Reservoir permeability prediction model generation and prediction method, device, equipment and medium

Abstract

This application provides a reservoir permeability prediction model generation and prediction method, apparatus, equipment, and medium, relating to the field of petroleum exploration technology. The model generation method includes: generating multiple target data; determining multiple reservoir permeability intervals and their corresponding target datasets based on the target data; constructing membership functions corresponding to the reservoir permeability intervals based on fuzzy logic and determining the parameter values ​​in the membership functions; determining the target sample set corresponding to each reservoir permeability interval according to the membership degree corresponding to each target data under different membership functions; dividing the target sample set corresponding to the reservoir permeability interval into a first sample set and a second sample set, and training the initial reservoir permeability prediction model separately to obtain two target reservoir permeability prediction models corresponding to the reservoir permeability interval. This application can effectively solve the problem of low prediction accuracy in predicting reservoir permeability based on the correlation between porosity and reservoir permeability.

Description

Technical Field

[0001] This application relates to the field of petroleum exploration technology, specifically to a method and apparatus for generating reservoir permeability prediction models, a reservoir permeability prediction method and apparatus, electronic equipment, and a machine-readable storage medium. Background Technology

[0002] Reservoir permeability is an important physical property parameter reflecting the fluid flow capacity in rocks. It is an indispensable basic data in reservoir evaluation, production capacity prediction, oilfield development scheme design, and reservoir numerical simulation.

[0003] Typically, in conventional reservoirs with good porosity and permeability, there is a certain correlation between porosity and reservoir permeability. Through simple linear regression analysis, a relatively reliable reservoir permeability prediction model can be established, which can effectively reflect the impact of porosity changes on reservoir permeability.

[0004] However, for unconventional reservoirs, due to their strong heterogeneity and other factors, the traditional porosity-reservoir permeability linear regression method is difficult to accurately predict their reservoir permeability. Summary of the Invention

[0005] The purpose of this application is to provide a method and apparatus for generating a reservoir permeability prediction model, a reservoir permeability prediction method and apparatus, an electronic device, and a machine-readable storage medium, so as to solve the problem of low prediction accuracy in the prior art when predicting reservoir permeability based on the correlation between porosity and reservoir permeability.

[0006] To achieve the above objectives, the first aspect of this application provides a method for generating a reservoir permeability prediction model, the method comprising: Multiple target data are generated; among them, the target data includes multiple target logging curves and corresponding reservoir permeability curves, and each target logging curve corresponds to a target type. Based on the target data, multiple reservoir permeability intervals are determined, and the target dataset corresponding to each reservoir permeability interval is determined; wherein, the reservoir permeability curves contained in all target data in the target dataset belong to the same reservoir permeability interval. For each reservoir permeability interval, a membership function corresponding to the reservoir permeability interval is constructed based on fuzzy logic, and the parameter values ​​in the membership function corresponding to the reservoir permeability interval are determined according to the target dataset corresponding to the reservoir permeability interval. Based on the membership degree of each target data under different membership functions, the target sample set corresponding to each reservoir permeability interval is determined; wherein, the target sample set includes multiple target data. For each reservoir permeability interval, the target sample set corresponding to the reservoir permeability interval is divided into a first sample set and a second sample set. The pre-constructed initial reservoir permeability prediction model is trained using the first sample set and the second sample set respectively, to obtain two target reservoir permeability prediction models corresponding to the reservoir permeability interval.

[0007] In this embodiment of the application, before generating multiple target data, the method further includes: Acquire multiple raw data sets; these raw data sets include multiple raw logging curves and corresponding reservoir permeability curves, with each raw logging curve corresponding to a specific type. The logging curves in each raw data set are depth-corrected to obtain multiple intermediate data sets; these intermediate data sets include multiple corrected logging curves and corresponding reservoir permeability curves. The corrected logging curves from each intermediate data point are preprocessed to obtain the data to be analyzed; the data to be analyzed includes multiple logging curves and corresponding reservoir permeability curves.

[0008] In this embodiment of the application, multiple target data are generated, including: For each set of data to be analyzed, determine the correlation degree of each logging curve to be analyzed in the data; whereby the correlation degree characterizes the degree of correlation between the logging curve to be analyzed and the corresponding reservoir permeability curve. For each type, the target correlation for that type is determined based on the correlation of each logging curve to be analyzed for that type. Based on the relevance of each target, a set of target types is determined; wherein, the set of target types includes at least one target type.

[0009] In this application embodiment, multiple reservoir permeability ranges are determined based on various target data, including: Based on the permeability curves of each reservoir, multiple reservoir permeability intervals were determined using a fuzzy C-means clustering algorithm.

[0010] In this embodiment of the application, for each reservoir permeability interval, the parameter values ​​in the membership function corresponding to the reservoir permeability interval are determined based on the target dataset corresponding to the reservoir permeability interval, including: Based on the permeability curves of each reservoir in the target dataset corresponding to the reservoir permeability range, determine the mean and standard deviation of the reservoir permeability. The mean and standard deviation of reservoir permeability are defined as the mean and standard deviation of the membership function corresponding to the reservoir permeability interval.

[0011] In this embodiment of the application, the target sample set corresponding to each reservoir permeability interval is determined based on the membership degree of each target data under different membership functions, including: For each target data, determine the membership degree of the target data under each membership function, and determine the target sample set to which the target data belongs based on the membership function corresponding to the maximum membership degree.

[0012] In this embodiment of the application, for each reservoir permeability range, the target sample set corresponding to the reservoir permeability range is divided into a first sample set and a second sample set, including: According to the fuzzy rules, the target sample set corresponding to the reservoir permeability range is divided into the first sample set and the second sample set.

[0013] In this embodiment of the application, for each reservoir permeability interval, a pre-constructed initial reservoir permeability prediction model is trained using a first sample set and a second sample set respectively to obtain two target reservoir permeability prediction models corresponding to that reservoir permeability interval, including: The first sample set is divided into a first training set and a first test set, and the second sample set is divided into a second training set and a second test set; The initial reservoir permeability prediction model was trained using the first training set and the second training set, respectively, to obtain the first model to be analyzed and the second model to be analyzed. The first model to be analyzed is tested using the first test set to obtain the first test result; the second model to be analyzed is tested using the second test set to obtain the second test result. If both the first test result and the second test result meet the preset conditions, then both the first model to be analyzed and the second model to be analyzed are determined as target reservoir permeability prediction models; if the first test result or the second test result does not meet the preset conditions, then return to the step of determining multiple reservoir permeability intervals based on each target data and determining the target dataset corresponding to each reservoir permeability interval, until both the first test result and the second test result meet the preset conditions.

