Method and apparatus for predicting fracture network complexity in deep coal and rock gas formations

Abstract

This application provides a method and apparatus for predicting the fracture network complexity of deep coal and shale gas formations. The method includes: obtaining multiple fracturing parameters based on fracturing operation curve data of the target well; constructing a three-dimensional mesh data model based on seismic data, well logging data, and core analysis data of the target well's location; obtaining multiple geological parameters based on the three-dimensional mesh data model; standardizing the multiple fracturing parameters to obtain standardized multiple fracturing parameters; standardizing the multiple geological parameters to obtain standardized multiple geological parameters; and inputting the standardized multiple fracturing parameters and standardized multiple geological parameters into a trained fracture network complexity prediction model, causing the trained fracture network complexity prediction model to output the fracture network complexity. This method improves the accuracy of predicting the fracture network complexity of deep coal and shale gas formations.

Description

Technical Field

[0001] This application relates to the field of oil and gas field development, and in particular to a method and apparatus for predicting the fracture network complexity of deep coal and rock gas formations. Background Technology

[0002] Deep coalbed methane is an unconventional natural gas resource found in coal-bearing strata. Its development requires hydraulic fracturing to create complex fracture networks to improve reservoir permeability. However, deep coalbed methane reservoirs are generally characterized by high stress, low permeability, and complex tectonic movements, resulting in complex and unpredictable fracture network morphologies after fracturing operations.

[0003] Currently, traditional methods mostly rely on empirical formulas or single parameters to predict the fracture network complexity of deep coal and rock gas formations. For example, they often only consider pressure drop rates or peak pressures, lacking systematic mining of multi-dimensional data. This leads to significant discrepancies between the predicted fracture network complexity and the actual fracture network complexity. Summary of the Invention

[0004] This application provides a method and apparatus for predicting the fracture network complexity of deep coal and rock gas, which can solve the problem of significant deviation between the predicted fracture network complexity and the actual fracture network complexity.

[0005] In a first aspect, embodiments of this application provide a method for predicting the fracture network complexity of deep coal and rock gas formations, comprising: obtaining multiple construction parameters based on fracturing construction curve data of a target well; constructing a three-dimensional grid data model based on seismic data, well logging curve data, and core analysis data of the area where the target well is located; obtaining multiple geological parameters based on the three-dimensional grid data model; standardizing the multiple construction parameters to obtain standardized multiple construction parameters; standardizing the multiple geological parameters to obtain standardized multiple geological parameters; and inputting the standardized multiple construction parameters and the standardized multiple geological parameters into a trained fracture network complexity prediction model, so that the trained fracture network complexity prediction model outputs the fracture network complexity.

[0006] In one possible implementation, multiple construction parameters are obtained based on the fracturing construction curve data of the target well, including: generating fracture closure identification curve data and instantaneous pump stop pressure curve data based on the fracturing construction curve data; obtaining the number of fracture closure points and the closure injection ratio of each fracture closure point based on the fracture closure identification curve data; wherein the closure injection ratio is the ratio of the closure time of each fracture closure point to the fracturing injection time; obtaining multiple pressure values ​​under fracture network conditions based on the fracturing construction curve data, fracture closure identification curve, and instantaneous pump stop pressure curve; wherein the fracture network condition refers to the complex fracture network formed after fracturing construction; obtaining the horizontal stress difference coefficient based on the multiple pressure values; and determining the number of fracture closure points, the closure injection ratio of each fracture closure point, and the horizontal stress difference coefficient as multiple construction parameters.

[0007] In one possible implementation, obtaining the horizontal stress difference coefficient based on multiple pressure values ​​includes: obtaining the maximum and minimum horizontal principal stresses based on the multiple pressure values; and calculating the horizontal stress difference coefficient based on the maximum and minimum horizontal principal stresses.

[0008] In one possible implementation, a three-dimensional grid data model is constructed based on seismic data, well logging data, and core analysis data of the target well area. This includes: using sequence stratigraphy to obtain seismic data, well logging data, and core analysis data of the target coal seam; and using Kriging interpolation to construct a three-dimensional grid data model based on the seismic data, well logging data, and core analysis data of the target coal seam.

[0009] In one possible implementation, multiple geological parameters are obtained based on a three-dimensional mesh data model, including: using an anisotropic parameter calculation tool to obtain anisotropic parameters based on the three-dimensional mesh data model; using a spatial surface analysis tool to obtain the coal seam structural dip angle based on the three-dimensional mesh data model; using a structural curvature quantification analysis tool to obtain the structural curvature based on the three-dimensional mesh data model; and determining the anisotropic parameters, coal seam structural dip angle, and structural curvature as multiple geological parameters.

[0010] In one possible implementation, standardized construction parameters and standardized geological parameters are input into a trained fracture network complexity prediction model, causing the trained fracture network complexity prediction model to output fracture network complexity. This includes: inputting standardized construction parameters and standardized geological parameters into the trained fracture network complexity prediction model, and having the trained fracture network complexity prediction model perform the following steps: obtaining a fracture complexity index based on the standardized construction parameters and standardized geological parameters; obtaining a reservoir stimulation volume based on the standardized construction parameters and standardized geological parameters; and outputting the fracture network complexity based on the fracture complexity index and the reservoir stimulation volume.

[0011] In one possible implementation, the method further includes: acquiring sample data from multiple wells; wherein the sample data from each well includes multiple construction parameters, multiple geological parameters, fracture complexity index, and reservoir stimulation volume; dividing the sample data from multiple wells into a training set and a test set; using the training set, training multiple candidate fracture network complexity prediction models to obtain multiple trained candidate fracture network complexity prediction models; inputting the test set into each trained candidate fracture network complexity prediction model for testing and outputting test values; and obtaining the coefficient of determination based on the test values ​​and the actual test values; and determining the trained candidate fracture network complexity prediction model with the highest coefficient of determination as the trained fracture network complexity prediction model.

[0012] In one possible implementation, the method further includes: obtaining the oil and gas production of each well and performing a logarithmic transformation on the oil and gas production of each well to obtain the logarithmic value of the oil and gas production of each well; obtaining multiple standardized construction parameters and multiple standardized geological parameters corresponding to each well; inputting the multiple standardized construction parameters and multiple standardized geological parameters corresponding to each well into a trained fracture network complexity prediction model; enabling the trained fracture network complexity prediction model to obtain a fracture complexity index based on the multiple standardized construction parameters and multiple standardized geological parameters; obtaining the reservoir stimulation volume based on the multiple standardized construction parameters and multiple standardized geological parameters; determining the fracture complexity index and reservoir stimulation volume as comparison parameters; determining the correlation coefficient between the oil and gas production of each well and any comparison parameter based on the logarithmic value of the oil and gas production of each well and any comparison parameter; wherein the correlation coefficient is used to verify whether the trained fracture network complexity prediction model is effective.

