A method, apparatus, medium, and equipment for predicting shale oil production capacity.
A shale oil production capacity prediction model was constructed using deep learning methods. By utilizing the mapping relationship between seismic attribute data and production capacity data, the model hyperparameters were optimized, solving the problem of inaccuracy in shale oil production capacity prediction and achieving high-precision production capacity prediction.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2025-05-30
- Publication Date
- 2026-04-03
AI Technical Summary
In existing technologies, shale oil production capacity prediction methods suffer from low accuracy. The prediction results of multi-parameter planar overlay method and geophysical response prediction method are inaccurate, and well logging modeling evaluation method cannot fully include shale oil sweet spot evaluation parameters.
By employing deep learning methods, a mapping relationship is constructed by acquiring shale oil production capacity data and seismic attribute data from geological layers. A shale oil production capacity prediction model is trained, and the model hyperparameters are optimized using a fully connected neural network and a batch normalization layer, combined with grid search and tri-fold cross-validation, to achieve quantitative prediction of shale oil production capacity.
It improves the accuracy of shale oil production capacity prediction, solves the problems of high uncertainty and subjectivity in traditional methods, and realizes intelligent quantitative prediction from seismic characteristics to production capacity.
Smart Images
Figure CN120254967B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of deep learning technology, and in particular to a method, apparatus, medium and equipment for predicting shale oil production capacity. Background Technology
[0002] Shale oil sweet spots refer to shale oil-rich areas with superior source rocks, reservoir properties, oil-bearing characteristics, brittleness, and geostress characteristics, coupled with relatively low engineering modification costs, and good development benefits under current economic and technological conditions. Establishing a shale oil sweet spot evaluation system is crucial for achieving optimal shale oil selection and efficient development. Shale oil sweet spot evaluation typically includes geological sweet spots, engineering sweet spots, and economic sweet spots (Sun et al., 2021). Geological sweet spots mainly focus on the comprehensive evaluation of geological characteristics such as source rocks, reservoirs, natural fractures, formation energy, and local structures (Qian et al., 2018); engineering sweet spots mainly include a comprehensive evaluation of rock fracturability and geostress anisotropy (Hou et al., 2021; Lu et al., 2022); and economic sweet spots typically encompass a comprehensive evaluation of economic influencing factors such as resource abundance, surface conditions, and development difficulty (Wu et al., 2023; Zeng et al., 2024). The parameters selected for shale oil sweet spots vary in different regions, and the numerical ranges of the evaluation indicators are also different, resulting in different predictions of shale oil production.
[0003] Existing technologies for predicting shale oil production include multi-parameter planar overlay methods, geophysical response prediction methods, and well logging modeling evaluation methods.
[0004] Among the aforementioned prediction methods, the multi-parameter planar overlay method can quickly select favorable areas. However, considering the large thickness and strong heterogeneity of continental shale layers, the prediction accuracy after multi-parameter planar overlay is reduced. The geophysical response prediction method is limited by the seismic inversion scale and the currently available geophysical response characteristics, resulting in inaccurate prediction results. The well logging modeling evaluation method is limited by the numerous types of shale oil sweet spot evaluation parameters, which cannot fully encompass all shale oil sweet spot evaluation parameters, leading to inaccurate prediction results. In summary, existing technologies suffer from inaccurate shale oil production capacity predictions due to various factors. Summary of the Invention
[0005] Therefore, it is necessary to provide a method, apparatus, medium, and equipment for predicting shale oil production capacity to address the aforementioned technical problems.
[0006] The following technical solution is adopted in this specification:
[0007] This specification provides a method for predicting shale oil production capacity, including:
[0008] Obtain shale oil production data from horizontal wells in shale oil-bearing geological formations;
[0009] Seismic attributes within a specified time window are extracted from geological strata containing shale oil to obtain initial seismic attribute data;
[0010] Using the geodetic coordinates of the target horizontal well as an index, a mapping relationship between the initial seismic attribute data and shale oil production capacity is constructed based on the initial seismic attribute data and shale oil production capacity data;
[0011] A shale oil production capacity prediction model is constructed, and the model is trained using mapping relationships to obtain a well-trained shale oil production capacity prediction model.