[0014] A second aspect of this application provides a method for predicting reservoir permeability, comprising: Obtain the raw logging data to be predicted; wherein, the raw logging data includes multiple measured logging curves, each of which corresponds to a target type; The original logging data is preprocessed to obtain target logging data; wherein, the target logging data includes multiple processed logging curves; Determine the target reservoir permeability range corresponding to the target logging data; The target logging data is input into two target reservoir permeability prediction models corresponding to the target reservoir permeability range obtained by the reservoir permeability prediction model generation method described in the first aspect above, and the reservoir permeability prediction result corresponding to the target logging data is determined.

[0015] In this embodiment of the application, determining the target reservoir permeability range corresponding to the target logging data includes: Determine the membership degree of the target logging data under the membership function corresponding to each preset reservoir permeability interval, and determine the target reservoir permeability interval corresponding to the target logging data based on the membership function corresponding to the maximum membership degree.

[0016] In this embodiment of the application, the target logging data is input into two target reservoir permeability prediction models corresponding to the target reservoir permeability interval obtained by the reservoir permeability prediction model generation method described in the first aspect above, and the reservoir permeability prediction result corresponding to the target logging data is determined, including: The first and second prediction results output by the two target reservoir permeability prediction models are weighted and fused to obtain the reservoir permeability prediction result corresponding to the target logging data.

[0017] A third aspect of this application provides a reservoir permeability prediction model generation apparatus, the apparatus comprising: The target data generation module is used to generate multiple target data; the target data includes multiple target logging curves and corresponding reservoir permeability curves, and each target logging curve corresponds to a target type. The interval and dataset determination module is used to determine multiple reservoir permeability intervals based on various target data, and to determine the target dataset corresponding to each reservoir permeability interval; wherein, the reservoir permeability curves contained in all target data in the target dataset belong to the same reservoir permeability interval. The membership function determination module is used to construct the membership function corresponding to each reservoir permeability interval based on fuzzy logic, and determine the parameter values ​​in the membership function corresponding to the reservoir permeability interval based on the target dataset corresponding to the reservoir permeability interval. The sample set determination module is used to determine the target sample set corresponding to each reservoir permeability interval based on the membership degree of each target data under different membership functions; wherein, the target sample set includes multiple target data. The model training module is used to divide the target sample set corresponding to each reservoir permeability interval into a first sample set and a second sample set, and train the pre-built initial reservoir permeability prediction model using the first sample set and the second sample set respectively, so as to obtain two target reservoir permeability prediction models corresponding to the reservoir permeability interval.

[0018] A fourth aspect of this application provides a reservoir permeability prediction device, the device comprising: The data acquisition module is used to acquire the raw logging data to be predicted; wherein, the raw logging data includes multiple measured logging curves, and each measured logging curve corresponds to a target type. The data preprocessing module is used to preprocess the raw logging data to obtain target logging data; wherein, the target logging data includes multiple processed logging curves; The target interval determination module is used to determine the target reservoir permeability interval corresponding to the target logging data; The reservoir permeability prediction module is used to input the target logging data into two target reservoir permeability prediction models corresponding to the target reservoir permeability interval obtained by the reservoir permeability prediction model generation method described in the first aspect above, and to determine the reservoir permeability prediction result corresponding to the target logging data.

[0019] The fifth aspect of this application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the reservoir permeability prediction model generation method described in the first aspect or the reservoir permeability prediction method described in the second aspect.

[0020] The sixth aspect of this application provides a machine-readable storage medium storing instructions that, when executed by a processor, configure the processor to perform the reservoir permeability prediction model generation method described in the first aspect or the reservoir permeability prediction method described in the second aspect.

[0021] The reservoir permeability prediction model generation method, apparatus, equipment, and medium provided in this application combine fuzzy logic and machine learning. First, fuzzy logic is used to process the uncertainty and ambiguity in well logging data, effectively aiding in data classification and resolving the issue of ambiguity in data belonging to multiple reservoir permeability ranges. Then, machine learning is used to extract potential patterns and features from the fuzzy logic-processed data, thereby constructing a highly accurate reservoir permeability prediction model. This method can not only handle complex geological data but also address issues such as reservoir heterogeneity and data incompleteness, thus providing more accurate and stable permeability prediction results. It effectively solves the problem of low prediction accuracy in existing technologies that predict reservoir permeability based on the correlation between porosity and reservoir permeability.

[0022] Other features and advantages of the embodiments of this application will be described in detail in the following detailed description section. Attached Figure Description

[0023] The accompanying drawings are provided to further illustrate the embodiments of this application and form part of the specification. They are used together with the following detailed description to explain the embodiments of this application, but do not constitute a limitation on the embodiments of this application. In the drawings: Figure 1 The schematic diagram illustrates a process flow diagram of the reservoir permeability prediction model generation method according to an embodiment of this application; Figure 2 The diagram illustrates the curve of the loss function during model training in an embodiment of this application. Figure 3 This illustration shows a comparison between the prediction results of the LSTM model on the data processed by fuzzy logic and the actual reservoir permeability in an embodiment of this application. Figure 4 This illustration shows a comparison between the prediction results of the LSTM model for data without fuzzy logic processing and the actual reservoir permeability in an embodiment of this application. Figure 5 The schematic diagram illustrates a flow chart of the reservoir permeability prediction method according to an embodiment of this application; Figure 6 This schematic diagram illustrates the structural block diagram of a reservoir permeability prediction model generation device according to an embodiment of this application; Figure 7 This schematic diagram illustrates the structural block diagram of a reservoir permeability prediction device according to an embodiment of this application; Figure 8 The diagram illustrates the internal structure of a computer device according to an embodiment of this application.

[0024] Explanation of reference numerals in the attached figures A01 - Processor; A02 - Network Interface; A03 - Internal Memory; A04 - Display Screen; A05 - Input Device; A06 - Non-volatile Storage Media; B01 - Operating System; B02 - Computer Program. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only for illustration and explanation of the embodiments of this application and are not intended to limit the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0026] It should be noted that if the embodiments of this application involve directional indicators (such as up, down, left, right, front, back, etc.), the directional indicators are only used to explain the relative positional relationship and movement of each component in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indicators will also change accordingly.