[0013] In one possible implementation, the correlation coefficient between oil and gas production and any comparison parameter is determined based on the logarithmic value of oil and gas production from each well and any comparison parameter, using the following formula:

[0014]

[0015]

[0016] In the formula, This represents the correlation coefficient between the oil and gas production of the k-th well and the i-th comparison parameter; This represents the logarithmic value of the oil and gas production of the k-th well. This represents the i-th comparison parameter of the k-th well. This represents the absolute value of the difference between the logarithmic value of the oil and gas production of the k-th well and the i-th comparison parameter. Represents the resolution coefficient; The coefficient represents the correlation between oil and gas production and the i-th comparison parameter, where k represents the k-th well and n represents the number of wells.

[0017] Secondly, embodiments of this application provide a device for predicting the fracture network complexity of deep coal and rock gas formations, comprising: a first acquisition module for acquiring multiple construction parameters based on fracturing construction curve data of a target well; a construction module for constructing a three-dimensional grid data model based on seismic data, well logging curve data, and core analysis data of the area where the target well is located; a second acquisition module for acquiring multiple geological parameters based on the three-dimensional grid data model; a first standardization processing module for standardizing the multiple construction parameters to obtain standardized multiple construction parameters; a second standardization processing module for standardizing the multiple geological parameters to obtain standardized multiple geological parameters; and an output module for inputting the standardized multiple construction parameters and the standardized multiple geological parameters into a trained fracture network complexity prediction model, so that the trained fracture network complexity prediction model outputs the fracture network complexity.

[0018] Thirdly, embodiments of this application provide an electronic device, including: a memory and a processor;

[0019] The memory stores instructions that the computer executes;

[0020] The processor executes computer execution instructions stored in memory, causing the processor to perform the first aspect and / or various possible implementations of the first aspect as described above.

[0021] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the first aspect and / or various possible implementations of the first aspect.

[0022] Fifthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the first aspect and / or various possible implementations of the first aspect.

[0023] The fracture network complexity prediction method and apparatus for deep coal and gas formations provided in this application reflect the dynamic characteristics of fracture propagation during fracturing operations using construction parameters. By constructing a three-dimensional grid data model and acquiring multiple geological parameters, the influence on fracture propagation direction can be quantified, providing geological constraints for fracture network complexity prediction. The multiple construction and geological parameters cover the main construction and geological control factors of fracture network complexity. Standardization eliminates the dimensional differences between different parameters. Combining the standardized construction and geological parameters constructs multi-source input data, solving the problem of traditional methods relying on only a single parameter and insufficient information mining. The multi-source input data enables the trained fracture network complexity prediction model to simultaneously capture the dynamic characteristics of fracture propagation and geological constraints, thereby more comprehensively characterizing fracture network complexity and improving the prediction accuracy. Attached Figure Description

[0024] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0025] Figure 1 A schematic diagram illustrating a scenario for the method of predicting the fracture network complexity of deep coal and rock gas provided in this application embodiment;

[0026] Figure 2 A flowchart illustrating the method for predicting the fracture network complexity of deep coal and rock gas provided in this application embodiment. Figure 1 ;

[0027] Figure 3 A linear correlation diagram of oil and gas production and fracture complexity index for any well provided in the embodiments of this application;

[0028] Figure 4 Linear correlation diagram of oil and gas production and reservoir stimulation volume for any well provided in the embodiments of this application;

[0029] Figure 5 A flowchart illustrating the method for predicting the fracture network complexity of deep coal and rock gas provided in this application embodiment. Figure 2 ;

[0030] Figure 6 A schematic diagram of the structure of the deep coal and rock gas fracture network complexity prediction device provided in the embodiments of this application;

[0031] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0032] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0033] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0034] Figure 1 This is a schematic diagram illustrating a scenario for the method of predicting the fracture network complexity of deep coal and rock gas provided in the embodiments of this application. Figure 1 As shown, it includes: a receiving device 101, a processing device 102, and a display device 103.

[0035] It is understood that the structures illustrated in the embodiments of this application do not constitute a specific limitation on the method for predicting the fracture network complexity of deep coal and rock gas. In other feasible embodiments of this application, the above architecture may include more or fewer components than illustrated, or combine some components, or split some components, or arrange different components, which can be determined according to the actual application scenario and is not limited here. Figure 1 The components shown can be implemented in hardware, software, or a combination of both.

[0036] In the specific implementation process, the receiving device 101 can be an input / output interface or a communication interface, which can acquire the fracturing construction curve data of the target well, as well as the seismic data, well logging curve data and core analysis data of the area where the target well is located.

[0037] The processing device 102 can obtain multiple construction parameters based on the fracturing construction curve data of the target well; construct a three-dimensional grid data model based on the seismic data, well logging curve data and core analysis data of the area where the target well is located; obtain multiple geological parameters based on the three-dimensional grid data model; standardize the multiple construction parameters to obtain standardized multiple construction parameters; standardize the multiple geological parameters to obtain standardized multiple geological parameters; and input the standardized multiple construction parameters and standardized multiple geological parameters into a trained fracture network complexity prediction model, so that the trained fracture network complexity prediction model outputs the fracture network complexity.

[0038] The display device 103 can be used to display the complexity of the stitching mesh.

[0039] It should be understood that the aforementioned processor can be implemented by reading instructions from memory and executing those instructions, or it can be implemented through chip circuitry.

[0040] Furthermore, the network architecture and business scenarios described in the embodiments of this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided in the embodiments of this application. As those skilled in the art will know, with the evolution of network architecture and the emergence of new business scenarios, the technical solutions provided in the embodiments of this application are also applicable to similar technical problems.

[0041] To address the aforementioned technical problems, this application proposes the following technical approach: Traditional methods often rely on empirical formulas or single parameters to predict the fracture network complexity of deep coalbed methane formations, leading to significant discrepancies between predicted and actual fracture network complexity. The inventors have devised a method that considers both construction-controlled and geological-controlled factors, including multiple construction parameters and multiple geological parameters. Standardization eliminates dimensional differences between parameters, and the standardized construction and geological parameters are combined to construct multi-source input data. Using a trained fracture network complexity prediction model allows for the simultaneous capture of dynamic characteristics of fracture propagation and geological constraints, avoiding the problem of insufficient information mining that arises with empirical formulas. This results in a more comprehensive characterization of fracture network complexity and improved prediction accuracy.