[0012] The seismic attribute data of the entire target area is input into the trained shale oil production capacity prediction model to obtain the shale oil production capacity data of the target area.
[0013] Optionally, the model architecture of the shale oil production capacity prediction model includes an input layer, a hidden layer, a BN layer, and an output layer;
[0014] The number of hidden layers is at least two, and every two adjacent hidden layers are connected by a BN layer.
[0015] The shale oil production prediction model uses a grid search method to determine the number of hidden layers and the number of neurons in the hidden layers.
[0016] The shale oil production capacity prediction model uses a three-fold cross-validation method to determine the target hyperparameter combination from the preset hyperparameter combinations.
[0017] Optionally, before constructing the mapping relationship between the initial seismic attribute data and shale oil production capacity based on the initial seismic attribute data and shale oil production capacity, using the geodetic coordinates of the target horizontal well as an index, the method further includes:
[0018] The initial seismic attribute data is preprocessed to obtain preprocessed seismic attribute data.
[0019] Based on the correlation between each seismic attribute data and shale oil production capacity in the preprocessed seismic attribute data, the preprocessed seismic attribute data is filtered to obtain seismic attribute data.
[0020] Based on the initial seismic attribute data and shale oil production capacity data, a mapping relationship between the initial seismic attribute data and shale oil production capacity is constructed, specifically including:
[0021] An initial mapping relationship between seismic attribute data and shale oil production capacity was constructed based on seismic attribute data and shale oil production capacity data.
[0022] Optionally, based on the correlation between each seismic attribute data point and shale oil production capacity in the preprocessed seismic attribute data, the preprocessed seismic attribute data is filtered to obtain seismic attribute data, specifically including:
[0023] Hierarchical clustering was used to calculate the correlation between each seismic attribute data and shale oil production capacity in the preprocessed seismic attribute data, and the similarity calculation results were obtained.
[0024] Based on the similarity calculation results, the seismic attribute data is determined;
[0025] Seismic attribute data includes the average negative polarity amplitude, average amplitude, average positive polarity amplitude, average energy, asymmetry, root mean square amplitude, kurtosis, and maximum amplitude.
[0026] Optionally, obtain shale oil production data from horizontal wells in shale oil-bearing geological formations, including:
[0027] The shale oil production of each single well in each fracturing section, the contribution rate of the fracturing section, and the concentration of oil phase tracer in the fracturing section were obtained.
[0028] Based on the shale oil production of each single well in each fracturing section, the contribution rate of the fracturing section, and the concentration of oil phase tracer in the fracturing section, the shale oil production capacity data is calculated using a preset formula for calculating the production capacity of the fracturing section.
[0029] Optionally, the preset formula for calculating the production capacity of the shale oil fracturing section is:
[0030] ;
[0031] in, Indicates the first Heavenly Shale oil production per fractured section Indicates the number of wells in the first... Shale oil production at opportune times Indicates the first Heavenly Contribution rate at each fracturing segment Indicates the first Heavenly Oil phase tracer concentration at each fracturing section No. The total concentration of oil phase tracer in all fracturing sections within a day.
[0032] Optionally, the method further includes:
[0033] Calculate the cumulative oil production based on shale oil production capacity data;
[0034] Calculate the cumulative probability of oil production based on the cumulative oil production.
[0035] Based on the cumulative oil production and cumulative oil production probability, a relationship diagram is drawn, and the oil-producing layers in the target area are classified according to the Lorenz cumulative probability curve relationship diagram to determine the type of oil-producing layer.
[0036] This specification provides a shale oil production capacity prediction device, including:
[0037] The data acquisition module is specifically used to acquire shale oil production data of horizontal wells in geological strata containing shale oil; extract seismic attributes within a specified time window in geological strata containing shale oil to obtain initial seismic attribute data; and construct a mapping relationship between initial seismic attribute data and shale oil production capacity based on the geodetic coordinates of the target horizontal well and the initial seismic attribute data and shale oil production capacity data.
