Analytical methods, systems, equipment and media for fracturing tight, low-permeability oil and gas reservoirs

Abstract

This invention discloses an analysis method, system, equipment, and medium for fracturing tight, low-permeability oil and gas reservoirs, relating to the field of tight reservoir stimulation technology. The key technical points are: collecting characteristic parameters corresponding to multiple oil and gas wells; classifying the characteristic parameters to obtain training and testing sets, wherein the classification is implemented using a relative distance classification algorithm; constructing a fracturing analysis network model using the training and testing sets, introducing objective and probability functions to adaptively correct the slope and intercept of the linear function of the fracturing analysis network model, updating the fracturing analysis network model based on the adaptive correction results, and obtaining a trained fracturing analysis network model; analyzing the characteristic parameters of actual oil and gas wells using the trained fracturing analysis network model to obtain production rates under different fracturing parameters, and optimizing the fracturing parameters with the goal of maximizing production rates.

Description

Technical Field

[0001] This invention relates to the field of tight reservoir stimulation technology, and more specifically, to analytical methods, systems, equipment, and media for fracturing tight, low-permeability oil and gas reservoirs. Background Technology

[0002] With the large-scale application of domestic oil and gas exploration and development and reservoir stimulation technologies (such as hydraulic fracturing exploration operations in coalbed methane, tight sandstone, and shale oil and gas), a wealth of geological and engineering data has been accumulated. This data is characterized by its large volume, diverse information types, and the fact that traditional fracturing design methods are time-consuming and redundant in their calculations, failing to meet the demands of analyzing such massive amounts of data. Furthermore, the hydraulic fracturing process involves numerous complex physical processes with strong coupling between various physical parameters, making it impossible for traditional design models to accurately reveal the relationship between fracture morphology and production parameters in actual formations. Summary of the Invention

[0003] This invention provides an analysis method, system, equipment, and medium for fracturing tight, low-permeability oil and gas reservoirs. It solves the problem that traditional fracturing design methods based on large datasets are time-consuming and redundant, failing to meet the analysis requirements of existing large amounts of data. This invention utilizes data classification processing to process characteristic parameters, identifying the primary and secondary factors affecting oil and gas well production, forming useful training and testing sets. Then, this invention uses objective and probability functions to adaptively modify the constructed fracturing analysis network model, thereby improving model convergence speed and analysis accuracy. Finally, the established fracturing analysis network model is used to analyze the production corresponding to different fracturing parameters under actual formation conditions of oil and gas wells, thereby optimizing the fracturing parameters.

[0004] The above-mentioned technical objective of the present invention is achieved through the following technical solution:

[0005] The first aspect of this application provides an analytical method for fracturing tight, low-permeability oil and gas reservoirs, the method comprising:

[0006] Collect characteristic parameters corresponding to multiple oil and gas wells, including geological parameters, drilling parameters, fracturing parameters, and production parameters;

[0007] The feature parameters are classified to obtain a training set and a test set, wherein the classification is implemented using a relative distance classification algorithm.

[0008] A fracturing analysis network model is constructed using the training and testing sets. An objective function and a probability function are introduced to adaptively correct the slope and intercept of the linear function of the fracturing analysis network model. The fracturing analysis network model is updated based on the adaptive correction results to obtain a trained fracturing analysis network model.

[0009] The characteristic parameters of actual oil and gas wells are analyzed using a trained fracturing analysis network model to obtain the production rate under different fracturing parameters, and the fracturing parameters are optimized with the goal of maximizing production rate.

[0010] In one implementation scheme, the feature parameters are classified, specifically as follows:

[0011] Construct a sample matrix and a target matrix for multiple oil and gas wells, wherein the row vector of the sample matrix is ​​the sequence number of the oil and gas well, and the column vector is the geological parameters, drilling parameters, and fracturing parameters; the row vector of the target matrix is ​​the sequence number of the oil and gas well, and the column vector is the production parameters.

[0012] The target matrix is ​​divided into three categories of target sub-matrices according to the production output of the oil and gas wells corresponding to the production parameters; the three categories of target sub-matrices represent the excellent, medium and poor production output of oil and gas wells, respectively;

[0013] Three target sub-matrices corresponding to oil and gas wells are randomly selected from the three types of target sub-matrices, and the first relative distance between the multiple oil and gas wells and the randomly selected three oil and gas wells is calculated.