[0042] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0043] Figure 2 A flowchart illustrating the method for predicting the fracture network complexity of deep coal and rock gas provided in this application embodiment. Figure 1 The methods include:

[0044] S201: Obtain multiple construction parameters based on the fracturing construction curve data of the target well.

[0045] Among them, the fracturing operation curve is a curve that records time series data during the fracturing operation process, including data such as time series, construction pressure, proppant displacement, and proppant concentration.

[0046] The unit for construction pressure is MPa, the unit for sand-carrying fluid discharge is m³ / min, and the unit for proppant concentration is kg / m³.

[0047] Among them, construction parameters refer to the parameters extracted from the fracturing construction curve to characterize the dynamics of crack propagation.

[0048] Specifically, step S201 includes S2011 to S2015:

[0049] S2011: Generate fracture closure identification curve data and instantaneous pump stop pressure curve data based on fracturing construction curve data.

[0050] In this embodiment, professional fracturing construction analysis software is used to generate fracture closure identification curve data and instantaneous pump stop pressure curve data based on fracturing construction curve data.

[0051] Specifically, open the fracturing operation analysis software and import the fracturing operation curve data; call the fracture closure identification function to generate fracture closure identification curve data. Call the instantaneous pump stop pressure extraction function to extract pressure data 5-10 minutes before and after the instant of pump stop (the end of the operation) and generate the instantaneous pump stop pressure curve.

[0052] S2012: Based on the fracture closure identification curve data, obtain the number of fracture closure points and the closure injection ratio of each fracture closure point; where the closure injection ratio is the ratio of the closure time of each fracture closure point to the fracturing injection time.

[0053] Among them, crack closure identification curve data can also be called G function curve data.

[0054] In this embodiment, the number of fracture closure points is counted based on fracture closure identification curve data. After fracturing operations are completed, the pressure gradually decreases, and the closure of different fractures can cause abrupt changes in the pressure drop pattern. The fracture closure identification curve data can amplify these abrupt changes, and by checking whether the tangent line passing through the origin coincides with the curve, it can be determined whether each abrupt change is a fracture closure point. Finally, the number of tangent lines is the total number of fracture closure points.

[0055] In this embodiment, the fracture closure identification curve data transforms the pressure-time curve after fracturing pump shutdown into a G-value-pressure superimposed derivative curve. The G-value is a mathematical transformation of the time after fracturing pump shutdown; it is not a directly measured physical quantity, but a standardized time index calculated by fracturing construction analysis software based on pump shutdown time and construction parameters. Its core function is to amplify the pressure signal of fracture closure, making the fracture closure point easier to identify.

[0056] In this embodiment, in the fracture closure identification curve data, if the pressure drop superimposed derivative curve shows multiple upward convexities and deviations in the graph, and the tangent line passes through the origin, the number of fracture closure points can be determined by calculating the number of tangent lines. Each tangent line represents one fracture closure point in the reservoir.

[0057] For example, during the first stage of fracturing in the DJX1 well, three rays passing through the origin can be drawn that partially overlap with the superimposed derivative curve of the pressure drop, thus identifying three fracture closure points in this well section.

[0058] In this embodiment, when a complex network of fractures forms underground, the geometric morphology, stress state, and fluid filtration characteristics of different fractures vary significantly, resulting in different closure times, which are represented by multiple independent closure points on the curve. Therefore, the more closure points there are, the more it indicates the formation of a complex network of fractures underground.

[0059] In this embodiment, the pump injection time essentially reflects the effective time for fracturing fluid to maintain fracture propagation. Under the same pumping capacity, a smaller closure injection ratio indicates that the fracturing fluid energy is dispersed and consumed by multiple fracture branches in a complex fracture network, rather than by the extension of a single fracture, indirectly proving the high complexity of the fracture network. Therefore, the smaller the closure injection ratio, the more it indicates the formation of complex fractures underground.

[0060] For example, the first closure time of DJX1 well is 14.1 min, the pump injection time is 183.8 min, and the calculated closure injection ratio is 0.076.

[0061] S2013: Based on the fracturing operation curve data, fracture closure identification curve and instantaneous pump stop pressure curve, obtain multiple pressure values ​​under fracture network conditions; where fracture network conditions refer to the complex fracture network formed after the fracturing operation is completed.

[0062] Optionally, multiple pressure values ​​include stable construction pressure, ground rupture pressure, ground closure pressure, and instantaneous pump stop pressure.

[0063] In this embodiment, the stable construction pressure and ground fracturing pressure are extracted from the fracturing construction curve data, the ground closure pressure is extracted from the fracture closure identification curve, and the instantaneous pump stop pressure is extracted from the instantaneous pump stop pressure curve.

[0064] Among them, stable construction pressure refers to the average value of the pressure during the middle stage of fracturing, ground rupture pressure refers to the peak value before the pressure drops sharply in the early stage of construction, ground closure pressure refers to the pressure value corresponding to the closure point of the fracture, and instantaneous pump stop pressure refers to the pressure peak value at the moment of pump stop.

[0065] S2014: Obtain the horizontal stress difference coefficient based on multiple pressure values.

[0066] Among them, the Horizontal Stress Difference Ratio (HSDR) is a key parameter characterizing the degree of anisotropy of the horizontal principal stress in the formation.

[0067] Specifically, the maximum and minimum horizontal principal stresses are obtained based on multiple pressure values; and the horizontal stress difference coefficient is calculated based on the maximum and minimum horizontal principal stresses.

[0068] In this embodiment, the formula for calculating HSDR is: . Indicates the maximum horizontal principal stress. This represents the minimum horizontal principal stress.

[0069] In this embodiment, when the HSDR value is small, the difference in horizontal principal stress is small, and the directional constraint of the geostress field on fracture propagation is weakened. At this time, the fracturing fluid is more likely to break through natural fractures or weak surfaces, forming a complex fracture network extending in multiple directions. Furthermore, the fracture propagation exhibits a composite mode of tensile fracturing and shear slip, promoting the interweaving of the fracture network and making it easier to form a complex fracture network.

[0070] In this embodiment, the maximum and minimum horizontal principal stresses are obtained based on multiple pressure values ​​using the stress calculation function of the fracturing construction analysis software.

[0071] For example, the first segment of DJX1 well =42.5MPa, =36.8MPa, and the calculated HSDR is 0.155.