[0038] The model training module is specifically used to construct a shale oil production capacity prediction model and train the shale oil production capacity prediction model using mapping relationships to obtain a trained shale oil production capacity prediction model.
[0039] The production capacity calculation module is specifically used to input the seismic attribute data of the entire target area into the trained shale oil production capacity prediction model to obtain the shale oil production capacity data of the target area.
[0040] This specification provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described shale oil production capacity prediction method.
[0041] This specification provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the above-described shale oil production capacity prediction method.
[0042] The above-mentioned technical solutions adopted in this specification can achieve the following beneficial effects:
[0043] The shale oil production capacity prediction method provided in this specification first extracts seismic attributes within a specified time window from the geological strata containing shale oil as initial data, and calculates the shale oil production capacity of horizontal wells using a preset formula; then, based on the geographical coordinates of the horizontal wells, a mapping relationship between seismic attributes and production capacity is established, and a machine learning model is trained based on this; finally, using the trained model, the distribution of shale oil production capacity is predicted based on the seismic attribute data of the entire target area, realizing quantitative intelligent prediction from seismic characteristics to production capacity.
[0044] This invention uses a data-driven approach to establish the relationship between shale oil production capacity and seismic attributes, realizing the transformation from the traditional method of qualitatively identifying shale oil fields based on geological features to a deep learning-based nonlinear big data method for identifying shale oil fields. It fully explores the implicit relationships between seismic attributes, solves the problems of high uncertainty and high subjectivity in traditional methods, and improves the accuracy of shale oil production capacity prediction. Attached Figure Description
[0045] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:
[0046] Figure 1 This document provides a schematic flowchart of a shale oil production capacity prediction method.
[0047] Figure 2 A schematic diagram illustrating the oil production contribution rate of the fracturing section;
[0048] Figure 3 This is a schematic diagram showing the average daily oil production distribution in the fracturing section.
[0049] Figure 4 A schematic diagram illustrating the link between shale oil production capacity and seismic attributes;
[0050] Figure 5 This is a model architecture diagram for a shale oil production capacity prediction model.
[0051] Figure 6 This is a schematic diagram illustrating the model accuracy of the shale oil production capacity prediction model training results.
[0052] Figure 7 A schematic diagram showing the distribution of shale oil production capacity in the target area;
[0053] Figure 8 A graph showing the relationship between daily oil production and cumulative probability of daily oil production;
[0054] Figure 9 A schematic diagram of the advantageous areas after classifying the target area;
[0055] Figure 10 This is a schematic diagram of a shale oil production capacity prediction device provided in this specification;
[0056] Figure 11 This is a schematic diagram of a computer device for implementing a shale oil production capacity prediction method, as provided in this specification. Detailed Implementation
[0057] To make the objectives, technical solutions, and advantages of this specification clearer, the technical solutions of this application will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments in this specification without creative effort are within the scope of protection of this application.
[0058] The technical solutions provided by the various embodiments of this application are described in detail below with reference to the accompanying drawings.
[0059] Figure 1 This is a flowchart illustrating a shale oil production capacity prediction method described in this specification, which specifically includes the following steps:
[0060] S101: Obtain shale oil production data for horizontal wells in geological formations containing shale oil.
[0061] For example, this embodiment focuses on the prediction of shale oil production capacity. Furthermore, since the G9 layer is mainly composed of organic-rich shale with thin layers of siltstone or calcareous layers, exhibiting typical shale oil reservoir characteristics, the geological strata containing shale oil in this embodiment can refer to the G9 layer.
[0062] Based on this, in one or more embodiments of this specification, the executing entity can be a hardware device or system with multimodal data acquisition, processing and analysis capabilities, including servers, edge computing devices, etc.