[0014] Traverse the first relative distance, divide the multiple oil and gas wells into three clusters according to the minimum relative distance, and calculate the cluster center of the three clusters;

[0015] Calculate the second relative distance between all sample matrices and the class centers, reclassify and iterate according to the minimum relative distance. If the sum of squared errors is less than the threshold, the classification ends and the classification result is obtained; otherwise, continue iterating until convergence.

[0016] In one implementation, based on the classification results, sample matrices corresponding to the three types of oil and gas wells are selected, and training data and test data of the sample matrices corresponding to the three types of oil and gas wells are selected according to a preset ratio. Training sets and test sets are obtained based on the training data and test data, respectively.

[0017] In one embodiment, the fracturing analysis network model consists of an initiation layer, an intermediate layer, and a terminal layer, wherein the initiation layer, the intermediate layer, and the terminal layer are composed of different numbers of grid cells. The grid cells of the initiation layer and the grid cells of the intermediate layer are connected by a linear function. The grid cells of the intermediate layer are connected by a hyperbolic tangent function to the values ​​calculated by the linear function, and the values ​​calculated by the hyperbolic tangent function are passed to the terminal layer.

[0018] In one implementation, the objective function is calculated as F(i) = k∑(t) l -z k ), where z k This represents the value calculated by the grid cell in the terminal layer, where k is a constant, l represents the l-th calculation, and t lRepresents the actual value of the terminal layer mesh;

[0019] The formula for calculating the probability function is: Where j represents the j-th grid cell, N represents the total number of computational grid cells, and i represents the i-th grid cell.

[0020] In one implementation scheme, an objective function and a probability function are introduced to adaptively correct the slope and intercept of the linear function of the fracturing analysis network model, specifically including:

[0021] Based on the calculated value of the objective function, adaptive adjustments are made based on a given probability threshold;

[0022] When the selection probability calculated by the probability function of the grid cells in the initial layer and the grid cells in the intermediate layer is less than the probability threshold, the slope and intercept of the linear function remain unchanged; or, when the selection probability calculated by the probability function of the grid cells in the initial layer and the grid cells in the intermediate layer is less than the probability threshold, the slope and intercept of the linear function are adaptively corrected.

[0023] In one implementation scheme, the formula for adaptively correcting the slope and intercept of the linear function is as follows: Where b represents a random number in the interval [0,1], and w j This represents the slope after adaptive correction. w represents the slope of the l-th operation of the linear function. j l+1 Let θ represent the slope of the (l+1)th operation of the linear function. j θ represents the adaptively corrected intercept. j l θ represents the intercept of the l-th operation of the linear function. j l+1 It is the intercept of the (l+1)th operation of the linear function.

[0024] A second aspect of this application provides an analytical system for fracturing tight, low-permeability oil and gas reservoirs, the system comprising:

[0025] The data acquisition module is used to collect characteristic parameters corresponding to multiple oil and gas wells, including geological parameters, drilling parameters, fracturing parameters, and production parameters.

[0026] The data processing module is used to classify the feature parameters to obtain a training set and a test set, wherein the classification is implemented using a relative distance classification algorithm.

[0027] The model construction module is used to construct a fracturing analysis network model using the training set and the test set. It introduces an objective function and a probability function to adaptively correct the slope and intercept of the linear function of the fracturing analysis network model. Based on the adaptive correction results, it updates the fracturing analysis network model to obtain a trained fracturing analysis network model.

[0028] The analysis module is used to analyze the characteristic parameters of actual oil and gas wells using a trained fracturing analysis network model, obtain the production rate under different fracturing parameters, and optimize the fracturing parameters with the goal of maximizing the production rate.

[0029] A third aspect of this application provides an electronic device comprising a processor, a memory, and a computer program stored in the memory and executable by the processor, wherein the computer program, when executed by the processor, implements the steps of the analytical method for fracturing tight, low-permeability oil and gas reservoirs as described in the first aspect of this application.

[0030] A third aspect of this application provides a computer-readable storage medium storing a computer program, wherein when executed by a processor, the computer program implements the steps of the analytical method for fracturing tight, low-permeability oil and gas reservoirs as described in the first aspect of this application.