[0072] S2015: The number of crack closure points, the closure injection ratio of each crack closure point, and the horizontal stress difference coefficient are determined as multiple construction parameters.

[0073] S202: Construct a three-dimensional grid data model based on seismic data, well logging data, and core analysis data of the target well area.

[0074] Specifically, sequence stratigraphy was used to obtain seismic data, well logging data, and core analysis data of the target coal seam; and kriging interpolation was used to construct a three-dimensional grid data model based on the seismic data, well logging data, and core analysis data of the target coal seam.

[0075] In this embodiment, seismic data, well logging data, and core analysis data of the target well area are imported into specialized geological modeling software. This geological modeling software supports three-dimensional visualization analysis of the seismic data, well logging data, and core analysis data.

[0076] The logging data includes GR, resistivity, and sonic transit time.

[0077] In this embodiment, the sequence stratigraphy method built into the geological modeling software is used to delineate the target layer, i.e. the target coal seam, and to lock the area to be studied, such as 20m below the top boundary of the coal seam, so as to avoid interference from irrelevant strata in the calculation.

[0078] In this embodiment, three-dimensional modeling is selected in the geological modeling software, Kriging interpolation is selected, and the grid size is set, such as 5m×5m×2m.

[0079] S203: Obtain multiple geological parameters based on the three-dimensional grid data model.

[0080] Specifically, step S203 includes S2031~S2034:

[0081] S2031: Anisotropic parameters are obtained using an anisotropic parameter calculation tool based on a three-dimensional mesh data model.

[0082] Among them, anisotropy parameters are quantitative indicators that characterize the degree of difference in the physical properties of rocks in different directions. The core of anisotropy parameters is to reflect the elastic heterogeneity of rocks. This heterogeneity is mainly caused by geological features such as fractures and stratigraphic stratification. It is a key geological parameter for judging whether fracturing fluid can form a complex fracture network.

[0083] In this embodiment, in the rock physical properties module of the geological modeling software, shear wave logging data is called, and anisotropic parameters are calculated based on Thomsen theory using an anisotropic parameter calculation tool.

[0084] Among them, shear wave logging data is measured using the XMAC instrument. The shear wave logging data includes vertical shear wave velocity, horizontal shear wave velocity, and x-direction shear wave velocity.

[0085] The role of Thomsen's theory is to transform the difference in shear wave velocity into quantifiable anisotropic parameters. Because underground rocks have cracks and strata are layered, shear waves propagate at different velocities in different directions. Thomsen's theory is a mature algorithm for converting the difference in shear wave velocity into anisotropic parameters.

[0086] In this embodiment, the anisotropy parameter is the anisotropy intensity parameter.

[0087] The anisotropic strength parameter is used to quantify the overall elastic non-uniformity. Anisotropic strength parameter = (Vsh² - Vsv²) / (2Vsv²) - 2 × (Vsx² - Vsv²) / Vsv², where Vsh represents the horizontal shear wave velocity, Vsv represents the vertical shear wave velocity, and Vsx is the shear wave velocity in the x-direction.

[0088] S2032: Using spatial surface analysis tools, the dip angle of the coal seam structure is obtained based on the three-dimensional mesh data model.

[0089] In this embodiment, the coal seam dip angle includes the principal curvature dip angle along the x-axis, the principal curvature dip angle along the y-axis, and the xy-plane twist rate.

[0090] In this embodiment, the target layer structural surface is selected in the structural interpretation module of the geological modeling software, the spatial surface analysis tool is enabled, and the x-axis principal curvature dip angle, y-axis principal curvature dip angle, and xy-plane torsion rate are checked. The spatial surface analysis tool automatically calculates according to the formula:

[0091]

[0092]

[0093]

[0094] Wherein, Kx represents the x-axis principal curvature dip angle, that is, the principal curvature dip angle along the x-axis direction; Ky represents the y-axis principal curvature dip angle, that is, the principal curvature dip angle along the y-axis direction; Kxy represents the xy plane twist rate, reflecting the degree of twist of the structural surface in the two-dimensional plane.

[0095] For example, the first target layer of well DJX1 has Kx=5.2°, Ky=3.8°, and Kxy=1.5°, indicating that the coal seam structure dips steeper along the x-axis.

[0096] S2033: Using a construction curvature quantization analysis tool, the construction curvature is obtained based on a three-dimensional mesh data model.

[0097] In this embodiment, the construction curvature quantification analysis tool is enabled, and the average curvature is selected. The construction curvature quantification analysis tool automatically calculates the average curvature according to the formula: average curvature = (k1 + k2) / 2, where k1 and k2 are the two principal curvatures.

[0098] At any point on the structural surface, there are two mutually perpendicular directions. The curvature in these two directions is the maximum and minimum among all directions at that point. These are called principal curvatures, denoted as k1, the maximum principal curvature, and k2, the minimum principal curvature, respectively.

[0099] Here, the average curvature reflects the average degree of formation curvature. For example, the average curvature of the first segment of well DJX1 = .

[0100] The greater the structural curvature, the more severe the structural deformation.

[0101] S2034: Anisotropy parameters, coal seam dip angle, and structural curvature are determined as multiple geological parameters.

[0102] S204: Standardize multiple construction parameters to obtain standardized construction parameters.

[0103] In this embodiment, several construction parameters include the number of crack closure points, the closure injection ratio of each crack closure point, and the horizontal stress difference coefficient.

[0104] In this embodiment, construction parameters of different units and magnitudes are converted into a unified scale of 0 to 1 to avoid the trained seam mesh complexity prediction model being affected by the unit.

[0105] S205: Standardize multiple geological parameters to obtain standardized geological parameters.

[0106] In this embodiment, multiple geological parameters include anisotropy parameters, coal seam dip angle, and structural curvature.

[0107] In this embodiment, geological parameters of different units and magnitudes are converted into a unified scale of 0 to 1 to avoid the trained fracture network complexity prediction model being affected by the unit.

[0108] S206: Input the standardized construction parameters and the standardized geological parameters into the trained seam complexity prediction model, so that the trained seam complexity prediction model outputs the seam complexity.

[0109] Specifically, the standardized construction parameters and standardized geological parameters are input into the trained fracture network complexity prediction model, which then performs the following steps: obtaining the fracture complexity index based on the standardized construction parameters and standardized geological parameters; obtaining the reservoir stimulation volume based on the standardized construction parameters and standardized geological parameters; and outputting the fracture network complexity based on the fracture complexity index and reservoir stimulation volume.