[0027] Furthermore, if the embodiments of this application involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, features defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the technical solutions of various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. If the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed in this application.

[0028] In view of the problem of low prediction accuracy in related technologies that predict reservoir permeability based on the correlation between porosity and reservoir permeability, this application provides a reservoir permeability prediction model generation and prediction method, apparatus, equipment and medium. The reservoir permeability prediction model generation and prediction method, apparatus, equipment and medium provided in this application will be described in detail below with reference to the accompanying drawings and through specific embodiments and implementation methods.

[0029] Figure 1 The illustration shows a schematic flowchart of the reservoir permeability prediction model generation method according to an embodiment of this application. Figure 1 As shown in one embodiment of this application, a method for generating a reservoir permeability prediction model is provided, the method including the following steps.

[0030] Step A200 generates multiple target data.

[0031] The target data includes multiple target logging curves and corresponding reservoir permeability curves, with each target logging curve corresponding to a target type.

[0032] In a specific example, to generate multiple target data, the following steps are included before step A200.

[0033] Step A110: Obtain multiple raw data sets.

[0034] The raw data includes multiple raw logging curves and corresponding reservoir permeability curves, with each raw logging curve corresponding to a different type.

[0035] In this example, the raw data includes eight raw logging curves: natural gamma ray (GR), caliper (CAL), self-potential (SP), deep resistivity (RXO), shallow resistivity (RT), sonic transit time (AC), compensated neutron (CNL), density (DEN), and porosity (PHI). Different raw logging curves represent different types.

[0036] Step A120 involves performing depth correction processing on the logging curves in each raw data set to obtain multiple intermediate data sets.

[0037] The intermediate data includes multiple corrected logging curves and corresponding reservoir permeability curves.

[0038] Specifically, since the data obtained by logging instruments are closely related to depth, ensuring the accurate correspondence between depth data and other measurement data is crucial to the reliability of geological and physical information. Therefore, depth correction is a key step in the preprocessing of logging data.

[0039] In the process of depth correction, a method based on correlation function is often used. The principle is roughly as follows: the two original logging curves that need to be corrected for depth are regarded as two discrete sequences of equal length. Each discrete sequence contains N sampling points, which represent the measurement values ​​at different depth positions. By analyzing the linear correlation between the two discrete sequences, it is determined whether the measurement values ​​of the two original logging curves at the same depth position are similar, thereby achieving depth correction.

[0040] In this example, step A120 uses the depth correction function built into the CIFlog software to perform depth correction processing on the logging curves in the original data. Depth correction ensures accurate correspondence between the logging curves, better reflecting actual geological conditions and providing a reliable basis for subsequent target data determination and analysis.

[0041] Step A130: Perform data preprocessing on the corrected logging curves in each intermediate data set to obtain the data to be analyzed.

[0042] The data to be analyzed includes multiple logging curves and corresponding reservoir permeability curves.

[0043] In practical applications, data sources are diverse and formats are complex. Raw data may contain duplicates, missing data, and outliers. Duplicate data can be removed; missing data can introduce uncertainty, and handling methods include deletion, replacement, or imputation. Deletion is suitable for large samples, replacement methods are determined based on the type of logging curve, and imputation methods mainly include regression imputation and multiple imputation. Outliers are often considered noise data, and handling methods include limiting the data range, logical verification, and establishing matching rules. Preprocessing the corrected logging curves using simple methods can effectively improve data quality.

[0044] In this example, for each corrected logging curve, step A130 performs the following data preprocessing operation.

[0045] Step A131: Clean the corrected logging curves to obtain the first preprocessed curve.

[0046] Data cleaning includes handling duplicate, missing, and outlier values. Specifically, missing and duplicate values ​​can be removed using deletion methods; outliers can be handled using the isolation forest method. Additionally, for missing values, imputation methods (such as mean imputation or regression imputation) can be considered to avoid data loss.

[0047] Step A132: Perform data augmentation on the first preprocessed curve to obtain the second preprocessed curve.

[0048] Specifically, since the number of sample points after data cleaning is usually reduced, data augmentation is used to process the data to achieve a more balanced distribution of the original samples.

[0049] Step A133: Perform data transformation on the first preprocessed curve to obtain the logging curve to be analyzed.

[0050] Data transformation may include normalization (such as scaling the data to the range of [0, 1]) to meet the needs of the algorithm or analysis model, thereby preparing for subsequent modeling.

[0051] This example demonstrates how data preprocessing improves data quality, which in turn enhances the performance of the trained model, enabling it to make better predictions, reduce overfitting, and improve its accuracy and generalization ability.

[0052] In this embodiment of the application, step A200 includes the following steps.

[0053] Step A210: For each piece of data to be analyzed, determine the correlation degree of each logging curve to be analyzed in that piece of data.

[0054] Correlation characterizes the degree of correlation between the logging curve to be analyzed and the corresponding reservoir permeability curve.

[0055] Specifically, the correlation of each logging curve to be analyzed can be determined by methods such as the Pearson correlation coefficient. This application does not specifically limit the calculation method of the correlation.

[0056] Step A220: For each type, determine the target correlation corresponding to that type based on the correlation of each logging curve to be analyzed for that type.

[0057] In a specific example, step A220 can determine the average correlation of each logging curve of the same type as the target correlation of that type.

[0058] Step A230: Determine the set of target types based on the relevance of each target.

[0059] The target type set includes at least one target type.

[0060] In a specific example, step A230 sorts the target relevances in descending order and determines the type corresponding to the top N (which can be preset, for example, 5) target relevances as the target type. For example, porosity curves (PHI), density curves (DEN), acoustic transit time curves (AC), compensated neutron curves (CNL), and natural gamma curves (GR), which are representative and can reflect the reservoir pore structure, are selected as target logging curves, i.e., used as input training variables for the model.

[0061] In another specific example, step A230 determines the type corresponding to the relevance greater than a preset relevance threshold as the target type.

[0062] Well logging instruments acquire a wide variety of logging curves. In this embodiment, the target data is selected by determining the target type. Compared with the original data, the target data contains fewer logging curves. By discarding logging curves that are unrelated to reservoir permeability and selecting logging curves that are related to reservoir permeability, the input dimensions of the model are reduced, thereby improving model performance.

[0063] Step A300: Based on the target data, determine multiple reservoir permeability intervals and determine the target dataset corresponding to each reservoir permeability interval.

