and oil production wells are located at different locations is different. For example, when the water injection well is at position A, the reservoir thickness at its location is 45m, and when the water injection well is at position B, the reservoir thickness at its location is 40m. Similarly, when the oil production well is at position C, the reservoir thickness at its location is 38m, and when the oil production well is at position D, the reservoir thickness at its location is 42m.

[0208] The location coordinates refer to the specific location of the water injection well or oil production well in the formation.

[0209] By acquiring the first static sample data of the water injection well and the second static sample data of the oil production well, it is possible to more accurately evaluate the impact of the geological conditions at the locations of the water injection well and the oil production well on the mining work.

[0210] It is understandable that, first of all, the permeability of the reservoir will not only directly affect the water injection effect of the water injection well, but also have a significant impact on the production efficiency of the oil production well. For example, a reservoir with high permeability is easier for fluids (including injected water and crude oil in the formation) to flow, thereby improving the water injection efficiency of the water injection well and the oil production efficiency of the oil production well.

[0211] Secondly, the reservoir thickness not only determines how large the reservoir area the water injected by the injection well can penetrate, but also determines how large the reservoir area the oil well can effectively extract crude oil from.

[0212] In addition, the location coordinates reflect not only the actual location of the water injection well in the formation, but also the actual location of the oil production well in the formation.

[0213] Therefore, by acquiring the first static sample data of the water injection well and the second static sample data of the oil production well, the production work of the water injection well and the oil production well can be accurately evaluated according to the geological characteristics of their locations.

[0214] S504: Determine an initial adjacency matrix according to the first static sample data and the second static sample data.

[0215] The purpose of this step is to obtain a matrix that can represent the spatial influence relationship between water injection wells and oil production wells.

[0216] It can be understood that the first static sample data of the water injection well is used to characterize the physical properties of the reservoir where the water injection well is located. These data include the permeability of the reservoir, which reflects the difficulty of fluid passing through the reservoir rock; the reservoir thickness, which indicates the scale of the reservoir rock in the vertical direction; and the location coordinates, which accurately locate the location of the water injection well in the formation. These parameters together constitute a comprehensive description of the geological characteristics of the reservoir where the water injection well is located.

[0217] The second static sample data of the oil well is used to characterize the production potential of the reservoir where the oil well is located. Similarly, these data also include key parameters such as reservoir permeability, reservoir thickness and location coordinates, which together constitute a comprehensive description of the geological characteristics of the reservoir where the oil well is located.

[0218] Therefore, by comprehensively considering the first static sample data of the water injection well and the second static data sample of the oil production well, it means that the intrinsic relationship between these data can be further analyzed, so as to construct an initial adjacency matrix that can represent the spatial influence relationship between the water injection well and the oil production well.

[0219] Optionally, the present application provides a possible implementation method, including: using the following seepage resistance coefficient formula, combining the first static sample data and the second static sample data, to determine the initial adjacency matrix.

[0220]

[0221] in, It represents the interwell seepage resistance coefficient between the water injection well and the oil production well, which is a dimensionless quantity; is the crude oil viscosity; Indicates the distance between the water injection well and the oil production well; is the permeability at the location of injection well i, is the permeability at the location of oil well j; is the oil layer thickness at the location of injection well i, is the oil layer thickness at the location of oil well j.

[0222] The initial adjacency matrix is Each element in the initial adjacency matrix represents the inverse of the seepage resistance coefficient between the water injection well and the oil production well. The larger the value of this element, the stronger the well connectivity based on geological data and the greater the impact of the water injection well on the oil well. The expression is as follows:

[0223]

[0224] in, It represents the interwell seepage resistance coefficient between the first oil production well and the first water injection well; It represents the interwell seepage resistance coefficient between the first oil production well and the nth water injection well; It represents the interwell seepage resistance coefficient between the mth oil production well and the first water injection well; It represents the interwell seepage resistance coefficient between the mth oil production well and the nth water injection well.