[0110] Optionally, the fracture complexity index can be standardized to obtain a standardized fracture complexity index; the reservoir stimulation volume can be standardized to obtain a standardized reservoir stimulation volume. Both the standardized fracture complexity index and reservoir stimulation volume are within the range of 0 to 1.

[0111] Optionally, the weights of the fracture complexity index, the reservoir stimulation volume, and the ranges of different fracture network complexity levels can be preset and adjusted according to actual conditions.

[0112] Optionally, a weighted sum is performed based on the standardized fracture complexity index, the standardized reservoir stimulation volume, the weight of the fracture complexity index, and the weight of the reservoir stimulation volume to obtain a comprehensive fracture network complexity score. If the comprehensive fracture network complexity score is greater than or equal to 8, it is considered high complexity; if the comprehensive fracture network complexity score is greater than or equal to 5 and less than 8, it is considered medium complexity; and if the comprehensive fracture network complexity score is less than 5, it is considered low complexity.

[0113] Optionally, the trained fracture network complexity prediction model is used to predict the fracture network complexity of 120 wells. In this embodiment, the effectiveness of the trained fracture network complexity prediction model is verified, as detailed below:

[0114] SA: Obtain the oil and gas production of each well, and perform a logarithmic transformation on the oil and gas production of each well to obtain the logarithmic value of the oil and gas production of each well.

[0115] The oil and gas production of each well is the oil and gas production of each well at different times, which is essentially a sequence.

[0116] In this embodiment, logarithmic transformation is performed to make the data more stable and more suitable for correlation analysis.

[0117] SB: Obtain the standardized construction parameters and standardized geological parameters for each well.

[0118] SC: Input the standardized construction parameters and standardized geological parameters corresponding to each well into the trained fracture network complexity prediction model; so that the trained fracture network complexity prediction model can obtain the fracture complexity index based on the standardized construction parameters and standardized geological parameters; and obtain the reservoir stimulation volume based on the standardized construction parameters and standardized geological parameters.

[0119] SD: The fracture complexity index and reservoir stimulation volume are determined as comparison parameters.

[0120] Among them, the fracture complexity index is the fracture complexity index of each well at different times, which is essentially a sequence.

[0121] Among them, the reservoir stimulation volume is the reservoir stimulation volume of each well at different times, which is essentially a sequence.

[0122] Optionally, the fracture complexity index can be determined as the first comparison parameter and the reservoir stimulation volume as the second comparison parameter; alternatively, the reservoir stimulation volume can be determined as the first comparison parameter and the fracture complexity index as the second comparison parameter.

[0123] SE: Based on the logarithmic value of oil and gas production of each well and any comparison parameter, determine the correlation coefficient between oil and gas production and any comparison parameter; the correlation coefficient is used to verify whether the trained fracture network complexity prediction model is effective.

[0124] Specifically, based on the logarithmic values ​​of oil and gas production from each well and any comparison parameter, the correlation coefficient between oil and gas production and any comparison parameter is determined using the following formula:

[0125]

[0126]

[0127] In the formula, This represents the correlation coefficient between the oil and gas production of the k-th well and the i-th comparison parameter; This represents the logarithmic value of the oil and gas production of the k-th well. This represents the i-th comparison parameter of the k-th well. This represents the absolute value of the difference between the logarithmic value of the oil and gas production of the k-th well and the i-th comparison parameter. Represents the resolution coefficient; The coefficient represents the correlation between oil and gas production and the i-th comparison parameter, where k represents the k-th well and n represents the number of wells.

[0128] Optionally, the resolution coefficient can be 0.5.

[0129] For example, a correlation coefficient greater than or equal to 0.8 is considered a strong correlation, greater than or equal to 0.6 and less than 0.8 is considered a moderate correlation, and less than 0.6 is considered a weak correlation. The first correlation coefficient is 0.831, and the second correlation coefficient is 0.815. This verifies that the trained seam complexity prediction model is effective.

[0130] For example, multiple construction parameters and multiple geological parameters of any well are input into a trained fracture network complexity prediction model, which outputs a fracture complexity index of 2.32 and a reservoir stimulation volume of... The measured fracture complexity index of the microseismic event was 2.45, and the reservoir stimulation volume was... The relative errors were 5.3% and 3.4% respectively for the oil and gas production of this well. The first correlation coefficient was calculated to be 0.842, verifying that the trained seam complexity prediction model is effective.

[0131] refer to Figure 3-4 , Figure 3 A linear correlation diagram of oil and gas production and fracture complexity index for any well provided in the embodiments of this application; Figure 4 Linear correlation diagram of oil and gas production and reservoir stimulation volume for any well provided in the embodiments of this application.

[0132] like Figure 3 As shown, the linear fitting trend of oil and gas production and fracture complexity index corresponds to the equation y = 6 × 10 -5 x + 0.155, coefficient of determination R² = 0.905.

[0133] like Figure 4 As shown, the linear fitting trend of oil and gas production and reservoir stimulation volume corresponds to the equation y=0.0886x-248.95, with a coefficient of determination R²=0.803.

[0134] In summary, construction parameters reflect the dynamic characteristics of fracture propagation during fracturing. By constructing a three-dimensional mesh data model and acquiring multiple geological parameters, the influence on fracture propagation direction can be quantified, providing geological constraints for fracture network complexity prediction. Multiple construction and geological parameters encompass the main construction and geological controlling factors of fracture network complexity. Standardization eliminates the dimensional differences between parameters, and combining the standardized construction and geological parameters constructs multi-source input data, solving the problem of traditional methods relying on only a single parameter. This multi-source input data allows the trained fracture network complexity prediction model to simultaneously capture the dynamic characteristics of fracture propagation and geological constraints, avoiding insufficient information mining issues that arise from empirical formulas. This results in a more comprehensive characterization of fracture network complexity and improved prediction accuracy.

[0135] refer to Figure 5 , Figure 5 A flowchart illustrating the method for predicting the fracture network complexity of deep coal and rock gas provided in this application embodiment. Figure 2 Based on the above embodiments, this embodiment describes the training process of the trained seam complexity prediction model, as detailed below:

[0136] S501: Obtain sample data from multiple wells; the sample data for each well includes multiple construction parameters, multiple geological parameters, fracture complexity index, and reservoir stimulation volume.