[0064] In this target dataset, all reservoir permeability curves contained in the target data belong to the same reservoir permeability range.

[0065] In this embodiment of the application, step A300 includes the following steps.

[0066] Step A310: Based on the permeability curves of each reservoir, multiple reservoir permeability intervals are determined using a fuzzy C-means clustering algorithm.

[0067] Specifically, the Fuzzy-cmeans (FCM) clustering algorithm is used to analyze reservoir permeability curves, which can flexibly and accurately divide reservoir permeability intervals, adapting to the complexity and uncertainty of data. This method has higher accuracy and adaptability than traditional hard clustering methods, which helps to better understand reservoir characteristics and provides a more reliable foundation for subsequent reservoir analysis and modeling.

[0068] Step A400: For each reservoir permeability interval, construct the membership function corresponding to the reservoir permeability interval based on fuzzy logic, and determine the parameter values ​​in the membership function corresponding to the reservoir permeability interval based on the target dataset corresponding to the reservoir permeability interval.

[0069] Fuzzy logic, specifically designed to handle uncertainty and fuzziness, differs from traditional binary logic in that it allows elements to have continuous membership degrees, enabling more flexible descriptions of complex situations and fuzzy concepts. By defining the membership degrees of fuzzy sets through membership functions and combining them with "If-Then" form fuzzy rules for reasoning, fuzzy logic effectively handles fuzzy conditional and concluding relationships, enhancing the system's analytical and decision-making capabilities. In summary, fuzzy logic improves prediction accuracy by quantifying errors and fuzziness and integrating them with measurement data.

[0070] In a specific example, based on the continuous and relatively smooth distribution characteristics of the well logging data, the membership function is chosen to be a Gaussian function.

[0071] In this embodiment of the application, step A400, for each reservoir permeability interval, determines the parameter value in the membership function corresponding to that reservoir permeability interval, including the following steps.

[0072] Step A410: Determine the mean and standard deviation of reservoir permeability based on the permeability curves of each reservoir in the target dataset corresponding to the reservoir permeability range.

[0073] Step A420: The mean and standard deviation of reservoir permeability are determined as the mean and standard deviation of the membership function corresponding to the reservoir permeability interval.

[0074] The parameters in the membership function include the mean and standard deviation.

[0075] This application embodiment analyzes the reservoir permeability curve and reasonably sets the mean and standard deviation of the membership function to adapt to different reservoir permeability ranges.

[0076] Step A500: Based on the membership degree of each target data under different membership functions, determine the target sample set corresponding to each reservoir permeability interval.

[0077] The target sample set includes multiple target data.

[0078] In this embodiment of the application, step A500 includes the following steps.

[0079] Step A510: For each target data, determine the membership degree of the target data under each membership function, and determine the target sample set to which the target data belongs based on the membership function corresponding to the maximum membership degree.

[0080] This application embodiment selects a membership function based on the maximum membership degree of the target data, thereby determining the target sample set corresponding to each reservoir permeability interval. This can improve the flexibility and adaptability of subsequent classification decisions in uncertain and fuzzy environments.

[0081] Step A600: For each reservoir permeability interval, the target sample set corresponding to the reservoir permeability interval is divided into a first sample set and a second sample set. The pre-constructed initial reservoir permeability prediction model is trained using the first sample set and the second sample set respectively to obtain two target reservoir permeability prediction models corresponding to the reservoir permeability interval.

[0082] In this embodiment of the application, for each reservoir permeability range, step A600 includes the following steps.

[0083] Step A610: According to the fuzzy rules, the target sample set corresponding to the reservoir permeability range is divided into a first sample set and a second sample set.

[0084] Specifically, the data classified by fuzzy logic is divided into a first sample set and a second sample set according to fuzzy rules, based on the target sample set corresponding to the reservoir permeability range. The aim is to classify the samples in a targeted manner according to different permeability characteristics, thereby improving the accuracy and flexibility of classification decisions.

[0085] Step A620: Divide the first sample set into a first training set and a first test set, and divide the second sample set into a second training set and a second test set.

[0086] In a specific example, step A620 divides the first sample set and the second sample set according to a preset division ratio (such as 7:3 or 8:2).

[0087] Step A630: Train the initial reservoir permeability prediction model using the first training set and the second training set respectively to obtain the first model to be analyzed and the second model to be analyzed.

[0088] In a specific example, the initial reservoir permeability prediction model employs a Long Short-Term Memory (LSTM) network model, which is better able to handle sequential data and capture long-term dependencies in the data.

[0089] The LSTM model consists of two LSTM layers and two fully connected (dense) layers. To improve computational efficiency, the number of nodes in the two fully connected layers is set to 16 and 8, respectively. Furthermore, an early stopping strategy is employed to monitor the mean absolute error (MAE) on the test sets (i.e., the first and second test sets) to prevent overfitting. The Adam optimization algorithm is used for training, and Dropout regularization is combined to further reduce the risk of overfitting.

[0090] In this example, the LSTM model uses Mean Squared Error (MSE) as the loss function; for activation functions, ReLU is used in the fully connected layers, and tanh is used in the LSTM layers. During training, the model's ability to fit the data is compared based on the loss function. The parameters of the neural network are adjusted through multiple training iterations, and the trend of the loss function ultimately indicates whether the model has achieved a good fit.

[0091] Of course, it is understandable that other existing models can be used for the initial reservoir permeability prediction model, and this application does not specifically limit the structure of the initial reservoir permeability prediction model.

[0092] Step A640: Test the first model to be analyzed using the first test set to obtain the first test result; test the second model to be analyzed using the second test set to obtain the second test result.

[0093] The test results of the test sets (i.e., the first test set and the second test set) can be obtained by calculating the similarity or error between the reservoir permeability curve predicted by the model and the actual reservoir permeability curve (i.e., the reservoir permeability curve contained in the target data).

[0094] For example, mean squared error (MSE) or mean absolute error (MAE) can be used to quantitatively measure the prediction error of a model, or other indicators (such as the Pearson correlation coefficient) can be used to measure the similarity between two curves, thereby determining the test results.

[0095] Step A650: If both the first test result and the second test result meet the preset conditions, then both the first model to be analyzed and the second model to be analyzed are determined as target reservoir permeability prediction models; if the first test result or the second test result does not meet the preset conditions, then return to execute step A300 until both the first test result and the second test result meet the preset conditions.