[0031] Compared with the prior art, the present invention has the following beneficial effects:

[0032] This invention provides an analytical method for fracturing tight, low-permeability oil and gas reservoirs. It utilizes data classification to process characteristic parameters, identifying primary and secondary factors affecting oil and gas well production, and forming useful training and testing sets. Then, the invention adaptively modifies the constructed fracturing analysis network model using objective and probability functions, thereby improving model convergence speed and analytical accuracy. Finally, the established fracturing analysis network model is used to analyze the production rates corresponding to different fracturing parameters under actual formation conditions, thus optimizing the fracturing parameters.

[0033] Furthermore, the second to fourth aspects of this application also provide an analysis system, equipment, and medium for fracturing tight, low-permeability oil and gas reservoirs, which have the same technical effects as the aforementioned analysis method for fracturing tight, low-permeability oil and gas reservoirs, and will not be described in detail here. Attached Figure Description

[0034] The accompanying drawings, which are included to provide a further understanding of embodiments of the invention and form part of this application, do not constitute a limitation thereof. In the drawings:

[0035] Figure 1 A flowchart illustrating an analytical method for fracturing tight, low-permeability oil and gas reservoirs provided in this application embodiment;

[0036] Figure 2 This is a data graph of geological parameters, drilling parameters, fracturing parameters, and production parameters provided in the embodiments of this application.

[0037] Figure 3 This is a classification map of the training sample classification contour values ​​provided in the embodiments of this application;

[0038] Figure 4 This is a graph showing the results of the production sample classification calculation provided in the embodiments of this application;

[0039] Figure 5 This is a graph showing the ranking of the impact of geological, fracturing, and production parameters on production output, as provided in the embodiments of this application.

[0040] Figure 6 This is a comparison chart of the analysis results and the actual results of the training set provided in the embodiments of this application;

[0041] Figure 7 This is a comparison chart of the analysis results and actual results of the test set provided in the embodiments of this application;

[0042] Figure 8 This is a schematic diagram of the fracturing analysis network model provided in the embodiments of this application;

[0043] Figure 9 This is a schematic diagram of the analysis system for fracturing tight, low-permeability oil and gas reservoirs provided in an embodiment of this application. Detailed Implementation

[0044] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the embodiments and accompanying drawings. The illustrative embodiments and descriptions of the present invention are only used to explain the present invention and are not intended to limit the present invention.

[0045] It should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0046] As mentioned in the background, with the large-scale application of domestic oil and gas exploration and development and reservoir stimulation technologies (such as hydraulic fracturing operations in coalbed methane, tight sandstone, and shale oil and gas industries), a large amount of geological and engineering data has been accumulated. This data is characterized by its large volume, diverse information types and categories, and the fact that traditional fracturing design methods are time-consuming and redundant in their calculations, failing to meet the analysis needs of this massive amount of data. Furthermore, the hydraulic fracturing process involves many complex physical processes with strong coupling between various physical parameters, making it impossible for traditional design models to accurately reveal the relationship between the fracture morphology and production parameters in actual formations. Therefore, in view of the deficiencies described in the background art, this application provides an analysis method for fracturing tight, low-permeability oil and gas reservoirs. It utilizes data classification processing to process characteristic parameters, identify the primary and secondary factors affecting oil and gas well production, and form useful training and testing sets. Then, the invention uses objective and probability functions to adaptively modify the constructed fracturing analysis network model, thereby improving the model's convergence speed and analysis accuracy. Finally, the established fracturing analysis network model is used to analyze the production corresponding to different fracturing parameters under actual formation conditions of oil and gas wells, thereby optimizing the fracturing parameters.

