Design method, device and medium for high-gas coal roadway CO2 high-pressure gas fracturing

Abstract

The invention discloses a high-gas coal roadway CO2 high-pressure gas fracturing design method, device and medium, and relates to the technical field of coal seam gas governing, the method comprises the following steps: carrying out structured processing and stress field and crack distribution analysis on a coal seam basic parameter data set, carrying out virtual simulation and dynamic updating, and generating a digital twin coal seam model; cO2 high-pressure gas is input into the digital twin coal seam model, node response simulation and fracture propagation calculation are carried out, and a fracture evolution prediction data set is obtained; performing node feasibility verification, risk assessment and parameter iterative optimization on the optimized injection scheme to generate an injection scheme check data set; and performing multi-dimensional evaluation and comprehensive judgment on the injection scheme check data set to obtain a crack effect evaluation result. Through cleaning, structuring and dynamic modeling of coal seam environment parameters, a digital twin coal seam model capable of reflecting coal seam stress and crack distribution characteristics in real time is generated, and accurate reproduction and dynamic updating of a coal seam complex structure are achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of coal seam gas control, in particular to a design method, equipment and medium for CO2 high-pressure gas fracturing of high-gas coal roadway. BACKGROUND

[0002] With the continuous increase of coal mining intensity, the safety and production efficiency of high-gas coal roadway area have been paid more and more attention. In order to effectively release coal seam gas and improve the mechanical properties of coal body, fracturing technology has gradually become an important auxiliary means in coal mining. Among them, the method of using high-pressure fluid as fracturing medium is relatively mature, especially CO2 fracturing, which gradually becomes the key research direction of gas control and coal seam reconstruction due to its phase change expansion effect, inert gas characteristics and strong permeation driving capacity.

[0003] The existing related research and application are mostly concentrated in the "empirical" or "single parameter configuration" fracturing design idea, and lack of systematic modeling and dynamic prediction mechanism for complex coal seam environmental conditions. In high-gas coal roadway, the coal seam structure often has multi-dimensional heterogeneity, and the stress distribution and crack propagation law have significant differences. Only relying on static parameters or single test results for injection design may have uncertainty in crack direction, extension range and connectivity effect. SUMMARY

[0004] In view of the above existing problems, the present application is proposed.

[0005] Therefore, the present application provides a design method for CO2 high-pressure gas fracturing of high-gas coal roadway to solve the problems of uncertainty of crack evolution prediction and insufficient optimization of fracturing scheme under complex coal seam conditions.

[0006] To solve the above technical problems, the present application provides the following technical solutions:

[0007] In the first aspect, the present application provides a design method for CO2 high-pressure gas fracturing of high-gas coal roadway, which includes collecting coal seam environmental parameters of target coal seam for preprocessing and comprehensive analysis, and summarizing and arranging to form a coal seam basic parameter data set;

[0008] The coal seam basic parameter data set is subjected to structured processing, stress field and crack distribution analysis, and virtual simulation and dynamic updating to generate a digital twin coal seam model;

[0009] The CO2 high-pressure gas is input into the digital twin coal seam model for node response simulation and crack propagation calculation to obtain a crack evolution prediction data set;

[0010] The crack evolution prediction data set is subjected to parameter analysis and working condition matching to obtain an optimized injection scheme;

[0011] The node feasibility verification, risk assessment and parameter iterative optimization of the optimized injection scheme are carried out, and an injection scheme verification dataset is generated.

[0012] A multi-dimensional evaluation and comprehensive judgment were performed on the injection scheme verification dataset to obtain the crack effect assessment results.

[0013] As a preferred embodiment of the CO2 high-pressure gas fracturing design method for high-gas coal roadways described in this invention, the following steps are taken: The collected coal seam environmental parameters are preprocessed and comprehensively analyzed, and then summarized to form a coal seam basic parameter dataset.

[0014] The coal seam environmental parameters are cleaned, denoised, and smoothed to obtain a standardized dataset;

[0015] The standardized dataset is processed through unified normalization, comprehensive analysis, and interactive parameter calculation to generate comprehensive coal seam basic parameter indicators.

[0016] The comprehensive coal seam basic parameter indicators are summarized, structured, and uniformly formatted to form a coal seam basic parameter dataset.

[0017] As a preferred embodiment of the CO2 high-pressure gas fracturing design method for high-gas coal roadways described in this invention, the specific steps for structuring the coal seam basic parameter dataset, analyzing the stress field and fracture distribution, performing virtual simulation and dynamic updates, and generating a digital twin coal seam model are as follows.