[0137] In this embodiment, the direct measurement of fracture complexity index and reservoir stimulation volume heavily relies on microseismic monitoring technology. By capturing microseismic signals of rock fracturing during hydraulic fracturing, the spatial distribution of the fracture network is inverted, and the fracture complexity index and reservoir stimulation volume are calculated. However, this technology is costly and difficult to implement in the field. Of the 132 wells in the study area, only 12 wells have complete microseismic monitoring data, and data scarcity has become a bottleneck for the large-scale evaluation of parameters. To overcome the data limitation, a machine learning method is used to construct a fracture network complexity prediction model.

[0138] In this embodiment, multiple construction parameters include the number of fracture closure points, the closure injection ratio of each fracture closure point, and the horizontal stress difference coefficient; multiple geological parameters include anisotropy parameters, coal seam dip angle, and structural curvature.

[0139] In this embodiment, a feature matrix is ​​constructed by combining multiple standardized construction parameters and multiple standardized geological parameters; and a label matrix is ​​constructed by combining the fracture complexity index and reservoir stimulation volume.

[0140] In this embodiment, the Z-score normalization method is used to perform Min-Max normalization on the feature matrix and the label matrix, mapping them to the [0,1] interval.

[0141] S502: Divide the sample data from multiple wells into training and test sets.

[0142] In this embodiment, the sample data from multiple wells are divided into a training set and a test set in an 8:2 ratio.

[0143] S503: Using the training set, train multiple candidate seam complexity prediction models to obtain multiple trained candidate seam complexity prediction models.

[0144] In this embodiment, multiple candidate seam complexity prediction models are constructed using Python language and various ensemble learning methods such as random forest, XGBoost, CatBoost, LightGBM, and neural networks.

[0145] In this embodiment, the training set is used to train each candidate seam complexity prediction model to obtain a trained candidate seam complexity prediction model.

[0146] S504: Input the test set into each trained candidate seam complexity prediction model, perform the test, output the test value, and obtain the coefficient of determination based on the test value and the actual test value.

[0147] In this embodiment, the prediction accuracy of each trained candidate seam complexity prediction model is tested using a test set. The accuracy is quantified by the coefficient of determination; the higher the coefficient of determination, the more accurate the model prediction.

[0148] Specifically, the coefficient of determination is obtained based on the test values ​​output by each trained candidate seam complexity prediction model on the test set, and the labels of the test set, i.e., the true test values. The coefficient of determination ranges from 0 to 1, with R² ≥ 0.8 indicating a high-precision model and R² ≥ 0.7 indicating a usable model, where R² represents the coefficient of determination.

[0149] Referring to Table 1, which shows the determination coefficient results for each algorithm provided in the embodiments of this application, the determination coefficients include those for the fracture complexity index and the reservoir stimulation volume. The candidate fracture network complexity prediction model constructed by the XGBoost algorithm outperforms other algorithms on the test set with R²(FCI) = 0.83 and R²(SRV) = 0.80. Here, FCI is the fracture complexity index, and SRV is the reservoir stimulation volume.

[0150] Table 1. Determination coefficient results for each algorithm

[0151]

[0152] S505: The candidate seam complexity prediction model with the highest coefficient of determination is selected as the trained seam complexity prediction model.

[0153] In summary, sample data from multiple wells were acquired and divided into training and testing sets. After training multiple candidate fracture network complexity prediction models using the training set, these models were tested using the testing set. The determination coefficients of each trained candidate model were obtained, and these coefficients were used to quantify the model. The candidate model with the highest determination coefficient was then selected as the final trained fracture network complexity prediction model. A higher determination coefficient results in smaller residuals and more controllable prediction errors, further improving the accuracy of fracture network complexity prediction for deep coal and gas formations.

[0154] Figure 6 This is a schematic diagram of the structure of the deep coal and rock gas fracture network complexity prediction device provided in the embodiments of this application, as shown below. Figure 6 As shown, the deep coal and rock gas fracture network complexity prediction device provided in this embodiment includes: a first acquisition module 601, a construction module 602, a second acquisition module 603, a first standardization processing module 604, a second standardization processing module 605, and an output module 606.

[0155] The first acquisition module 601 is used to acquire multiple construction parameters based on the fracturing construction curve data of the target well.

[0156] Module 602 is used to construct a three-dimensional grid data model based on seismic data, well logging curve data and core analysis data of the target well area.

[0157] The second acquisition module 603 is used to acquire multiple geological parameters based on the three-dimensional grid data model.

[0158] The first standardization processing module 604 is used to standardize multiple construction parameters to obtain standardized construction parameters.

[0159] The second standardization processing module 605 is used to standardize multiple geological parameters to obtain standardized geological parameters.

[0160] The output module 606 is used to input multiple standardized construction parameters and multiple standardized geological parameters into the trained seam complexity prediction model, so that the trained seam complexity prediction model outputs the seam complexity.

[0161] In one possible implementation, the first acquisition module 601 is specifically used for: generating fracture closure identification curve data and instantaneous pump stop pressure curve data based on fracturing construction curve data; acquiring the number of fracture closure points and the closure injection ratio of each fracture closure point based on the fracture closure identification curve data; wherein the closure injection ratio is the ratio of the closure time of each fracture closure point to the fracturing injection time; acquiring multiple pressure values ​​under fracture network conditions based on the fracturing construction curve data, fracture closure identification curve, and instantaneous pump stop pressure curve; wherein the fracture network conditions refer to the complex fracture network formed after fracturing construction; acquiring the horizontal stress difference coefficient based on the multiple pressure values; and determining the number of fracture closure points, the closure injection ratio of each fracture closure point, and the horizontal stress difference coefficient as multiple construction parameters.

[0162] In one possible implementation, the first acquisition module 601 is further configured to: acquire the maximum horizontal principal stress and the minimum horizontal principal stress based on multiple pressure values; and calculate the horizontal stress difference coefficient based on the maximum horizontal principal stress and the minimum horizontal principal stress.

[0163] In one possible implementation, the construction module 602 is specifically used to: acquire seismic data, well logging data and core analysis data of the target coal seam using sequence stratigraphy; and construct a three-dimensional grid data model based on the seismic data, well logging data and core analysis data of the target coal seam using kriging interpolation.

[0164] In one possible implementation, the second acquisition module 603 is specifically used to: acquire anisotropic parameters using an anisotropic parameter calculation tool based on a three-dimensional mesh data model; acquire the coal seam structural dip angle using a spatial surface analysis tool based on a three-dimensional mesh data model; acquire the structural curvature using a structural curvature quantification analysis tool based on a three-dimensional mesh data model; and determine the anisotropic parameters, coal seam structural dip angle, and structural curvature as multiple geological parameters.