[0047] The method provided in this embodiment will be described in detail below with reference to specific implementation methods. Please refer to [link / reference]. Figure 1 , Figure 1 A flowchart illustrating the analytical method for fracturing tight, low-permeability oil and gas reservoirs provided in this application embodiment. The method includes the following steps:

[0048] S110, Collect characteristic parameters corresponding to multiple oil and gas wells, wherein the characteristic parameters include geological parameters, drilling parameters, fracturing parameters and production parameters;

[0049] Specifically, geological parameters mainly include reservoir characteristics such as well number, geological structure, reservoir depth, reservoir thickness, permeability, porosity, oil, gas and water saturation, TOC content, temperature, formation pressure, triaxial stress, brittleness index, and fracture development parameters. Drilling parameters mainly include horizontal section length, drilling success rate, recorded gas content, vertical depth, and drilling fluid density. Fracturing parameters mainly include section length, number of perforation clusters, segment distance, perforation cluster spacing, perforation cluster number, fracturing flow rate, sand ratio, total fluid volume, total sand volume, fluid intensity, sand intensity, maximum operating pressure, average operating pressure, pump shutdown pressure, and microseismic monitoring fracture sweep length, height and width, and flowback rate. Production parameters mainly include oil, gas, and water production, wellhead pressure, fluid level, bottom hole pressure, water production rate, and gas-liquid ratio.

[0050] S120, the feature parameters are classified to obtain a training set and a test set, wherein the classification is implemented using a relative distance classification algorithm.

[0051] In this embodiment, fractured oil and gas wells from the same block or with similar formations are selected, and then wells with complete data are further screened. Finally, a database is established based on well number, geology, drilling, fracturing, and production data. Figure 2 The sample database matrix shown.

[0052] S130, a fracturing analysis network model is constructed using the training set and the test set. An objective function and a probability function are introduced to adaptively correct the slope and intercept of the linear function of the fracturing analysis network model. The fracturing analysis network model is updated based on the adaptive correction results to obtain a trained fracturing analysis network model.

[0053] Specifically, in one embodiment, the fracturing analysis network model consists of an initial layer, an intermediate layer, and a terminal layer. Each layer comprises a different number of grid cells. The grid cells in the initial layer and the intermediate layer are connected via a linear function. The grid cells in the intermediate layer are connected using a hyperbolic tangent function to calculate the values ​​obtained from the linear function, and the calculated values ​​are then passed to the terminal layer. In this embodiment, a fracturing analysis network model is constructed using generated training and testing sets. The network includes an initial layer, an intermediate layer, and a terminal layer, each composed of a different number of grid cells. The grid cells in the initial layer and the intermediate layer are connected via a linear function. The intermediate layer grid cells then calculate the values ​​obtained from the linear function using a hyperbolic tangent function and pass the result to the terminal layer. Based on the error between the terminal value and the true value, an objective function and a probability function are introduced to adaptively correct the slope and intercept of the linear function between the initial layer and the intermediate layer grid cells. Iterative calculations are then performed using the topological network model until the error between the calculated value and the true value reaches a set value. The grid cells in the initial layer can input feature parameters from the training set.

[0054] S140: The trained fracturing analysis network model is used to analyze the characteristic parameters of the actual oil and gas well to obtain the production under different fracturing parameters, and the fracturing parameters are optimized with the maximum production as the objective.

[0055] In this embodiment, analyzing the characteristic parameters of actual oil and gas wells based on a trained network model to obtain production rates under different fracturing parameters is a conventional technique for those skilled in the art, and will not be elaborated upon here. Those skilled in the art should understand that the actual characteristic parameters refer to geological parameters, drilling parameters, and fracturing parameters. Geological parameters mainly include reservoir characteristics such as well number, geological structure, reservoir depth, reservoir thickness, permeability, porosity, oil, gas, and water saturation, TOC content, temperature, formation pressure, triaxial stress, brittleness index, and fracture development parameters. Drilling parameters mainly include parameters such as horizontal section length, drilling success rate, recorded gas content, vertical depth, and drilling fluid density. Fracturing parameters mainly include parameters such as section length, number of perforation clusters, segment distance, perforation cluster spacing, perforation cluster number, fracturing flow rate, sand ratio, total fluid volume, total sand volume, fluid intensity, sand intensity, maximum operating pressure, average operating pressure, pump shutdown pressure, and parameters such as fracture sweep length, height, and width monitored by microseismic monitoring, and flowback rate.