[0096] The preset condition can be defined as the test result being greater than a preset judgment threshold. That is, when the first test result is greater than the preset judgment threshold, it indicates that the first test result satisfies the preset condition.

[0097] Understandably, when the first or second test result does not meet the preset conditions, it indicates that the currently trained model does not yet meet the actual requirements and needs further adjustment. In this case, return to step A300, and generate a new reservoir permeability range by adjusting the relevant parameters in the fuzzy C-means clustering algorithm (such as the number of clusters (C), fuzziness parameter (m), tolerance error (ε), and maximum number of iterations (max_iter)).

[0098] The following application example will be used to explain step A600.

[0099] We have 1542 well logging data and permeability samples from a certain oilfield (i.e., the target sample set corresponding to a certain reservoir permeability range includes 1542 target data points). After fuzzy logic classification, the data is divided into two samples (i.e., the first sample set and the second sample set) according to fuzzy rules. For the sample with the larger data volume, it is divided into a training set and a test set (i.e., the first training set and the first test set) in an 8:2 ratio. The training set is then input into an LSTM model for training. The training results are as follows... Figure 2 As shown, according to Figure 2 It can be seen that after 800 iterations, the model parameters gradually converged, the loss function tended to stabilize, and finally a fitting model that met expectations was obtained.

[0100] After completing model training, the predictive ability of the trained model (i.e., the first model to be analyzed) is tested using a test set. The test results are as follows: Figure 3 As shown. According to Figure 3 It can be seen that the LSTM model's prediction results for reservoir permeability (corresponding to...) Figure 3 LSTM predicted permeability and actual reservoir permeability (corresponding to) Figure 3 The correlation coefficient of the actual penetration rate in the middle ( The accuracy reached 0.86, with a mean absolute error of 0.058. Furthermore, the prediction results of this LSTM model on data without fuzzy logic processing are as follows: Figure 4 As shown. According to Figure 4 It can be seen that the LSTM model's prediction results for reservoir permeability (corresponding to...) Figure 4 The unfuzzy LSTM predicted permeability in the reservoir (and the actual reservoir permeability) are compared with the actual reservoir permeability (corresponding to...). Figure 4 The correlation coefficient of the actual penetration rate in the middle ( The accuracy reached 0.75, with a mean absolute error of 0.098. Figure 3 and Figure 4 By comparison, it can be seen that the LSTM model built from large sample data processed by fuzzy logic classification has... The improvement of 0.11 indicates that, with the same sample size, the model built by combining fuzzy logic classification can improve performance while maintaining the original accuracy. Compared with traditional regression models, machine learning methods perform better in reservoir permeability prediction, and selecting the appropriate model based on different permeability characteristics yields better fitting performance. Therefore, combining fuzzy logic with machine learning techniques can significantly improve the accuracy and reliability of reservoir property estimation, while also possessing good generalization ability, thus enhancing the overall accuracy of the prediction task.

[0101] As can be seen, the reservoir permeability prediction model generation method provided in this application combines fuzzy logic and machine learning. First, fuzzy logic is used to process the uncertainty and ambiguity in well logging data, effectively helping to classify the data and resolve the problem of ambiguity in the classification of data across multiple reservoir permeability ranges. Then, machine learning is used to extract potential patterns and features from the data processed by fuzzy logic, thereby constructing a highly accurate reservoir permeability prediction model. This method can not only handle complex geological data but also address issues such as reservoir heterogeneity and data incompleteness, thus providing more accurate and stable permeability prediction results. It effectively solves the problem of low prediction accuracy in existing technologies that predict reservoir permeability based on the correlation between porosity and reservoir permeability.

[0102] Specifically, in well logging data, especially when dealing with complex low-permeability reservoirs, the data often contains significant noise and uncertainty. Fuzzy logic can handle this fuzzy information through "if-then" rules, overcoming the dependence of traditional methods on data quality. Fuzzy logic technology, by introducing membership functions and fuzzy rules, can address the uncertainty and fuzziness in the data. By using fuzzy rules to divide the data into different samples, targeted prediction models for different reservoir permeability ranges can be established. Compared to traditional data processing methods, fuzzy logic does not require explicit classification criteria for the data and can still generate reliable prediction results even with incomplete or noisy data.

[0103] Furthermore, traditional machine learning methods primarily rely on data quality and the explicitness of features. This method, by incorporating fuzzy logic, makes the machine learning model more accurate in handling fuzzy and nonlinear relationships. By using the classification results of fuzzy logic as input to machine learning, it can not only process complex well logging data more accurately but also improve accuracy in prediction, especially in low-permeability reservoirs, demonstrating higher reliability and wider applicability.

[0104] Based on the principle of model selection, this application establishes reservoir permeability prediction models for each reservoir permeability range, taking into account the different data characteristics that may exist in different reservoir permeability ranges. This ensures that each model achieves a balance between fitting ability and generalization ability, thereby better adapting to the characteristics of different reservoir permeability ranges and improving the accuracy of reservoir permeability prediction.

[0105] Figure 1 This is a flowchart illustrating a method for generating a reservoir permeability prediction model in one embodiment. It should be understood that, although... Figure 1 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise explicitly stated herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 1 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0106] Figure 5 A schematic flowchart illustrating the reservoir permeability prediction method according to an embodiment of this application is shown. Figure 5 As shown in one embodiment of this application, a reservoir permeability prediction method is provided, which may include the following steps.

[0107] Step B100: Obtain the raw logging data to be predicted.

[0108] The original logging data includes multiple measured logging curves, each of which corresponds to a target type.

[0109] Step B200: Perform data preprocessing on the original logging data to obtain the target logging data.

[0110] The target logging data includes multiple processed logging curves.

[0111] In this embodiment, the method for preprocessing the raw logging data in step B200 is the same as the method for preprocessing the corrected logging curve in step A130 of the above embodiment, and will not be described again here. By preprocessing the raw logging data, this embodiment improves data quality, thereby ensuring the accuracy of subsequent model predictions.

[0112] Step B300: Determine the target reservoir permeability range corresponding to the target logging data.

[0113] The target reservoir permeability range is one of the multiple reservoir permeability ranges determined by the reservoir permeability prediction model generation method described in the above embodiments.

[0114] In this embodiment of the application, step B300 includes the following steps.

[0115] Step B310: Determine the membership degree of the target logging data under the membership function corresponding to each preset reservoir permeability interval, and determine the target reservoir permeability interval corresponding to the target logging data based on the membership function corresponding to the maximum membership degree.