[0165] In one possible implementation, the output module 606 is specifically used to: input multiple standardized construction parameters and multiple standardized geological parameters into a trained fracture network complexity prediction model, and the trained fracture network complexity prediction model performs the following steps: the steps include: obtaining a fracture complexity index based on the multiple standardized construction parameters and multiple standardized geological parameters; obtaining a reservoir stimulation volume based on the multiple standardized construction parameters and multiple standardized geological parameters; and outputting the fracture network complexity based on the fracture complexity index and the reservoir stimulation volume.

[0166] In one possible implementation, the deep coal and rock gas fracture network complexity prediction device further includes a training module, which is specifically used for: acquiring sample data from multiple wells; wherein the sample data from each well includes multiple construction parameters, multiple geological parameters, fracture complexity index, and reservoir stimulation volume; dividing the sample data from multiple wells into a training set and a test set; using the training set to train multiple candidate fracture network complexity prediction models respectively, obtaining multiple trained candidate fracture network complexity prediction models; inputting the test set into each trained candidate fracture network complexity prediction model respectively, conducting tests, and outputting test values; and obtaining the determination coefficient based on the test values ​​and the actual test values; and determining the trained candidate fracture network complexity prediction model with the highest determination coefficient as the trained fracture network complexity prediction model.

[0167] In one possible implementation, the deep coal and rock gas fracture network complexity prediction device further includes a verification module. The verification module is specifically used for: acquiring the oil and gas production of each well and performing a logarithmic transformation on the oil and gas production of each well to obtain the logarithmic value of the oil and gas production of each well; acquiring multiple standardized construction parameters and multiple standardized geological parameters corresponding to each well; inputting the multiple standardized construction parameters and multiple standardized geological parameters corresponding to each well into a trained fracture network complexity prediction model; enabling the trained fracture network complexity prediction model to obtain a fracture complexity index based on the multiple standardized construction parameters and multiple standardized geological parameters; obtaining the reservoir stimulation volume based on the multiple standardized construction parameters and multiple standardized geological parameters; determining the fracture complexity index and reservoir stimulation volume as comparison parameters; and determining the correlation coefficient between the oil and gas production of each well and any comparison parameter based on the logarithmic value of the oil and gas production of each well and any comparison parameter; wherein the correlation coefficient is used to verify whether the trained fracture network complexity prediction model is effective.

[0168] In one possible implementation, the correlation coefficient between oil and gas production and any comparison parameter is determined based on the logarithmic value of oil and gas production from each well and any comparison parameter, using the following formula:

[0169]

[0170]

[0171] In the formula, This represents the correlation coefficient between the oil and gas production of the k-th well and the i-th comparison parameter; This represents the logarithmic value of the oil and gas production of the k-th well. This represents the i-th comparison parameter of the k-th well. This represents the absolute value of the difference between the logarithmic value of the oil and gas production of the k-th well and the i-th comparison parameter. Represents the resolution coefficient; The coefficient represents the correlation between oil and gas production and the i-th comparison parameter, where k represents the k-th well and n represents the number of wells.

[0172] The deep coal and rock gas fracture network complexity prediction device provided in this embodiment can execute the method provided in the above method embodiment. Its implementation principle and technical effect are similar, and will not be described in detail here.

[0173] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 7 As shown, the electronic device provided in this embodiment includes at least one processor 701 and a memory 702. Optionally, the electronic device further includes a communication component 703. The processor 701, memory 702, and communication component 703 are connected via a bus.

[0174] In a specific implementation, at least one processor 701 executes computer execution instructions stored in memory 702, causing at least one processor 701 to perform the above-described method.

[0175] The specific implementation process of processor 701 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0176] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0177] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0178] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0179] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0180] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described method.

[0181] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0182] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.

[0183] The division of units is merely a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0184] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0185] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0186] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0187] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0188] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A method for predicting the network complexity of deep coal rock gas, characterized in that, Applied to electronic devices, including: Based on the fracturing operation curve data of the target well, generate fracture closure identification curve data and instantaneous pump stop pressure curve data; Based on the fracture closure identification curve data, the number of fracture closure points and the closure injection ratio of each fracture closure point are obtained; wherein the closure injection ratio is the ratio of the closure time of each fracture closure point to the fracturing injection time. Based on the fracturing operation curve data, the fracture closure identification curve, and the instantaneous pump stop pressure curve, multiple pressure values ​​under fracture network conditions are obtained; wherein the fracture network conditions refer to the complex fracture network formed after the fracturing operation is completed. Based on the multiple pressure values, the horizontal stress difference coefficient is obtained; The number of crack closure points, the closure injection ratio of each crack closure point, and the horizontal stress difference coefficient are determined as multiple construction parameters; A three-dimensional grid data model is constructed based on the seismic data, well logging data, and core analysis data of the target well area. An anisotropy parameter calculation tool is used to obtain an anisotropy parameter according to the three-dimensional mesh data model; the anisotropy parameter calculation formula is: , wherein Vsh represents a horizontal transverse wave velocity, Vsv represents a vertical transverse wave velocity, and Vsx represents an x-direction transverse wave velocity. Using a spatial surface analysis tool, the dip angle of the coal seam structure is obtained based on the three-dimensional mesh data model; wherein, the dip angle of the coal seam structure includes the principal curvature dip angle along the x-axis, the principal curvature dip angle along the y-axis, and the xy-plane twist rate; The construction curvature is obtained using a construction curvature quantization analysis tool based on the three-dimensional mesh data model; wherein the construction curvature is the average curvature, the average curvature = (k1+k2) / 2, and k1 and k2 are the two principal curvatures; The anisotropy parameter, the coal seam dip angle, and the structural curvature are defined as multiple geological parameters; The multiple construction parameters are standardized to obtain standardized construction parameters. The multiple geological parameters are standardized to obtain standardized geological parameters. The standardized construction parameters and the standardized geological parameters are input into the trained seam complexity prediction model, so that the trained seam complexity prediction model outputs the seam complexity. The method further includes: The oil and gas production of each well is obtained, and the oil and gas production of each well is logarithmically transformed to obtain the logarithmic value of the oil and gas production of each well. Obtain the standardized multiple construction parameters and the standardized multiple geological parameters corresponding to each well; The standardized construction parameters and standardized geological parameters corresponding to each well are input into the trained fracture network complexity prediction model; the trained fracture network complexity prediction model obtains the fracture complexity index based on the standardized construction parameters and standardized geological parameters; and obtains the reservoir stimulation volume based on the standardized construction parameters and standardized geological parameters. The fracture complexity index and the reservoir stimulation volume are determined as comparison parameters; Based on the logarithmic value of the oil and gas production of each well and any comparison parameter, determine the correlation coefficient between the oil and gas production and the comparison parameter; wherein the correlation coefficient is used to verify whether the trained fracture network complexity prediction model is effective. The formula for the correlation coefficient is: In the formula, represents the correlation coefficient of the oil and gas production of the kth well and the ith comparison parameter; represents the logarithmic value of the oil and gas production of the kth well, represents the ith comparison parameter of the kth well, represents the absolute value of the difference between the logarithmic value of the oil and gas production of the kth well and the ith comparison parameter, represents the resolution coefficient; represents the correlation coefficient of the oil and gas production and the ith comparison parameter, k represents the kth well, and n represents the number of wells.