[0225] For example, there are 4 water injection wells and 5 oil production wells. Based on the above information, first, it is necessary to obtain the first static sample data corresponding to each of the 4 water injection wells and the second static sample data corresponding to each of the 5 oil production wells. Secondly, the initial adjacency matrix can be determined based on the seepage resistance coefficient formula and in combination with the first static sample data and the second static sample data. The initial adjacency matrix is ​​expressed as:

[0226]

[0227] Among them, the first row of data represents the inter-well seepage resistance coefficient between the first water injection well and the five oil production wells; the second row of data represents the inter-well seepage resistance coefficient between the second water injection well and the five oil production wells; the third row of data represents the inter-well seepage resistance coefficient between the third water injection well and the five oil production wells; the fourth row of data represents the inter-well seepage resistance coefficient between the second water injection well and the five oil production wells.

[0228] S505. Adjust the parameters and initial adjacency matrix of the oil production prediction model based on the predicted oil production and the oil production label.

[0229] The explanation of step S505 is similar to that of step S403, and will not be repeated here.

[0230] The training method of the oil production prediction model provided in the embodiment of the present application first obtains sample data, which covers the first dynamic sample data of the water injection well, the second dynamic sample data of the oil production well, and the oil production label. At the same time, the first static sample data of the water injection well and the second static sample data of the oil production well are obtained. Subsequently, the initial adjacency matrix is ​​determined based on the first static sample data and the second static sample data, and the initial adjacency matrix is ​​embedded in the feature extraction network of the oil production prediction model. Finally, these dynamic sample data are input into the oil production prediction model, and the prediction model can predict the oil production in combination with the initial adjacency matrix and the dynamic sample data. Finally, the predicted oil production output by the oil production prediction model is compared with the actual oil production label, and the parameters of the oil production prediction model and the initial adjacency matrix are adjusted according to the comparison results to improve the prediction accuracy of the model.

[0231] This method comprehensively considers the dynamic and static properties of water injection wells and oil production wells, as well as the spatial relationship between them, and effectively solves the technical problems of existing reservoir production prediction that only considers the properties of oil production wells themselves, ignores the spatial influence relationship between water injection wells and oil production wells, and does not fully utilize the static properties of water injection wells and oil production wells, resulting in reduced production prediction accuracy. It thus significantly improves the accuracy of oil production prediction.

[0232] Figure 6 A schematic diagram of the prediction structure of the reservoir production prediction method based on the adaptive graph convolutional network provided in this application, such as Figure 6 As shown, including:

[0233] In the first step, the first dynamic data of the water injection well and the second dynamic data of the oil production well are input into the encoding network;

[0234] In the second step, the encoding network performs convolution operation on the first dynamic data of the water injection well and the second dynamic data of the oil production well to obtain initial features;

[0235] The third step is to input the initial features into the first convolutional layer of the feature extraction network. The first convolutional layer uses the first temporal convolutional network to process the initial features, obtain the first sub-feature output by the first temporal convolutional network, and use the sigmoid activation function to continue feature processing to obtain the processed first sub-feature.

[0236] At the same time, the first convolution layer uses the second temporal convolutional network to process the initial features, obtain the first sub-feature output by the second temporal convolutional network, and uses the tanh activation function to continue feature processing to obtain the processed first sub-feature;

[0237] Step 4: multiply the processed first sub-feature and the processed second sub-feature to obtain the first feature;

[0238] The fifth step is to input the first feature of the first convolutional layer into the adaptive graph convolutional network built into the feature extraction network to obtain the second feature output by the adaptive graph convolutional network;

[0239] Step 6: Input the second feature output by the adaptive graph convolutional network into the second convolutional layer as the input vector of the second convolutional layer until the first feature generated by each of the L convolutional layers is obtained;

[0240] Step 7: After the first features generated by each of the L convolutional layers have completed the convolution operation, the L first features are accumulated to obtain the fused features;

[0241] In the eighth step, the fused features are input into the decoding network, and the fused features are decoded by two decoding layers in the decoding network to obtain the predicted oil production.