[0116] The preset reservoir permeability range is the reservoir permeability range determined by the reservoir permeability prediction model generation method described in the above embodiments.

[0117] Step B400: Input the target logging data into two target reservoir permeability prediction models corresponding to the target reservoir permeability interval obtained by the reservoir permeability prediction model generation method described in the above embodiment, and determine the reservoir permeability prediction result corresponding to the target logging data.

[0118] In a specific example, the method for determining the reservoir permeability prediction result corresponding to the target logging data in step B400 is as follows: weighted fusion of the first prediction result and the second prediction result output by the two target reservoir permeability prediction models to obtain the reservoir permeability prediction result corresponding to the target logging data.

[0119] In another specific example, the method for determining the reservoir permeability prediction result corresponding to the target logging data in step B400 is as follows: determine either the first prediction result or the second prediction result output by the two target reservoir permeability prediction models as the reservoir permeability prediction result corresponding to the target logging data.

[0120] It is worth mentioning that the reservoir permeability prediction results mentioned in the embodiments of this application are reservoir permeability prediction curves, and the first and second prediction results output by the two target reservoir permeability prediction models are also reservoir permeability prediction curves.

[0121] Considering that different reservoir permeability ranges may have different data characteristics, this application embodiment first determines the reservoir permeability range corresponding to the target logging data, and then uses the target reservoir permeability prediction model corresponding to the reservoir permeability range to make predictions, which can effectively improve the prediction accuracy of reservoir permeability.

[0122] Since the reservoir permeability prediction method provided in this application includes the reservoir permeability prediction model generation method provided in the above embodiments, it can also solve the problem of low prediction accuracy in the prior art when predicting reservoir permeability based on the correlation between porosity and reservoir permeability.

[0123] Figure 6 A schematic block diagram of a reservoir permeability prediction model generation apparatus according to an embodiment of this application is shown. Figure 6 As shown in one embodiment of this application, a reservoir permeability prediction model generation device is provided, which may include the following functional modules.

[0124] The target data generation module is used to generate multiple target data sets. These target data sets include multiple target logging curves and corresponding reservoir permeability curves, with each target logging curve corresponding to a specific target type.

[0125] The interval and dataset determination module is used to determine multiple reservoir permeability intervals based on various target data, and to determine the corresponding target dataset for each reservoir permeability interval. Specifically, all reservoir permeability curves contained in the target data within the target dataset belong to the same reservoir permeability interval.

[0126] The membership function determination module is used to construct the membership function corresponding to each reservoir permeability interval based on fuzzy logic, and determine the parameter values ​​in the membership function corresponding to the reservoir permeability interval based on the target dataset corresponding to the reservoir permeability interval.

[0127] The sample set determination module is used to determine the target sample set corresponding to each reservoir permeability interval based on the membership degree of each target data under different membership functions. The target sample set includes multiple target data sets.

[0128] The model training module is used to divide the target sample set corresponding to each reservoir permeability interval into a first sample set and a second sample set, and train the pre-built initial reservoir permeability prediction model using the first sample set and the second sample set respectively, so as to obtain two target reservoir permeability prediction models corresponding to the reservoir permeability interval.

[0129] In this embodiment of the application, the device may further include: The raw data acquisition module is used to acquire multiple raw data sets. These raw data sets include multiple raw logging curves and corresponding reservoir permeability curves, with each raw logging curve corresponding to a specific type.

[0130] The depth correction module is used to perform depth correction processing on the logging curves in each raw data set, obtaining multiple intermediate data sets. These intermediate data sets include multiple corrected logging curves and their corresponding reservoir permeability curves.

[0131] The preprocessing module is used to preprocess the corrected logging curves from various intermediate data sets to obtain the data to be analyzed. This data includes multiple logging curves and their corresponding reservoir permeability curves.

[0132] In this embodiment of the application, the target data generation module may include: The correlation calculation unit is used to determine the correlation degree of each logging curve in the data to be analyzed. The correlation degree characterizes the degree of correlation between the logging curve and the corresponding reservoir permeability curve.

[0133] The target correlation determination unit is used to determine the target correlation for each type based on the correlation of each well logging curve to be analyzed for that type.

[0134] The target type determination unit is used to determine a set of target types based on the relevance of each target. The set of target types includes at least one target type.

[0135] In this embodiment of the application, the interval and dataset determination module may include: The interval determination unit is used to determine multiple reservoir permeability intervals based on the permeability curves of each reservoir using a fuzzy C-means clustering algorithm.

[0136] In this embodiment of the application, the membership function determination module may include: The first parameter value determination unit is used to determine the mean and standard deviation of reservoir permeability based on the permeability curves of each reservoir in the target dataset corresponding to the reservoir permeability range.

[0137] The second parameter value determination unit is used to determine the mean and standard deviation of reservoir permeability as the mean and standard deviation of the membership function corresponding to the reservoir permeability interval.

[0138] In this embodiment of the application, the sample set determination module may include: The sample set determination unit is used to determine the membership degree of each target data under each membership function, and determine the target sample set to which the target data belongs based on the membership function corresponding to the maximum membership degree.

[0139] In this embodiment of the application, the model training module may include: The first sample partitioning unit is used to divide the target sample set corresponding to the reservoir permeability range into the first sample set and the second sample set according to fuzzy rules.

[0140] In this embodiment of the application, the model training module may further include: The second sample partitioning unit is used to partition the first sample set into a first training set and a first test set, and to partition the second sample set into a second training set and a second test set.

[0141] The model training unit is used to train the initial reservoir permeability prediction model using the first training set and the second training set, respectively, to obtain the first model to be analyzed and the second model to be analyzed.

[0142] The model testing unit is used to test the first model to be analyzed using the first test set to obtain the first test result; and to test the second model to be analyzed using the second test set to obtain the second test result.

[0143] The target model determination unit is used to determine the first and second models to be analyzed as target reservoir permeability prediction models if both the first test result and the second test result meet the preset conditions; if the first test result or the second test result does not meet the preset conditions, it returns to the step of determining multiple reservoir permeability intervals based on each target data and determining the target dataset corresponding to each reservoir permeability interval, until both the first test result and the second test result meet the preset conditions.

[0144] Since the reservoir permeability prediction model generation device provided in this application embodiment is a virtual device corresponding to the reservoir permeability prediction model generation method described in the above embodiment, it can also solve the problem of low prediction accuracy in the prior art of predicting reservoir permeability based on the correlation between porosity and reservoir permeability.