[0063] The server mentioned in this manual can be a server set up on a business platform, or a device such as a desktop computer or laptop computer capable of executing the solution described in this manual. For ease of explanation, the following description will only focus on the server as the execution subject.
[0064] In this embodiment, obtaining shale oil production data from horizontal wells in shale oil-bearing geological formations includes:
[0065] The shale oil production of each single well in each fracturing section, the contribution rate of the fracturing section, and the concentration of oil phase tracer in the fracturing section were obtained.
[0066] Based on the shale oil production of each single well in each fracturing section, the contribution rate of the fracturing section, and the concentration of oil phase tracer in the fracturing section, the shale oil production capacity data is calculated using a preset formula for calculating the production capacity of the fracturing section.
[0067] The pre-defined formula for calculating the production capacity of the shale oil fracturing section is:
[0068] ;
[0069] in, Indicates the first Heavenly Shale oil production per fractured section Indicates the number of wells in the first... Shale oil production at opportune times Indicates the first Heavenly Contribution rate at each fracturing segment Indicates the first Heavenly Oil phase tracer concentration at each fracturing section No. The total concentration of oil phase tracer in all fracturing sections within a day.
[0070] For example, taking QH2202Y-G9 as an example, as shown in Figure 2, Figure 2 This diagram illustrates the contribution rate of oil production to the fracturing stage. (Example:) Figure 3 As shown, Figure 3 This is a schematic diagram showing the average daily oil production distribution in the fracturing section.
[0071] S102: Extract seismic attributes within a specified time window from geological strata containing shale oil to obtain initial seismic attribute data.
[0072] In seismic data interpretation, a designated time window refers to a fixed time range selected around a geological stratum containing shale oil (such as the top and bottom interface of layer G9) to extract seismic attribute values within that range. The length of the time window can be determined based on the target layer thickness and seismic resolution, and the time window position can be a fixed time window or a sliding time window.
[0073] Optionally, before step S102, a seismic calibration method based on well logging data can be used. Based on shale oil seismic data and a small amount of stratigraphic interpretation and annotation data, seismic stratigraphic positions are interpreted, and the geological stratigraphic position containing shale oil is determined to be layer G9. Seismic attributes are then extracted along layer G9 within a specified time window. In this embodiment, 36 seismic attributes, such as amplitude, frequency, phase, and energy, are extracted for lithological interpretation and reservoir property prediction, as shown in Table 1. Table 1 provides examples of some seismic attributes. Among them, Ant average envelope represents the Ant average envelope value, Average energy represents the average energy of 3.9912e+05 and 1.3089e+06 respectively, Average magnitude represents the average amplitude of 732.9124 (first group) and 1355.3828 (second group), Average negative amplitude represents the average negative amplitude of 0, indicating that no negative fluctuations were detected, and Median represents the median of 557.5268 and 731.1199 respectively, which differ from the average value, indicating that the data distribution is asymmetrical.
[0074] Table 1 Examples of some earthquake attributes
[0075]
[0076] S103: Using the geodetic coordinates of the target horizontal well as an index, construct a mapping relationship between the initial seismic attribute data and shale oil production capacity based on the initial seismic attribute data and shale oil production capacity data.
[0077] For example, as shown in Table 2, Table 2 is an example table of the mapping relationship between shale oil production capacity and seismic attributes.
[0078] Table 2. Example of Shale Oil Production Capacity-Seismic Attribute Mapping Relationship
[0079]
[0080] In Table 2, (X, Y) represent the geodetic coordinates of the target horizontal well, Oil Production (t / d) represents oil production (tons / day), Ant average envelope represents the Ant average envelope value, Average energy represents the average energy, Average magnitude represents the average amplitude, Average negative amplitude represents the average negative amplitude, Average positive amplitude represents the average positive amplitude, and Median represents the median. Furthermore, as... Figure 4 As shown, Figure 4 This is a schematic diagram illustrating the link between shale oil production capacity and seismic attributes.