[0056] In one implementation scheme, the feature parameters are classified, specifically as follows:

[0057] Construct a sample matrix and a target matrix for multiple oil and gas wells. The row vectors of the sample matrix represent the well serial numbers, and the column vectors represent geological parameters, drilling parameters, and fracturing parameters. The row vectors of the target matrix represent the well serial numbers, and the column vectors represent production parameters. Divide the target matrix into three categories of target sub-matrices based on the production output of the oil and gas wells corresponding to the production parameters. These three categories of target sub-matrices represent excellent, medium, and poor oil and gas well production, respectively. Randomly select three oil and gas wells from the three categories of target sub-matrices and calculate the first relative distance between the multiple oil and gas wells and the randomly selected three wells. Iterate through the first relative distances and divide the multiple oil and gas wells into three clusters based on the minimum relative distance, calculating the cluster centers of the three clusters. Calculate the second relative distance between all sample matrices and the cluster centers, and reclassify and iterate based on the minimum relative distance. If the sum of squared errors is less than a threshold, the classification ends, and the classification result is obtained; otherwise, continue iterating until convergence.

[0058] In this embodiment, after collecting the feature data from step S110 above, an n×m two-dimensional matrix data (sample matrix) is first generated, with oil and gas well information data as rows and geological parameters, drilling parameters, fracturing parameters, and production parameters as columns. Here, 'a' represents the number of oil and gas wells (sample number), and 'm' represents the sum of geological parameters, drilling and completion parameters, and fracturing parameters (attribute feature parameters). The remaining production parameters can be used to generate an a×k (number of production parameters) two-dimensional matrix data (target matrix). The target matrix is ​​divided into three categories: excellent, medium, and poor production. Then, a 3×k matrix (three class centers) is randomly generated as initial values. The relative distance between well 'a' and the three extracted wells is calculated using the following formula:

[0059] In the formula: d(x, y) is the relative distance; x i Let 'a' be a column vectors representing the production parameters of each oil and gas well; y i This is the column vector of the selected oil and gas wells.

[0060] Compare the relative distances between well A and the three selected wells. Based on the minimum relative distance, well A can be initially divided into three clusters. The cluster centers of the three clusters after classification are recalculated. The specific calculation formula is as follows:

[0061] Then, recalculate the relative distances of all samples to the new class centers, and reclassify and iterate. If the sum of squared errors is less than the required value, the classification ends; otherwise, continue iterating until convergence. The formula for calculating the sum of squared errors is as follows:

[0062] In one embodiment, based on the classification results, sample matrices corresponding to three types of oil and gas wells are selected. Training and test data for these sample matrices are then selected according to a preset ratio, and training and test sets are obtained based on the training and test data, respectively. Specifically, after selecting sample matrices corresponding to the three types of wells, sample matrices and target matrices are selected from these three types of wells according to a preset ratio of 90%. The selected 90% forms the training set, and the remaining 10% forms the test set. As is common knowledge to those skilled in the art, the preset ratio can also be 85%, 80%, etc. For example, based on the above embodiment, this embodiment divides the target matrix into three categories: excellent, medium, and poor production. Then, a 3×k matrix (3 class centers) is randomly generated as the initial value. The relative distance between well a and the three extracted wells is calculated according to the data processing method of this invention. Then, the samples are classified according to the calculated relative distance. The distribution of the training sample classification profile values ​​is as follows: Figure 3 As shown, the oil production sample classification calculation results are as follows: Figure 4 As shown in the figure. The ranking of the impact of geological, fracturing, and production parameters on oil production is as follows. Figure 5 As shown. Finally, based on the classification results, sample matrices corresponding to the three types of wells are selected. Then, sample matrices and target matrices are selected from these three types of wells at a ratio of 90% each. The selected 90% can form the training set, and the remaining 10% is the test set.

[0063] Specifically, after the feature parameters of the training set are passed through the initial layer and intermediate networks, the grid input value of the intermediate layer is: h i =∑ i w ij x i -θ j The output value is: H j =f(∑ i wij x i +θ j The calculated value of the grid cell in the terminal layer is: z k =∑ j y j x j +θ j The corresponding mesh error of the fracturing analysis network model is: It should be noted that the calculation of grid error is a current technique.

[0064] In one embodiment, the objective function is calculated as F(i) = k∑(t) l -z k ), where z k This represents the value calculated by the grid cell in the terminal layer, where k is a constant, l represents the l-th calculation, and t l Represents the actual value of the terminal layer mesh;

[0065] The formula for calculating the probability function is: Where j represents the j-th grid cell, N represents the total number of computational grid cells, and i represents the i-th grid cell.