[0242] Figure 7 The schematic diagram of the structure of the oil reservoir production prediction device based on the adaptive graph convolutional network provided in this application is as follows: Figure 7 As shown, the oil reservoir production prediction device 700 based on the adaptive graph convolutional network provided in this embodiment includes:

[0243] The acquisition module 701 is used to acquire first dynamic data of the water injection well and second dynamic data of the oil production well, wherein the first dynamic data includes the water injection rate and / or bottom hole flow pressure of the water injection well, and the second dynamic data includes at least one of the oil production rate, water production rate, liquid production rate, and bottom hole flow pressure of the oil production well;

[0244] Input module 702 is used to input the first dynamic data and the second dynamic data into a pre-trained prediction model to obtain the predicted oil production output by the prediction model, wherein the feature extraction network in the prediction model includes an adaptive graph convolutional network, and the adjacency matrix in the adaptive graph convolutional network is used to characterize the spatial influence relationship between the water injection well and the oil production well.

[0245] In a possible implementation, the input module 702 is further used to input the first dynamic data and the second dynamic data into the encoding network to obtain initial features, use the initial features as input vectors of the first convolutional layer in the multiple cascaded convolutional layers, and for any convolutional layer, generate a first feature by passing the input vector of the convolutional layer through the timing processing module of the convolutional layer, then generate a second feature by passing the first feature through the adaptive graph convolutional network of the convolutional layer, and use the second feature as the input vector of the next convolutional layer, until the first features generated by each of the multiple cascaded convolutional layers are obtained;

[0246] The input module 702 is specifically used to input the first features generated by each of the multiple cascaded convolutional layers into the decoding network to obtain the predicted oil production.

[0247] In a possible implementation, the input module 702 is further configured to input the input vector of the convolutional layer into the first temporal convolutional network to obtain a first sub-feature;

[0248] The input module 702 is further used to input the input vector of the convolution layer into the second temporal convolutional network to obtain a second sub-feature;

[0249] The device further comprises: a product module 703;

[0250] The product module 703 is used to multiply the first sub-feature and the second sub-feature to obtain the first feature;

[0251] The device further includes: a generating module 704;

[0252] The generating module 704 is used to generate the second feature based on the first feature and the adjacency matrix in the adaptive graph convolutional network.

[0253] In a possible implementation, the device further includes: an accumulation module 705;

[0254] The accumulation module 705 is used to accumulate the first features generated by each of the multiple cascaded convolutional layers to obtain a fused feature;

[0255] The device further comprises: a decoding module 706;

[0256] The decoding module 706 is used to decode the fused features through one or more decoding layers of the decoding network to obtain the predicted oil production.

[0257] The device 700 for predicting the production of an injection-production oil reservoir provided in this embodiment can execute the method provided in the above method embodiment, and its implementation principle and technical effect are similar, so this embodiment will not be described in detail here.

[0258] Figure 8 A schematic diagram of the structure of the training device for the oil production prediction model provided in this application, such as Figure 8 As shown, the training device 800 for the oil production prediction model provided in this embodiment includes:

[0259] The acquisition module 801 is used to acquire sample data, wherein the sample data includes first dynamic sample data of the water injection well, second dynamic sample data of the oil production well, and an oil production label, wherein the first dynamic sample data includes the water injection rate and / or bottom hole flow pressure of the water injection well, and the second dynamic sample data includes at least one of the oil production rate, water production rate, liquid production rate, and bottom hole flow pressure of the oil production well;

[0260] An input module 802 is used to input the first dynamic sample data and the second dynamic sample data into an oil production prediction model to obtain a predicted oil production output by the oil production prediction model, wherein the feature extraction network in the oil production prediction model includes an adaptive graph convolution network, and the initial adjacency matrix in the adaptive graph convolution network is used to characterize the spatial influence relationship between the water injection well and the oil production well;

[0261] The adjustment module 803 is used to adjust the parameters of the oil production prediction model and the initial adjacency matrix based on the predicted oil production and the oil production label.