[0081] In this embodiment, before constructing the mapping relationship between the initial seismic attribute data and shale oil production capacity based on the initial seismic attribute data and shale oil production capacity using the geodetic coordinates of the target horizontal well as an index, the method further includes: performing data preprocessing on the initial seismic attribute data to obtain preprocessed seismic attribute data; and filtering the preprocessed seismic attribute data based on the correlation between each seismic attribute data and shale oil production capacity in the preprocessed seismic attribute data to obtain seismic attribute data.
[0082] The mapping relationship between initial seismic attribute data and shale oil production capacity is constructed based on initial seismic attribute data and shale oil production capacity data. Specifically, this includes: constructing the mapping relationship between initial seismic attribute data and shale oil production capacity based on initial seismic attribute data and shale oil production capacity data.
[0083] Optionally, the steps of preprocessing the initial seismic attribute data and filtering the preprocessed seismic attribute data can also be performed after S103, without specific restrictions.
[0084] Taking the steps of preprocessing the initial seismic attribute data and filtering the preprocessed seismic attribute data as an example, these steps can also be performed after S103. Because seismic attributes vary greatly in magnitude, directly inputting the raw data into the neural network for training may cause repeated oscillations during gradient updates, affecting network convergence and even leading to numerical out-of-bounds errors. Therefore, zero-mean normalization is used to standardize the data (i.e., data preprocessing), mapping the raw data to a distribution with a mean of 0 and a standard deviation of 1. The calculation formula is as follows:
[0085] ;
[0086] in, This represents the initial seismic attribute data. and These represent the mean and standard deviation of the initial seismic attribute data, respectively. This represents the preprocessed seismic attribute data.
[0087] For example, as shown in Table 3, Table 3 is an example table of standardized shale oil production capacity-seismic attribute data. Where (X, Y) represents the geodetic coordinates of the target horizontal well, Oil Production (t / d) represents oil production (tons / day), Antaverage envelope represents the Ant average envelope value, Average energy represents the average energy, average envelope represents the average envelope value, Average magnitude represents the average amplitude, Average negative amplitude represents the average negative amplitude, and Median represents the median.
[0088] Table 3. Example of standardized shale oil production capacity-seismic attribute data
[0089]
[0090] In this embodiment, the preprocessed seismic attribute data is filtered based on the correlation between each seismic attribute data and shale oil production capacity to obtain seismic attribute data. Specifically, this includes: using hierarchical clustering to calculate the correlation between each seismic attribute data and shale oil production capacity in the preprocessed seismic attribute data, and obtaining similarity calculation results; determining the seismic attribute data based on the similarity calculation results; the seismic attribute data includes the average negative polarity amplitude, average amplitude, average positive polarity amplitude, average energy, asymmetry, root mean square amplitude, kurtosis, and maximum amplitude.
[0091] S104: Construct a shale oil production capacity prediction model and train the shale oil production capacity prediction model using mapping relationships to obtain a trained shale oil production capacity prediction model.
[0092] In this embodiment, the model architecture of the shale oil production capacity prediction model includes an input layer, a hidden layer, a BN layer, and an output layer; wherein, the number of hidden layers is at least two, and every two adjacent hidden layers are connected by a BN layer; the shale oil production capacity prediction model determines the number of hidden layers and the number of neurons in the hidden layers through a grid search method; the shale oil production capacity prediction model determines the target hyperparameter combination from the preset hyperparameter combinations through a three-fold cross-validation method.
[0093] For example, a shale oil production prediction model can be obtained by adding a batch normalization (BN) layer to a fully connected neural network. A fully connected neural network is a classic feedforward neural network structure, typically composed of multiple stacked neuron layers. Each neuron in each layer is connected to all neurons in the preceding and following layers. Data propagates unidirectionally from the input layer through a series of hidden layers to the output layer, without feedback connections. Each neuron can be viewed as a computational unit; it receives input from all neurons in the preceding layer, performs a weighted summation, adds a bias term, and outputs the final result through an activation function.