[0018] The coal seam basic parameter dataset is mapped to a three-dimensional grid, and structured grid data is generated through classification, normalization and spatial interpolation.

[0019] Stress field and potential crack distribution analysis were performed on structured mesh data, and a preliminary digital twin model was generated by combining virtual simulation calculations and dynamic update processing.

[0020] The initial digital twin model is updated through time series simulation, dynamic correction, and real-time data feedback to generate a digital twin coal seam model.

[0021] As a preferred embodiment of the CO2 high-pressure gas-induced fracturing design method for high-gas coal roadways described in this invention, the specific steps of inputting CO2 high-pressure gas into a digital twin coal seam model to perform simulation calculations on fracture propagation and fracture network formation, and obtaining a fracture evolution prediction dataset, are as follows.

[0022] High-pressure CO2 gas is input into the digital twin coal seam model. Through stress-fracture interaction analysis and dynamic simulation, nodal fracture potential information is generated. Spatial interpolation and topological correlation analysis are then performed to generate a continuous three-dimensional fracture propagation field.

[0023] Dynamic simulation and time-step iteration of the continuous three-dimensional crack propagation field are performed to generate time-series crack evolution data;

[0024] Real-time monitoring, feedback correction, and iterative updates are performed on time-series crack evolution data to form a crack evolution prediction dataset.

[0025] As a preferred embodiment of the CO2 high-pressure gas fracturing design method for high-gas coal roadways described in this invention, the steps for analyzing the nodal fracture characteristics and matching the operating conditions of the fracture evolution prediction dataset to obtain an optimized injection scheme are as follows.

[0026] The nodal crack characteristics of the crack evolution prediction dataset are analyzed and the working condition matching calculation is performed to obtain the nodal injection optimization data. Crack-working condition coupling calculation is then performed to generate injection optimization index.

[0027] The injection optimization indicators are spatially integrated and sorted to form an injection optimization scoring field, and the pressure, flow rate and duration of each injection hole location are dynamically matched and calculated to generate the optimal combination of injection parameters.

[0028] The optimal injection parameter combinations are organized and summarized according to spatial location and time order to obtain the optimized injection scheme.

[0029] As a preferred embodiment of the CO2 high-pressure gas fracturing design method for high-gas coal roadways described in this invention, the specific steps of performing node feasibility verification, risk assessment, and parameter iterative optimization of the injection scheme to generate an injection scheme verification dataset are as follows.

[0030] Feasibility calculations and risk analyses of node injection parameters were performed on the optimized injection scheme to obtain an injection verification index dataset.

[0031] Node-level computation and neighborhood coupling analysis are performed on the injected verification index dataset to form node verification indices;

[0032] Spatial integration and sorting of node verification indicators are performed to generate a verification score field. Iterative adjustments and parameter optimizations are then performed to generate an injection scheme verification dataset.

[0033] As a preferred embodiment of the design method for CO2 high-pressure gas fracturing in high-gas coal roadways described in this invention, the specific steps for performing multi-dimensional evaluation and comprehensive judgment on the injection scheme verification dataset to obtain the fracturing effect evaluation result are as follows.

[0034] The injection scheme verification dataset is standardized and preprocessed to obtain multidimensional input data, and node-level nonlinear interaction analysis and mapping operations are performed to generate node mapping results.

[0035] The node mapping results are comprehensively judged at the node level to generate node crack effects. Spatial integration and temporal series synthesis operations are then performed to generate a continuous three-dimensional crack propagation effect field.

[0036] Deviation correction and iterative updates are performed on the continuous three-dimensional crack propagation effect field to obtain crack effect evaluation results.

[0037] As a preferred embodiment of the design method for CO2 high-pressure gas-induced fracturing in high-gas coal roadways described in this invention, the specific steps for correcting deviations and iteratively updating the continuous three-dimensional crack propagation effect field to obtain crack effect evaluation results are as follows.

[0038] By combining the continuous three-dimensional crack propagation effect field with the crack evolution prediction dataset, the crack deviation value is obtained through node deviation calculation, and then formed into a node crack effect index through iterative correction and spatial-temporal integration.

[0039] Spatial interpolation and time series synthesis operations are performed on the nodal crack effect index to form a continuous three-dimensional crack effect evaluation field. Crack effect evaluation results are generated by iterative adjustment of deviation correction and injection parameters.

[0040] In a second aspect, the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, it implements any step of the design method for CO2 high-pressure gas fracturing in high-gas coal roadways as described in the first aspect of the present invention.