[0262] In a possible implementation, the acquisition module 801 is further used to acquire first static sample data of the water injection well and second static sample data of the oil production well, wherein the first static sample data includes permeability, reservoir thickness and position coordinates of the reservoir at the location of the water injection well, and the second static sample data includes permeability, reservoir thickness and position coordinates of the reservoir at the location of the oil production well;

[0263] The apparatus further includes: a determination module 804;

[0264] The determining module 804 is used to determine the initial adjacency matrix according to the first static sample data and the second static sample data.

[0265] The oil production prediction model training device 800 provided in this embodiment can execute the method provided in the above method embodiment, and its implementation principle and technical effect are similar, which will not be described in detail in this embodiment.

[0266] Fig. 9 This is a schematic diagram of the structure of the oil reservoir production prediction device based on the adaptive graph convolutional network provided in this application. Fig. 9 As shown, the electronic device 900 provided in this embodiment includes: at least one processor 901 and a memory 902. Optionally, the device 900 also includes a communication component 903. The processor 901, the memory 902 and the communication component 903 are connected via a bus 904.

[0267] In a specific implementation process, at least one processor 901 executes the computer execution instructions stored in the memory 902, so that at least one processor 901 executes the above method.

[0268] The specific implementation process of the processor 901 can be found in the above method embodiment, and its implementation principle and technical effect are similar, so this embodiment will not be repeated here.

[0269] In the above embodiments, it should be understood that the processor can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), etc. A general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in the invention can be directly implemented as a hardware processor, or can be implemented by a combination of hardware and software modules in the processor.

[0270] The memory may include a high-speed memory (Random Access Memory, RAM), and may also include a non-volatile memory (NVM), such as at least one disk storage.

[0271] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, the bus in the drawings of this application is not limited to only one bus or one type of bus.

[0272] The present application also provides a computer program product, including a computer program, which implements the above method when executed by a processor.

[0273] The present application also provides a computer-readable storage medium, in which computer-executable instructions are stored. When a processor executes the computer-executable instructions, the above method is implemented.

[0274] The above-mentioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk. The readable storage medium can be any available medium that can be accessed by a general or special-purpose computer.

[0275] An exemplary readable storage medium is coupled to a processor so that the processor can read information from the readable storage medium and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can be located in an application specific integrated circuit (Application Specific Integrated Circuits, referred to as: ASIC). Of course, the processor and the readable storage medium can also exist in the device as discrete components.

[0276] The division of units is only a logical function division, and there may be other divisions in actual implementation, such as 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 an indirect coupling or communication connection through some interface, device or unit, which can be electrical, mechanical or other forms.

[0277] 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.

[0278] In addition, each functional unit in each embodiment of the present invention 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.

[0279] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0280] Those skilled in the art can understand that all or part of the steps of implementing the above-mentioned method embodiments can be completed by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, the steps of the above-mentioned method embodiments are executed; and the aforementioned storage medium includes: ROM, RAM, disk or optical disk and other media that can store program codes.

[0281] Finally, it should be noted that those skilled in the art will readily conceive of other embodiments of the present invention after considering the specification and practicing the invention disclosed herein. The present invention is intended to cover any variations, uses or adaptations of the present invention, which follow the general principles of the present invention and include common knowledge or customary technical means in the art not disclosed by the present invention, are not limited to the precise structure described above and shown in the drawings, and may be modified and changed in various ways without departing from the scope thereof. The scope of the present invention is limited only by the appended claims.

Claims

1. A method for predicting oil reservoir production based on an adaptive graph convolutional network, characterized in that: include: Acquire first dynamic data of the water injection well and second dynamic data of the oil production well, wherein the first dynamic data includes the water injection rate and / or bottom hole flow pressure of the water injection well, and the second dynamic data includes at least one of the oil production rate, water production rate, liquid production rate, and bottom hole flow pressure of the oil production well; The first dynamic data and the second dynamic data are input into a pre-trained prediction model to obtain the predicted oil production output by the prediction model, wherein the feature extraction network in the prediction model includes an adaptive graph convolutional network, and the adjacency matrix in the adaptive graph convolutional network is used to characterize the spatial influence relationship between the water injection well and the oil production well.