[0094] Optionally, to avoid overfitting, the Dropout layer can be used to randomly deactivate some neurons' outputs, reducing the correlation between neurons, decreasing the model's dependence on certain specific neurons, and improving the model's generalization ability.
[0095] During the training of deep neural networks, due to the high correlation and coupling between layers, parameter updates can lead to changes in the distribution of input data for subsequent layers. This phenomenon is called "Internal Covariate Shift," increasing the difficulty of training and slowing down the model's convergence speed. To address this, batch normalization (BN) layers are introduced into fully connected neural networks. The basic idea of BN is to insert a normalization layer at the input of each layer. By standardizing the output of the activation layers, the distribution of the input data for each layer is transformed into a standard normal distribution with a mean of 0 and a standard deviation of 1. The input training data passes through the fully connected layers and then enters the BN layer for batch normalization, restoring the data distribution to a standard normal distribution before being fed into the next layer for training. This accelerates the model's convergence speed and helps prevent problems such as vanishing and exploding gradients. Figure 5 As shown, Figure 5 This is a model architecture diagram for a shale oil production capacity prediction model.
[0096] In this embodiment, a grid search method is used to optimize the hyperparameters of the fully connected neural network (such as the number of intermediate hidden layers and the number of neurons in the hidden layers). Three-fold cross-validation is then used to evaluate the preset parameter combinations, thereby selecting the hyperparameter combination with the best performance (i.e., the target hyperparameter combination) to build the network model. Table 4 shows examples of preset parameter combinations. Here, layers represents the number of network layers, neurons represents the number of neurons, batch represents the batch size, epochs represents the number of training epochs, and dropout represents the dropout rate.
[0097] Table 4 Preset Parameter Combinations
[0098]
[0099] Optionally, four evaluation metrics can be used to assess the model's performance: Mean Squared Error (MSE), Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), and Coefficient of Determination (R²). The optimal model is then selected to determine the shale oil production capacity prediction model. Through grid search, the optimal grid parameters and the training results of the shale oil production capacity prediction model are obtained, as shown in Table 5. Table 5 presents the optimal grid parameters and the training results of the shale oil production capacity prediction model. Furthermore, as... Figure 6 As shown, Figure 6 This diagram illustrates the model accuracy of the shale oil production prediction model training results. In the diagram, layers represents the number of network layers, neurons represents the number of neurons, batch represents the batch size, epochs represents the number of training epochs, and dropout represents the dropout rate.
[0100] Table 5. Training Results of Optimal Grid Parameters and Shale Oil Production Prediction Model
[0101]
[0102] S105: Input the seismic attribute data of the entire target area into the trained shale oil production capacity prediction model to obtain the shale oil production capacity data of the target area.
[0103] In this embodiment, the method further includes:
[0104] Calculate the cumulative oil production based on shale oil production capacity data;
[0105] Calculate the cumulative probability of oil production based on the cumulative oil production.
[0106] Based on the cumulative oil production and cumulative oil production probability, a relationship diagram is drawn, and the oil-producing layers in the target area are classified according to the Lorenz cumulative probability curve relationship diagram to determine the type of oil-producing layer.
[0107] For example, such as Figure 7 As shown, Figure 7 This is a schematic diagram showing the distribution of shale oil production capacity in the target area.
[0108] Optionally, based on shale oil production capacity data, the formula for calculating cumulative oil production is as follows:
[0109] ;
[0110] in, This represents the cumulative oil production corresponding to each data point. This represents the daily oil production corresponding to each shale oil production capacity data point.
[0111] The formula used to calculate the cumulative probability of oil production based on cumulative oil production is as follows:
[0112] ;
[0113] in, This represents the cumulative probability of oil production corresponding to each shale oil production capacity data point. This indicates the number of samples for shale oil production capacity data.
[0114] like Figure 8 As shown, Figure 8 This is a graph showing the relationship between daily oil production and its cumulative probability. (Example:) Figure 9 As shown, Figure 9 A schematic diagram of the favorable areas after classifying the target area.