[0041] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the design method for CO2 high-pressure gas fracturing in high-gas coal roadways as described in the first aspect of the present invention.

[0042] The beneficial effects of this invention are as follows: By cleaning, structuring, and dynamically modeling the environmental parameters of the coal seam, a digital twin coal seam model that can reflect the stress and fracture distribution characteristics of the coal seam in real time is generated, realizing the accurate reproduction and dynamic updating of the complex structure of the coal seam; then, by inputting high-pressure CO2 gas into the digital twin coal seam model for stress-fracture interaction simulation and iterative evolution prediction, a fracture evolution prediction dataset is generated, realizing the visualization and trend control of the fracture network formation process, thereby providing a reliable basis for injection parameter optimization and scheme verification, and ultimately achieving the beneficial effect of improving the accuracy and adaptability of coal roadway fracture design. Attached Figure Description

[0043] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Fig. 1 A flowchart of the design method for CO2 high-pressure gas fracturing in high-gas coal roadways.

[0045] Fig. 2 Flowchart for building a coal seam basic parameter dataset.

[0046] Fig. 3 Flowchart for generating digital twin coal seam models.

[0047] Fig. 4 A flowchart for generating a crack evolution prediction dataset. Detailed Implementation

[0048] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0049] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0050] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0051] Reference Figs. 1-4 As one embodiment of the present invention, this embodiment provides a design method for CO2 high-pressure gas fracturing in high-gas coal roadways, comprising the following steps:

[0052] S1. Collect coal seam environmental parameters of the target coal seam, perform preprocessing and comprehensive analysis, and summarize and organize them to form a coal seam basic parameter dataset.

[0053] S1.1 Cleaning, denoising, and smoothing the coal seam environmental parameters to obtain a standardized dataset.

[0054] Furthermore, the collected coal seam environmental parameters include raw information such as pressure, temperature, stress, pore structure, and gas content. The parameters are first subjected to outlier detection and missing value processing, corrected by elimination or interpolation to ensure the integrity and continuity of the parameters. Noise suppression is then applied, for example, using moving average smoothing or weighted average smoothing to reduce high-frequency interference generated during data collection, smoothing fluctuations while preserving key trends. The smoothed parameters are then standardized according to uniform units and dimensions, such as through linear normalization or mean-variance standardization, to ensure comparability and consistency among different types of coal seam parameters. Finally, the processed parameters are organized and summarized according to time series and spatial location to form a standardized dataset.

[0055] S1.2. The standardized dataset is processed through unified normalization, comprehensive analysis, and interactive parameter calculation to generate comprehensive coal seam basic parameter indicators.

[0056] Furthermore, the standardized dataset is uniformly normalized according to the dimensions and range of variation of each coal seam environmental parameter. For example, linear normalization is used to convert pressure, temperature, stress, pore structure, and gas content to a unified numerical range to ensure that the environmental parameters of each coal seam are compared and analyzed comprehensively on the same scale. The normalized coal seam environmental parameters are then comprehensively analyzed. Weighted average, principal component analysis, or statistical correlation analysis methods are used to evaluate the interrelationships and overall trends among the characteristic parameters of fracture propagation in each coal seam, and to extract the characteristic parameters of fracture propagation that have a significant impact on the propagation of coal seams. Based on the characteristic parameters of fracture propagation in coal seams, stress-pore coupling indices or gas-pressure response indices under different parameter combinations are calculated to reflect the coupling effect and potential impact between coal seam environmental parameters. The normalization results, comprehensive analysis results, and interactive parameter calculation results are integrated in a unified format to generate comprehensive coal seam basic parameter indices.

[0057] S1.3. Summarize, structure, and format the comprehensive coal seam basic parameter indicators to form a coal seam basic parameter dataset.

[0058] Furthermore, the comprehensive coal seam basic parameter indicators are classified and sorted according to spatial location, depth level, and time series order, divided into spatial regions (e.g., working face, roadway, unmined area), depth levels (e.g., shallow, middle, deep), and time stages (e.g., pre-injection, injection, expansion period, stabilization period). Each coal seam environmental parameter and its corresponding parameters are summarized using unified field naming, units, and dimensions to ensure that comprehensive coal seam basic parameter indicators from different sources and types can be directly compared and accessed. The summarized comprehensive coal seam basic parameter indicators are then structured, for example, using tables. The pressure, temperature, stress, pore structure, and gas content indices, along with the calculated interactive parameters, are arranged using gridding, matrixing, or multidimensional array methods, ensuring that each record contains complete spatial, temporal, and parameter information. After structuring, the comprehensive coal seam basic parameter indices are uniformly formatted, for example, by standardizing the storage of records according to a unified encoding, file format, and storage structure. This ensures that all comprehensive coal seam basic parameter indices can be continuously read, indexed, and processed in subsequent operations, resulting in a coal seam basic parameter dataset containing complete, standardized, and structured comprehensive coal seam basic parameter indices.