2. The method of claim 1, wherein, The step of obtaining the horizontal stress difference coefficient based on the plurality of pressure values ​​includes: Based on the multiple pressure values, the maximum and minimum horizontal principal stresses are obtained; Calculate the horizontal stress difference coefficient based on the maximum horizontal principal stress and the minimum horizontal principal stress.

3. The method according to claim 1, characterized in that, The construction of a three-dimensional grid data model based on seismic data, well logging data, and core analysis data of the target well area includes: Sequence stratigraphy was used to obtain seismic data, well logging data, and core analysis data of the target coal seam. A three-dimensional grid data model was constructed using the Kriging interpolation method based on the seismic data, well logging data, and core analysis data of the target coal seam.

4. The method according to claim 1, characterized in that, The step of inputting the standardized multiple construction parameters and the standardized multiple geological parameters into the trained seam complexity prediction model, so that the trained seam complexity prediction model outputs the seam complexity, includes: The standardized construction parameters and the standardized geological parameters are input into the trained seam network complexity prediction model, and the trained seam network complexity prediction model performs the following steps; The steps include: The crack complexity index is obtained based on the standardized construction parameters and the standardized geological parameters. The reservoir stimulation volume is obtained based on the standardized construction parameters and the standardized geological parameters. The fracture network complexity is output based on the fracture complexity index and the reservoir stimulation volume.

5. The method according to claim 1, characterized in that, The method further includes: Obtain sample data from multiple wells; the sample data for each well includes multiple construction parameters, multiple geological parameters, fracture complexity index, and reservoir stimulation volume; The sample data from the multiple wells are divided into a training set and a test set; Using the training set, multiple candidate seam complexity prediction models are trained to obtain multiple trained candidate seam complexity prediction models. The test set is input into each trained candidate seam complexity prediction model to perform the test and output the test value; and the determination coefficient is obtained based on the test value and the actual test value. The candidate seam complexity prediction model with the highest determination coefficient is determined as the trained seam complexity prediction model.

6. A device for predicting the complexity of fracture networks in deep coal and rock gas formations, characterized in that, Applied to electronic devices, including: The first acquisition module is used to generate fracture closure identification curve data and instantaneous pump stop pressure curve data based on the fracturing construction curve data; Based on the fracture closure identification curve data, the number of fracture closure points and the closure injection ratio of each fracture closure point are obtained; wherein the closure injection ratio is the ratio of the closure time of each fracture closure point to the fracturing injection time. Based on the fracturing operation curve data, the fracture closure identification curve, and the instantaneous pump stop pressure curve, multiple pressure values ​​under fracture network conditions are obtained; wherein the fracture network conditions refer to the complex fracture network formed after the fracturing operation is completed. Based on the multiple pressure values, the horizontal stress difference coefficient is obtained; The number of crack closure points, the closure injection ratio of each crack closure point, and the horizontal stress difference coefficient are determined as multiple construction parameters; The building module is used to construct a three-dimensional grid data model based on seismic data, well logging data and core analysis data of the target well area; The second acquisition module is used to acquire anisotropic parameters based on the three-dimensional mesh data model using an anisotropic parameter calculation tool; the anisotropic parameter calculation formula is: anisotropic intensity parameter = Vsh represents the horizontal shear wave velocity, Vsv represents the vertical shear wave velocity, and Vsx represents the shear wave velocity in the x-direction. Using a spatial surface analysis tool, the coal seam structure dip angle is obtained based on the three-dimensional mesh data model. The coal seam structure dip angle includes the principal curvature dip angle along the x-axis, the principal curvature dip angle along the y-axis, and the xy-plane twist rate. The construction curvature is obtained using a construction curvature quantization analysis tool based on the three-dimensional mesh data model; wherein the construction curvature is the average curvature, the average curvature = (k1+k2) / 2, and k1 and k2 are the two principal curvatures; The anisotropy parameter, the coal seam dip angle, and the structural curvature are defined as multiple geological parameters; The first standardization processing module is used to standardize the multiple construction parameters to obtain standardized multiple construction parameters; The second standardization processing module is used to standardize the multiple geological parameters to obtain standardized multiple geological parameters; The output module is used to input the standardized multiple construction parameters and the standardized multiple geological parameters into the trained seam complexity prediction model, so that the trained seam complexity prediction model outputs the seam complexity. The verification module is used to obtain the oil and gas production of each well and perform logarithmic transformation on the oil and gas production of each well to obtain the logarithmic value of the oil and gas production of each well. Obtain the standardized multiple construction parameters and the standardized multiple geological parameters corresponding to each well; The standardized construction parameters and standardized geological parameters corresponding to each well are input into the trained fracture network complexity prediction model; the trained fracture network complexity prediction model obtains the fracture complexity index based on the standardized construction parameters and standardized geological parameters; and obtains the reservoir stimulation volume based on the standardized construction parameters and standardized geological parameters. The fracture complexity index and the reservoir stimulation volume are determined as comparison parameters; Based on the logarithmic value of the oil and gas production of each well and any comparison parameter, determine the correlation coefficient between the oil and gas production and the comparison parameter; wherein the correlation coefficient is used to verify whether the trained fracture network complexity prediction model is effective. The formula for the correlation coefficient is: In the formula, This represents the correlation coefficient between the oil and gas production of the k-th well and the i-th comparison parameter; This represents the logarithmic value of the oil and gas production of the k-th well. This represents the i-th comparison parameter of the k-th well. This represents the absolute value of the difference between the logarithmic value of the oil and gas production of the k-th well and the i-th comparison parameter. Represents the resolution coefficient; The correlation coefficient between the oil and gas production and the i-th comparison parameter is represented by k, where k represents the k-th well and n represents the number of wells.

Images