[0115] Optionally, based on the Lorenz cumulative probability curve relationship diagram, the oil-producing reservoirs can be classified into three types if the daily oil production is <0.17t / d; Class II reservoirs are those with a daily oil production of 0.17t / d < 0.61t / d; and Class I reservoirs are those with a daily oil production of >0.61t / d.
[0116] based on Figure 1 The shale oil production capacity prediction method shown adopts a data-driven approach to establish the relationship between the production capacity of favorable areas and seismic attributes. This transforms the traditional method of qualitatively identifying favorable areas based on geological features into a deep learning-based nonlinear big data method for identifying favorable areas. It fully explores the implicit relationships between seismic attributes, solves the problems of high uncertainty and high subjectivity in traditional methods, and improves the accuracy of shale oil production capacity prediction.
[0117] When applying the shale oil production capacity prediction method provided in this manual, it is not necessary to rely on... Figure 1 The steps shown are executed in sequence. The specific execution order of each step can be determined as needed, and this manual does not impose any restrictions on it.
[0118] The above are one or more embodiments of the shale oil production capacity prediction method provided in this specification. Based on the same idea, this specification also provides a corresponding shale oil production capacity prediction device, such as... Figure 10 As shown.
[0119] Figure 10 A schematic diagram of a shale oil production capacity prediction device provided in this specification includes:
[0120] The data acquisition module is specifically used to acquire shale oil production data of horizontal wells in geological strata containing shale oil; extract seismic attributes within a specified time window in geological strata containing shale oil to obtain initial seismic attribute data; and construct a mapping relationship between initial seismic attribute data and shale oil production capacity based on the geodetic coordinates of the target horizontal well and the initial seismic attribute data and shale oil production capacity data.
[0121] The model training module is specifically used to construct a shale oil production capacity prediction model and train the shale oil production capacity prediction model using mapping relationships to obtain a trained shale oil production capacity prediction model.
[0122] The production capacity calculation module is specifically used to input the seismic attribute data of the entire target area into the trained shale oil production capacity prediction model to obtain the shale oil production capacity data of the target area.
[0123] Specific limitations regarding the shale oil production capacity prediction device can be found in the limitations of the shale oil production capacity prediction method described above, and will not be repeated here. Each module in the aforementioned shale oil production capacity prediction device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each module.
[0124] This specification also provides a computer-readable storage medium storing a computer program that can be used to execute the above-described... Figure 1 The provided method for predicting shale oil production capacity.
[0125] This instruction manual also provides Figure 11 The schematic diagram of the computer device shown is as follows: Figure 11 At the hardware level, this computer device includes a processor, internal bus, network interface, memory, and non-volatile storage, and may also include other hardware required for business operations. The processor reads the corresponding computer program from the non-volatile storage into memory and then executes it to achieve the above. Figure 1 The provided method for predicting shale oil production capacity.
[0126] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, or optical storage, etc. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.