[0059] S2. The coal seam basic parameter dataset is structured, stress field and fracture distribution are analyzed, and virtual simulation and dynamic updates are performed to generate a digital twin coal seam model.

[0060] It should be noted that existing methods typically perform single static analysis on coal seam basic parameter datasets, obtaining coal seam characteristics through stress field calculations and fracture distribution predictions. The analysis process largely relies on fixed parameters or empirical formulas, and virtual simulations are mostly single simulations. Dynamic updates lack continuous time series processing, resulting in limited response to real-time changes in the coal seam environment.

[0061] This invention performs structured processing on the coal seam basic parameter dataset, combines stress field and fracture distribution analysis with virtual simulation, and forms a digital twin coal seam model through dynamic update processing, realizing continuous time series coal seam characteristic updates, enabling the model to reflect the changing trends and evolution processes of the coal seam environment under different working conditions.

[0062] S2.1 Map the coal seam basic parameter dataset to a three-dimensional grid, and generate structured grid data through classification, normalization and spatial interpolation.

[0063] Furthermore, the coal seam basic parameter dataset is mapped to a three-dimensional grid according to spatial coordinates and depth levels, ensuring that each grid cell contains comprehensive coal seam basic parameter indicators such as pressure, temperature, stress, pore structure, and gas content at the corresponding location. The comprehensive coal seam basic parameter indicators within the grid cell are normalized, for example through linear normalization or mean-variance standardization, so that different comprehensive coal seam basic parameter indicators can be calculated and compared at the same scale. For missing or sparsely distributed grid cells, spatial interpolation is performed, for example using nearest neighbor interpolation, bilinear interpolation, or cubic spline interpolation methods, to smoothly fill the comprehensive coal seam basic parameter indicators of surrounding grid cells into the target grid cell, so as to form a continuous three-dimensional parameter field. The generated structured grid data stores the comprehensive coal seam basic parameter indicators of each grid cell in a unified format, forming a continuous and analyzable three-dimensional spatial data structure.

[0064] S2.2 Analyze the stress field and potential crack distribution of the structured mesh data, and generate a preliminary digital twin model by combining virtual simulation calculation and dynamic update processing.

[0065] Furthermore, stress field calculations are performed on the comprehensive coal seam basic parameters, including pressure, temperature, stress, pore structure, and gas content, for each grid cell in the structured grid data. The stress distribution and stress concentration areas within the grid cells are determined using finite element analysis or mechanical analysis methods. Simultaneously, the locations and propagation directions of potential fractures are predicted and analyzed, forming preliminary information on potential fracture distribution. After the stress field and potential fracture distribution analysis are completed, the results are combined with virtual simulation calculations. The simulation simulates changes in coal seam stress and fracture evolution, evaluating the coal seam's response under different operating conditions. The structured grid data is dynamically updated based on the simulation results, for example, by adjusting the stress values ​​and potential fracture indices of the grid cells to reflect the impact of time series and operating condition changes. Finally, the stress field, potential fracture distribution, and dynamic update results are integrated to form a preliminary digital twin model.

[0066] S2.3. The preliminary digital twin model is updated through time series simulation, dynamic correction and real-time data feedback to generate a digital twin coal seam model.

[0067] Furthermore, each grid cell in the preliminary digital twin model is simulated according to a time series. The changes of the comprehensive coal seam basic parameters, including pressure, temperature, stress, pore structure, and gas content under historical working conditions, are gradually deduced at various points in time. These time-varying trajectories are then continuously connected to form a time series response. During the time series simulation, the stress field, potential fracture distribution, and related parameters in the preliminary digital twin model are dynamically corrected based on the simulation results, and the parameters within the grid cells are adjusted to reflect continuous changing trends. Simultaneously, the comprehensive coal seam basic parameters, including pressure, temperature, stress, pore structure, and gas content obtained from real-time field monitoring, are input into the update process to provide real-time data feedback correction to the preliminary digital twin model, ensuring that the parameter values ​​of the grid cells are consistent with the actual coal seam working conditions. After time series simulation, dynamic correction, and real-time data feedback updates, the stress field, potential fracture distribution, and comprehensive coal seam basic parameters of the grid cells are integrated to form a continuous, dynamic, and analyzable digital twin coal seam model.