[0145] Figure 7 A schematic block diagram of a reservoir permeability prediction device according to an embodiment of this application is shown. Figure 7 As shown in one embodiment of this application, a reservoir permeability prediction device is provided, which may include the following functional modules.

[0146] The data acquisition module is used to acquire the raw logging data to be predicted; wherein, the raw logging data includes multiple measured logging curves, and each measured logging curve corresponds to a target type. The data preprocessing module is used to preprocess the raw logging data to obtain target logging data. The target logging data includes multiple processed logging curves.

[0147] The target interval determination module is used to determine the target reservoir permeability interval corresponding to the target logging data.

[0148] The reservoir permeability prediction module is used to input the target logging data into two target reservoir permeability prediction models corresponding to the target reservoir permeability interval obtained by the reservoir permeability prediction model generation method described in the above embodiments, and to determine the reservoir permeability prediction result corresponding to the target logging data.

[0149] In this embodiment of the application, the target interval determination module may include: The target interval determination unit is used to determine the membership degree of the target logging data under the membership function corresponding to each preset reservoir permeability interval, and to determine the target reservoir permeability interval corresponding to the target logging data based on the membership function corresponding to the maximum membership degree.

[0150] In this embodiment of the application, the reservoir permeability prediction module may include: The reservoir permeability determination unit is used to perform weighted fusion processing on the first prediction result and the second prediction result output by the two target reservoir permeability prediction models to obtain the reservoir permeability prediction result corresponding to the target logging data.

[0151] Since the reservoir permeability prediction device provided in this application embodiment is a virtual device corresponding to the reservoir permeability prediction method described in the above embodiment, it can also solve the problem of low prediction accuracy in the prior art when predicting reservoir permeability based on the correlation between porosity and reservoir permeability.

[0152] This application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the reservoir permeability prediction model generation method or reservoir permeability prediction method described in the above embodiments.

[0153] The electronic device provided in this application includes a processor capable of running the reservoir permeability prediction model generation method of the aforementioned embodiments. Therefore, it can also solve the problem of low prediction accuracy in the prior art when predicting reservoir permeability based on the correlation between porosity and reservoir permeability.

[0154] This application provides a machine-readable storage medium storing instructions that, when executed by a processor, configure the processor to perform the reservoir permeability prediction model generation method or reservoir permeability prediction method described in the above embodiments.

[0155] The machine-readable storage medium provided in this application embodiment stores instructions for causing the machine to execute the reservoir permeability prediction model generation method described in the above embodiment. Therefore, it can also solve the problem of low prediction accuracy in the prior art when predicting reservoir permeability based on the correlation between porosity and reservoir permeability.

[0156] Figure 8 The diagram schematically illustrates the internal structure of a computer device according to an embodiment of this application. Figure 8 As shown in one embodiment of this application, a computer device is provided, which can be a terminal. The computer device includes a processor A01, a network interface A02, a display screen A04, an input device A05, and a memory (not shown) connected via a system bus. The processor A01 provides computing and control capabilities. The memory includes internal memory A03 and a non-volatile storage medium A06. The non-volatile storage medium A06 stores an operating system B01 and a computer program B02. The internal memory A03 provides an environment for the operation of the operating system B01 and the computer program B02 stored in the non-volatile storage medium A06. The network interface A02 is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor A01, it implements a reservoir permeability prediction model generation method or a reservoir permeability prediction method. The display screen A04 of the computer device can be an LCD screen or an e-ink screen. The input device A05 of the computer device can be a touch layer covering the display screen, or a button, trackball, or touchpad set on the casing of the computer device, or an external keyboard, touchpad, or mouse, etc.

[0157] Those skilled in the art will understand that Figure 8 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0158] In one embodiment, both the reservoir permeability prediction model generation device and the reservoir permeability prediction device provided in this application can be implemented as a computer program. The computer program can be implemented in various ways, such as... Figure 8 The device operates on the computer shown. The computer's memory can store various program modules that make up the reservoir permeability prediction model generation device or reservoir permeability prediction device. The computer program, composed of these program modules, causes the processor to execute the steps in the reservoir permeability prediction model generation method or reservoir permeability prediction method described in the various embodiments of this application.

[0159] For example, Figure 8 The computer equipment shown can be used as follows Figure 6 The target data generation module in the reservoir permeability prediction model generation device shown executes step A200, the interval and dataset determination module executes step A300, the membership function determination module executes step A400, the sample set determination module executes step A500, and the model training module executes step A600.

[0160] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0161] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1A device that provides the functions specified in one or more boxes.

[0162] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0163] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0164] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0165] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0166] Computer-readable media include both permanent and non-permanent, removable and non-removable media, which can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0167] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0168] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for generating a reservoir permeability prediction model, characterized in that, The method includes: Multiple target data are generated; among them, the target data includes multiple target logging curves and corresponding reservoir permeability curves, and each target logging curve corresponds to a target type. Based on the target data, multiple reservoir permeability intervals are determined, and the target dataset corresponding to each reservoir permeability interval is determined; wherein, the reservoir permeability curves contained in all target data in the target dataset belong to the same reservoir permeability interval. For each reservoir permeability interval, a membership function corresponding to the reservoir permeability interval is constructed based on fuzzy logic, and the parameter values ​​in the membership function corresponding to the reservoir permeability interval are determined according to the target dataset corresponding to the reservoir permeability interval. Based on the membership degree of each target data under different membership functions, the target sample set corresponding to each reservoir permeability interval is determined; wherein, the target sample set includes multiple target data. For each reservoir permeability interval, the target sample set corresponding to the reservoir permeability interval is divided into a first sample set and a second sample set. The pre-constructed initial reservoir permeability prediction model is trained using the first sample set and the second sample set respectively, to obtain two target reservoir permeability prediction models corresponding to the reservoir permeability interval.

2. The method according to claim 1, characterized in that, Before generating multiple target data, the method further includes: Acquire multiple raw data sets; these raw data sets include multiple raw logging curves and corresponding reservoir permeability curves, with each raw logging curve corresponding to a specific type. The logging curves in each raw data set are depth-corrected to obtain multiple intermediate data sets; these intermediate data sets include multiple corrected logging curves and corresponding reservoir permeability curves. The corrected logging curves from each intermediate data point are preprocessed to obtain the data to be analyzed; the data to be analyzed includes multiple logging curves and corresponding reservoir permeability curves.