[0127] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
Claims
1. A method for predicting shale oil production capacity, characterized in that, include: Obtain shale oil production data from horizontal wells in shale oil-bearing geological formations; Seismic attributes within a specified time window are extracted from geological strata containing shale oil to obtain initial seismic attribute data; Using the geodetic coordinates of the target horizontal well as an index, a mapping relationship between the initial seismic attribute data and the shale oil production capacity is constructed based on the initial seismic attribute data and the shale oil production capacity. Before constructing the mapping relationship between the initial seismic attribute data and the shale oil production capacity based on the initial seismic attribute data and the shale oil production capacity, using the geodetic coordinates of the target horizontal well as an index, the method further includes: The initial seismic attribute data is preprocessed to obtain preprocessed seismic attribute data; Based on the correlation between each seismic attribute data in the preprocessed seismic attribute data and the shale oil production capacity, the preprocessed seismic attribute data is filtered to obtain seismic attribute data; The step of filtering the preprocessed seismic attribute data based on the correlation between each seismic attribute data and the shale oil production capacity to obtain seismic attribute data specifically includes: Hierarchical clustering is used to calculate the correlation between each seismic attribute data in the preprocessed seismic attribute data and the shale oil production capacity, and the similarity calculation results are obtained. Based on the similarity calculation results, the seismic attribute data is determined; The seismic attribute data includes the average negative polarity amplitude, average amplitude, average positive polarity amplitude, average energy, asymmetry, root mean square amplitude, kurtosis, and maximum amplitude. The step of constructing a mapping relationship between the initial seismic attribute data and the shale oil production capacity based on the initial seismic attribute data and the shale oil production capacity specifically includes: Based on the seismic attribute data and the shale oil production capacity data, a mapping relationship between the initial seismic attribute data and the shale oil production capacity is constructed; A shale oil production capacity prediction model is constructed, and the mapping relationship is used to train the shale oil production capacity prediction model to obtain a trained shale oil production capacity prediction model. The model architecture of the shale oil production capacity prediction model includes an input layer, a hidden layer, a BN layer, and an output layer. The number of hidden layers is at least two, and every two adjacent hidden layers are connected by a BN layer. The shale oil production capacity prediction model determines the number of hidden layers and the number of neurons in the hidden layers using a grid search method. The shale oil production capacity prediction model determines the target hyperparameter combination from the preset hyperparameter combinations using the three-fold cross-validation method; Input the seismic attribute data of the entire target area into the trained shale oil production capacity prediction model to obtain the shale oil production capacity data of the target area; Also includes: Based on the shale oil production capacity data, calculate the cumulative oil production. Calculate the cumulative probability of oil production based on the cumulative oil production. Based on the cumulative oil production and the cumulative probability of oil production, a relationship diagram is drawn, and the oil-producing layers in the target area are classified according to the Lorenz cumulative probability curve relationship diagram to determine the type of oil-producing layer.
2. The shale oil production capacity prediction method as described in claim 1, characterized in that, The acquisition of shale oil production data from horizontal wells in shale oil-bearing geological formations includes: The shale oil production of each single well in each fracturing section, the contribution rate of the fracturing section, and the concentration of oil phase tracer in the fracturing section were obtained. Based on the shale oil production of each single well in each fracturing section, the contribution rate of the fracturing section, and the oil phase tracer concentration of the fracturing section, the shale oil production capacity data is calculated using a preset shale oil fracturing section production capacity calculation formula.
3. The shale oil production capacity prediction method as described in claim 2, characterized in that, The formula for calculating the production capacity of the pre-set shale oil fracturing section is as follows: in, Indicates the first i Heavenly j Shale oil production per fractured section Indicates the number of wells in the first... i Shale oil production at opportune times Indicates the first i Heavenly j Contribution rate at each fracturing segment Indicates the first i Heavenly j Oil phase tracer concentration at each fracturing section Indicates the first i The total concentration of oil phase tracer in all fracturing sections within a day.
4. A shale oil production capacity prediction device based on the shale oil production capacity prediction method according to any one of claims 1-3, characterized in that, include: The data acquisition module is specifically used to acquire shale oil production capacity data of horizontal wells in geological strata containing shale oil; extract seismic attributes within a specified time window in the geological strata containing shale oil to obtain initial seismic attribute data; and construct a mapping relationship between the initial seismic attribute data and the shale oil production capacity based on the geodetic coordinates of the target horizontal well and the initial seismic attribute data and the shale oil production capacity data, using the geodetic coordinates of the target horizontal well as an index. The model training module is specifically used to construct a shale oil production capacity prediction model and train the shale oil production capacity prediction model using the mapping relationship to obtain a trained shale oil production capacity prediction model. The production capacity calculation module is specifically used to input the seismic attribute data of the entire target area into the trained shale oil production capacity prediction model to obtain the shale oil production capacity data of the target area.
5. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the method described in any one of claims 1-3.
6. A computer device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method described in any one of claims 1-3.