[0066] Specifically, the slopes and intercepts of the intermediate and terminal layers are combined into an objective array. To find the optimal linear function intercept and slope, the objective function in this embodiment is set as F(i)=k∑(t l -z k Then, the selection probability of each grid is calculated, as follows: Here, k is usually taken as 1.

[0067] In one embodiment, an objective function and a probability function are introduced to adaptively correct the slope and intercept of the linear function of the fracturing analysis network model. Specifically, this includes: adaptively correcting the linear function based on a given probability threshold according to the calculated value of the objective function; when the selection probability calculated by the probability function of the grid cells in the initial layer and the grid cells in the intermediate layer is less than the probability threshold, the slope and intercept of the linear function remain unchanged; or, when the selection probability calculated by the probability function of the grid cells in the initial layer and the grid cells in the intermediate layer is less than the probability threshold, adaptively correcting the slope and intercept of the linear function.

[0068] In this embodiment, to accelerate the iterative convergence speed of the intermediate layer network units, the slope and intercept calculated by the linear function are adaptively corrected based on the calculated value of the objective function and a given probability threshold, as follows:

[0069]

[0070] Furthermore, the formula for adaptively correcting the slope and intercept of the linear function is as follows: Where b represents a random number in the interval [0, 1], w j This represents the slope after adaptive correction. This represents the slope of the linear function in the l-th operation. Let θ represent the slope of the (l+1)th operation of the linear function. j This represents the intercept after adaptive correction. This represents the intercept of the l-th operation of the linear function. It is the intercept of the linear function in its (1+1)th operation.

[0071] The corrected values ​​are used to update the topology network model, then the calculation is repeated, and finally the predicted values ​​and relative errors are calculated using RMSE and R². 2 The results are evaluated using three error metrics: MRE, M2RE, and M3RE. The specific calculation formulas for the three error metrics are as follows:

[0072] In the formula, y i y0 is the true value; y0 is the predicted value corresponding to the sample data; y m This represents the average value of the sample data. If the relative error does not meet the requirements, the slope and intercept values ​​of the grid cells are recalculated, and the network cells of the fracturing analysis network model are updated. The calculation is iterated again until the error index meets the requirements. This completes the construction and training of the fracturing analysis network model, resulting in the final fracturing analysis network model as shown below. Figure 8 As shown.

[0073] For example, based on the above-described implementation scheme for constructing and updating the fracturing analysis network model, the fracturing analysis network model obtains analysis results based on the training set, which are then compared with the actual results. Figure 6 As shown, the analysis results of this invention based on the test set are compared with the actual results, for example... Figure 7 As shown.

[0074] Corresponding to the embodiments of the above-described analytical methods for fracturing tight, low-permeability oil and gas reservoirs, this invention also provides an analytical system for fracturing tight, low-permeability oil and gas reservoirs, such as... Figure 9 As shown, the system includes:

[0075] The data acquisition module 910 is used to acquire characteristic parameters corresponding to multiple oil and gas wells, wherein the characteristic parameters include geological parameters, drilling parameters, fracturing parameters and production parameters;

[0076] The data processing module 920 is used to classify the feature parameters to obtain a training set and a test set, wherein the classification is implemented using a relative distance classification algorithm.

[0077] The model construction module 930 is used to construct a fracturing analysis network model using the training set and the test set, introduce an objective function and a probability function to adaptively correct the slope and intercept of the linear function of the fracturing analysis network model, update the fracturing analysis network model based on the adaptive correction result, and obtain a trained fracturing analysis network model.

[0078] Analysis module 940 is used to analyze the characteristic parameters of actual oil and gas wells using a trained fracturing analysis network model, obtain the production rate under different fracturing parameters, and optimize the fracturing parameters with the goal of maximizing the production rate.

[0079] In summary, the analysis system for fracturing tight, low-permeability oil and gas reservoirs provided in this invention has the following beneficial effects: It uses data classification to process characteristic parameters, identifies the primary and secondary factors affecting oil and gas well production, and forms useful training and testing sets. Then, the invention uses objective and probability functions to adaptively modify the constructed fracturing analysis network model, thereby improving model convergence speed and analysis accuracy. Finally, it uses the established fracturing analysis network model to analyze the production corresponding to different fracturing parameters under actual formation conditions of oil and gas wells, thereby optimizing the fracturing parameters.