2. The method according to claim 1, characterized in that The prediction model includes an encoding network, a feature extraction network and a decoding network, wherein the feature extraction network includes a plurality of cascaded convolutional layers, and each of the convolutional layers includes a timing processing module and the adaptive graph convolutional network; The step of inputting the first dynamic data and the second dynamic data into a pre-trained prediction model to obtain a predicted oil production output by the prediction model comprises: Inputting the first dynamic data and the second dynamic data into the encoding network to obtain initial features, using the initial features as input vectors of the first convolutional layer among the multiple cascaded convolutional layers, and for any convolutional layer, passing the input vector of the convolutional layer through the timing processing module of the convolutional layer to generate a first feature, then passing the first feature through the adaptive graph convolutional network of the convolutional layer to generate a second feature, and using the second feature as the input vector of the next convolutional layer, until the first features generated by each of the multiple cascaded convolutional layers are obtained; The first features generated by each of the multiple cascaded convolutional layers are input into the decoding network to obtain the predicted oil production.

3. The method according to claim 2, characterized in that The timing processing module includes a first timing convolutional network and a second timing convolutional network, wherein the input vector of the convolutional layer is passed through the timing processing module of the convolutional layer to generate a first feature, and then the first feature is passed through the adaptive graph convolutional network of the convolutional layer to generate a second feature, including: Inputting the input vector of the convolutional layer into the first temporal convolutional network to obtain a first sub-feature; Inputting the input vector of the convolutional layer into the second temporal convolutional network to obtain a second sub-feature; multiplying the first sub-feature and the second sub-feature to obtain the first feature; The second feature is generated based on the first feature and an adjacency matrix in the adaptive graph convolutional network.

4. The method according to claim 2, characterized in that: The step of inputting the first features generated by each of the plurality of cascaded convolutional layers into the decoding network to obtain the predicted oil production comprises: Accumulating the first features generated by each of the multiple cascaded convolutional layers to obtain a fused feature; The fused features are decoded by one or more decoding layers of the decoding network to obtain the predicted oil production.

5. The method according to any one of claims 1 to 4, characterized in that: The adjacency matrix is ​​obtained by training the prediction model based on the initial adjacency matrix, and the initial adjacency matrix is ​​determined based on the first static data of the water injection well and the second static data of the oil production well. The first static data includes the permeability, reservoir thickness and location coordinates of the reservoir where the water injection well is located, and the second static data includes the permeability, reservoir thickness and location coordinates of the reservoir where the oil production well is located.

6. A training method for an oil production prediction model, characterized in that: include: Acquire sample data, wherein the sample data includes first dynamic sample data of a water injection well, second dynamic sample data of an oil production well, and an oil production label, wherein the first dynamic sample data includes the water injection rate and / or bottom hole flow pressure of the water injection well, and the second dynamic sample data includes at least one of the oil production rate, water production rate, liquid production rate, and bottom hole flow pressure of the oil production well; Inputting the first dynamic sample data and the second dynamic sample data into an oil production prediction model to obtain a predicted oil production output by the oil production prediction model, wherein the feature extraction network in the oil production prediction model includes an adaptive graph convolution network, and the initial adjacency matrix in the adaptive graph convolution network is used to characterize the spatial influence relationship between the water injection well and the oil production well; The parameters of the oil production prediction model and the initial adjacency matrix are adjusted based on the predicted oil production and the oil production labels.

7. The method according to claim 6, characterized in that Also includes: Acquire the first static sample data of the water injection well and the second static sample data of the oil production well, wherein the first static sample data includes the permeability, reservoir thickness and position coordinates of the reservoir at the location of the water injection well, and the second static sample data includes the permeability, reservoir thickness and position coordinates of the reservoir at the location of the oil production well; The initial adjacency matrix is ​​determined according to the first static sample data and the second static sample data.

8. An oil reservoir production prediction device based on an adaptive graph convolutional network, characterized in that: include: Memory, processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory, so that the processor performs the method according to any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 7 when executed by a processor.

10. A computer program product, characterized in that The invention comprises a computer program, which implements the method according to any one of claims 1 to 7 when being executed by a processor.

Images