3. The method according to claim 2, characterized in that, Generate multiple target data, including: For each set of data to be analyzed, determine the correlation degree of each logging curve to be analyzed in the data; whereby the correlation degree characterizes the degree of correlation between the logging curve to be analyzed and the corresponding reservoir permeability curve. For each type, the target correlation for that type is determined based on the correlation of each logging curve to be analyzed for that type. Based on the relevance of each target, a set of target types is determined; wherein, the set of target types includes at least one target type.

4. The method according to claim 1, characterized in that, Based on the target data, multiple reservoir permeability ranges were determined, including: Based on the permeability curves of each reservoir, multiple reservoir permeability intervals were determined using a fuzzy C-means clustering algorithm.

5. The method according to claim 1, characterized in that, For each reservoir permeability interval, based on the target dataset corresponding to that interval, determine the parameter values ​​in the membership function for that interval, including: Based on the permeability curves of each reservoir in the target dataset corresponding to the reservoir permeability range, determine the mean and standard deviation of the reservoir permeability. The mean and standard deviation of reservoir permeability are defined as the mean and standard deviation of the membership function corresponding to the reservoir permeability interval.

6. The method according to claim 1, characterized in that, Based on the membership degree of each target data under different membership functions, determine the target sample set corresponding to each reservoir permeability interval, including: For each target data, determine the membership degree of the target data under each membership function, and determine the target sample set to which the target data belongs based on the membership function corresponding to the maximum membership degree.

7. The method according to claim 1, characterized in that, For each reservoir permeability interval, the target sample set corresponding to that reservoir permeability interval is divided into a first sample set and a second sample set, including: According to the fuzzy rules, the target sample set corresponding to the reservoir permeability range is divided into the first sample set and the second sample set.

8. The method according to claim 7, characterized in that, For each reservoir permeability interval, the pre-constructed initial reservoir permeability prediction model is trained using the first and second sample sets respectively, resulting in two target reservoir permeability prediction models corresponding to that reservoir permeability interval, including: The first sample set is divided into a first training set and a first test set, and the second sample set is divided into a second training set and a second test set; The initial reservoir permeability prediction model was trained using the first training set and the second training set, respectively, to obtain the first model to be analyzed and the second model to be analyzed. The first model to be analyzed is tested using the first test set to obtain the first test result; the second model to be analyzed is tested using the second test set to obtain the second test result. If both the first test result and the second test result meet the preset conditions, then both the first model to be analyzed and the second model to be analyzed are determined as target reservoir permeability prediction models; if the first test result or the second test result does not meet the preset conditions, then return to the step of determining multiple reservoir permeability intervals based on each target data and determining the target dataset corresponding to each reservoir permeability interval, until both the first test result and the second test result meet the preset conditions.

9. A method for predicting reservoir permeability, characterized in that, include: Obtain the raw logging data to be predicted; wherein, the raw logging data includes multiple measured logging curves, each of which corresponds to a target type; The original logging data is preprocessed to obtain target logging data; wherein, the target logging data includes multiple processed logging curves; Determine the target reservoir permeability range corresponding to the target logging data; The target logging data is input into two target reservoir permeability prediction models corresponding to the target reservoir permeability range obtained by the reservoir permeability prediction model generation method according to any one of claims 1 to 8, and the reservoir permeability prediction result corresponding to the target logging data is determined.

10. The method according to claim 9, characterized in that, Determining the target reservoir permeability range corresponding to the target logging data includes: Determine the membership degree of the target logging data under the membership function corresponding to each preset reservoir permeability interval, and determine the target reservoir permeability interval corresponding to the target logging data based on the membership function corresponding to the maximum membership degree.

11. The method according to claim 9, characterized in that, The target logging data is input into two target reservoir permeability prediction models corresponding to the target reservoir permeability interval obtained by the reservoir permeability prediction model generation method according to any one of claims 1 to 8, and the reservoir permeability prediction result corresponding to the target logging data is determined, including: The first and second prediction results output by the two target reservoir permeability prediction models are weighted and fused to obtain the reservoir permeability prediction result corresponding to the target logging data.

12. A reservoir permeability prediction model generation device, characterized in that, The device includes: The target data generation module is used to generate multiple target data; the target data includes multiple target logging curves and corresponding reservoir permeability curves, and each target logging curve corresponds to a target type. The interval and dataset determination module is used to determine multiple reservoir permeability intervals based on various target data, and to determine the target dataset corresponding to each reservoir permeability interval; wherein, the reservoir permeability curves contained in all target data in the target dataset belong to the same reservoir permeability interval. The membership function determination module is used to construct the membership function corresponding to each reservoir permeability interval based on fuzzy logic, and determine the parameter values ​​in the membership function corresponding to the reservoir permeability interval based on the target dataset corresponding to the reservoir permeability interval. The sample set determination module is used to determine the target sample set corresponding to each reservoir permeability interval based on the membership degree of each target data under different membership functions; wherein, the target sample set includes multiple target data. The model training module is used to divide the target sample set corresponding to each reservoir permeability interval into a first sample set and a second sample set, and train the pre-built initial reservoir permeability prediction model using the first sample set and the second sample set respectively, so as to obtain two target reservoir permeability prediction models corresponding to the reservoir permeability interval.

13. A reservoir permeability prediction device, characterized in that, The device includes: The data acquisition module is used to acquire the raw logging data to be predicted; wherein, the raw logging data includes multiple measured logging curves, and each measured logging curve corresponds to a target type. The data preprocessing module is used to preprocess the raw logging data to obtain target logging data; wherein, the target logging data includes multiple processed logging curves; The target interval determination module is used to determine the target reservoir permeability interval corresponding to the target logging data; The reservoir permeability prediction module is used to input the target logging data into two target reservoir permeability prediction models corresponding to the target reservoir permeability interval obtained by the reservoir permeability prediction model generation method according to any one of claims 1 to 8, and to determine the reservoir permeability prediction result corresponding to the target logging data.

14. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the reservoir permeability prediction model generation method according to any one of claims 1 to 8 or the reservoir permeability prediction method according to any one of claims 9 to 11.

15. A machine-readable storage medium storing instructions thereon, characterized in that, When executed by a processor, this instruction causes the processor to be configured to perform the reservoir permeability prediction model generation method according to any one of claims 1 to 8 or the reservoir permeability prediction method according to any one of claims 9 to 11.

Images