[0080] In another embodiment of the present invention, an electronic device is provided, comprising one or more processors; a memory, coupled to the processors, for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the steps of the analysis method for fracturing tight, low-permeability oil and gas reservoirs described in the above embodiments. The processor may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), off-the-shelf programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing and control core of the terminal, suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions in a computer storage medium to achieve the corresponding method flow or corresponding function; the processor described in the embodiments of the present invention can be used to execute the operation of the analysis method for fracturing tight, low-permeability oil and gas reservoirs.

[0081] In another embodiment of the present invention, a readable storage medium is provided, specifically a computer-readable storage medium (Memory), which is a memory device in a computer device used to store programs and data. It is understood that the computer-readable storage medium here can include both the built-in storage medium in the computer device and extended storage media supported by the computer device. The computer-readable storage medium provides storage space that stores the operating system of the terminal. Furthermore, the storage space also stores one or more instructions suitable for loading and execution by a processor. These instructions can be one or more computer programs (including program code). It should be noted that the computer-readable storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk storage device. The processor can load and execute one or more instructions stored in the computer-readable storage medium to implement the corresponding steps of the analytical method for fracturing tight, low-permeability oil and gas reservoirs in the above embodiments. Those skilled in the art should understand that the embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented 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.

[0082] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. An analytical method for fracturing tight, low-permeability oil and gas reservoirs, characterized in that, The methods include: Collect characteristic parameters corresponding to multiple oil and gas wells, including geological parameters, drilling parameters, fracturing parameters, and production parameters; The feature parameters are classified to obtain a training set and a test set, wherein the classification is implemented using a relative distance classification algorithm. A fracturing analysis network model is constructed using the training and testing sets. An objective function and a probability function are introduced to adaptively correct the slope and intercept of the linear function of the fracturing analysis network model. Based on the adaptive correction results, the fracturing analysis network model is updated to obtain a trained fracturing analysis network model. The objective function is calculated as follows: ,in, This represents the value calculated by the k-th grid cell in the terminal layer, where c is a constant, k represents the index of the terminal layer grid cell, and l represents the l-th calculation. This represents the true value of the terminal layer grid; the formula for calculating the probability function is: Where j represents the j-th grid cell, N represents the total number of computational grid cells, and i represents the i-th grid cell; An objective function and a probability function are introduced to adaptively correct the slope and intercept of the linear function in the fracturing analysis network model. Specifically, this includes: adaptive correction based on a given probability threshold according to the calculated value of the objective function; when the selection probability calculated by the probability function of the initial layer grid cells and the intermediate layer grid cells is greater than the probability threshold, the slope and intercept of the linear function remain unchanged; when the selection probability calculated by the probability function of the initial layer grid cells and the intermediate layer grid cells is less than or equal to the probability threshold, adaptive correction is performed on the slope and intercept of the linear function. The formula for adaptive correction of the slope and intercept of the linear function is as follows: Where b represents a random number in the interval [0,1], and w j This represents the slope after adaptive correction. This represents the slope of the linear function in the l-th operation. Let θ represent the slope of the (l+1)th operation of the linear function. j This represents the intercept after adaptive correction. This represents the intercept of the l-th operation of the linear function. It is the intercept of the (l+1)th operation of the linear function; The characteristic parameters of actual oil and gas wells are analyzed using a trained fracturing analysis network model to obtain the production rate under different fracturing parameters, and the fracturing parameters are optimized with the goal of maximizing production rate.

2. The analytical method for fracturing tight, low-permeability oil and gas reservoirs according to claim 1, characterized in that, The feature parameters are classified, specifically as follows: Construct a sample matrix and a target matrix for multiple oil and gas wells, wherein the row vector of the sample matrix is ​​the sequence number of the oil and gas well, and the column vector is the geological parameters, drilling parameters, and fracturing parameters; the row vector of the target matrix is ​​the sequence number of the oil and gas well, and the column vector is the production parameters. The target matrix is ​​divided into three categories of target sub-matrices according to the production output of the oil and gas wells corresponding to the production parameters; the three categories of target sub-matrices represent the excellent, medium and poor production output of oil and gas wells, respectively; Randomly select three target sub-matrices corresponding to oil and gas wells from the three types of target sub-matrices, and calculate the first relative distance between the multiple oil and gas wells and the randomly selected three oil and gas wells; Traverse the first relative distance, divide the multiple oil and gas wells into three clusters according to the minimum relative distance, and calculate the cluster center of the three clusters; Calculate the second relative distance between all sample matrices and the class centers, reclassify and iterate according to the minimum relative distance. If the sum of squared errors is less than the threshold, the classification ends and the classification result is obtained; otherwise, continue iterating until convergence.

3. The analytical method for fracturing tight, low-permeability oil and gas reservoirs according to claim 2, characterized in that, Based on the classification results, sample matrices corresponding to the three types of oil and gas wells are selected. Training data and test data of the sample matrices corresponding to the three types of oil and gas wells are selected according to a preset ratio. Training sets and test sets are obtained based on the training data and test data, respectively.

4. The analytical method for fracturing tight, low-permeability oil and gas reservoirs according to claim 1, characterized in that, The fracturing analysis network model consists of an initiation layer, an intermediate layer, and a terminal layer. Each of the initiation layer, intermediate layer, and terminal layer is composed of a different number of grid cells. The grid cells of the initiation layer and the grid cells of the intermediate layer are connected by a linear function. The grid cells of the intermediate layer are connected by a hyperbolic tangent function to the values ​​calculated by the linear function, and the values ​​calculated by the hyperbolic tangent function are then passed to the terminal layer.

5. An analytical system for fracturing tight, low-permeability oil and gas reservoirs, characterized in that the system... include: The data acquisition module is used to collect characteristic parameters corresponding to multiple oil and gas wells, including geological parameters, drilling parameters, fracturing parameters, and production parameters. The data processing module is used to classify the feature parameters to obtain a training set and a test set, wherein the classification is implemented using a relative distance classification algorithm. The model construction module is used to construct a fracturing analysis network model using the training and testing sets. It introduces an objective function and a probability function to adaptively correct the slope and intercept of the linear function of the fracturing analysis network model. Based on the adaptive correction results, the fracturing analysis network model is updated to obtain a trained fracturing analysis network model. The objective function is calculated as follows: ,in, This represents the value calculated by the k-th grid cell in the terminal layer, where c is a constant, k represents the index of the terminal layer grid cell, and l represents the l-th calculation. This represents the true value of the terminal layer grid; the formula for calculating the probability function is: Where j represents the j-th grid cell, N represents the total number of computational grid cells, and i represents the i-th grid cell; An objective function and a probability function are introduced to adaptively correct the slope and intercept of the linear function in the fracturing analysis network model. Specifically, this includes: adaptive correction based on a given probability threshold according to the calculated value of the objective function; when the selection probability calculated by the probability function of the initial layer grid cells and the intermediate layer grid cells is greater than the probability threshold, the slope and intercept of the linear function remain unchanged; when the selection probability calculated by the probability function of the initial layer grid cells and the intermediate layer grid cells is less than or equal to the probability threshold, adaptive correction is performed on the slope and intercept of the linear function. The formula for adaptive correction of the slope and intercept of the linear function is as follows: Where b represents a random number in the interval [0,1], and w j This represents the slope after adaptive correction. This represents the slope of the linear function in the l-th operation. Let θ represent the slope of the (l+1)th operation of the linear function. j This represents the intercept after adaptive correction. This represents the intercept of the l-th operation of the linear function. It is the intercept of the (l+1)th operation of the linear function; The analysis module is used to analyze the characteristic parameters of actual oil and gas wells using a trained fracturing analysis network model, obtain the production rate under different fracturing parameters, and optimize the fracturing parameters with the goal of maximizing the production rate.

6. An electronic device, characterized in that, The electronic device includes a processor, a memory, and a computer program stored in the memory and executable by the processor, wherein when the computer program is executed by the processor, it implements the steps of the analytical method for fracturing tight, low-permeability oil and gas reservoirs as described in any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, wherein when the computer program is executed by a processor, it implements the steps of the analytical method for fracturing tight, low-permeability oil and gas reservoirs as described in any one of claims 1 to 4.

Images