[0068] It should also be noted that the construction process of the digital twin coal seam model is based on a coal seam basic parameter dataset. Comprehensive coal seam basic parameter indicators are mapped onto a three-dimensional spatial grid to form structured grid data. Each grid cell contains comprehensive coal seam basic parameter indicators such as pressure, temperature, stress, pore structure, and gas content. Stress field analysis and potential fracture distribution analysis are performed on the structured grid data. The stress distribution and potential fracture locations within the grid cells are determined through finite element calculations or mechanical analysis. Simultaneously, virtual simulation calculations are combined to evaluate the coal seam response under different working conditions, generating a preliminary digital twin model. The comprehensive coal seam basic parameter indicators of pressure, temperature, stress, pore structure, and gas content in each grid cell of the preliminary digital twin coal seam model are then sequentially analyzed according to time sequence, depicting their changes under historical working conditions. By advancing the time steps sequentially, the parameter state at each time step is compared with the previous time step. The states at each time step are correlated to form a continuous time-series response. During the simulation, finite element analysis or physical constraints can be used to extrapolate the changing trends of the stress field and potential crack distribution of the grid cells over time, while updating the parameter values ​​at specific time steps to reflect the dynamic evolution process. The initial digital twin model is dynamically corrected during the time-series simulation, and the stress field and potential crack distribution of the grid cells are adjusted based on the simulation results. The comprehensive coal seam basic parameters, including pressure, temperature, stress, pore structure, and gas content obtained from real-time on-site monitoring, are input into the update process. The grid cell parameter values ​​are corrected through real-time data feedback to ensure that the initial digital twin model is consistent with the actual coal seam conditions. Finally, the results of the time-series simulation, dynamic correction, and real-time data feedback update are integrated to form a continuous and analyzable digital twin coal seam model.

[0069] The training process of the digital twin coal seam model involves mapping the comprehensive coal seam basic parameters, including pressure, temperature, stress, pore structure, and gas content, obtained from historical working conditions and real-time on-site monitoring, onto a three-dimensional spatial grid to form structured grid data. For each grid cell, the changes in the comprehensive coal seam basic parameters are analyzed according to the time series to form a time series response. At each time step, the grid cell response is compared with the observed value. By calculating the node deviation, the stress field, potential fracture distribution, and comprehensive coal seam basic parameters of the grid cell are iteratively corrected, allowing the digital twin coal seam model to gradually converge. Finally, the grid responses and correction results from all time steps are integrated to generate a continuous, dynamic, and analyzable digital twin coal seam model.

[0070] S3. Input high-pressure CO2 gas into the digital twin coal seam model to perform nodal response simulation and fracture propagation calculation, and obtain a fracture evolution prediction dataset.

[0071] It should be noted that existing methods typically perform single-point or single-hole static simulations of CO2 (carbon dioxide) high-pressure gas injection. Crack propagation calculations often rely on empirical formulas or simple finite element analysis, and nodal response analysis is mostly performed in isolation. It is difficult to achieve dynamic interaction among multiple holes and nodes, and the crack evolution results lack continuous time series and spatial integration.

[0072] This invention inputs high-pressure CO2 gas into a digital twin coal seam model, generates a continuous time series of fracture evolution prediction datasets through node response simulation and fracture propagation calculation, realizes node-level multidimensional interactive analysis, and combines spatial interpolation and topological correlation operations to enable fracture evolution information to be systematically reflected in three-dimensional space.

[0073] S3.1 Input high-pressure CO2 gas into the digital twin coal seam model, and through stress-fracture interaction analysis and dynamic simulation, form nodal fracture potential information, and perform spatial interpolation and topological correlation analysis to generate a continuous three-dimensional fracture propagation field.

[0074] Furthermore, high-pressure CO2 gas is injected into the digital twin coal seam model. Based on the pressure, temperature, stress, pore structure, and gas content of each grid cell, stress changes and fracture responses are calculated using comprehensive coal seam fundamental parameters. This generates nodal fracture potential information (e.g., fracture length potential, fracture connectivity index, fracture opening tendency, and fracture propagation direction vector), reflecting the probability of fractures occurring at each node under current operating conditions and the direction of fracture development. Spatial interpolation is then performed on the nodal fracture potential information, using methods such as nearest neighbor interpolation, three-dimensional bilinear interpolation, or cubic spline interpolation, to smoothly extend the fracture potential information of discrete nodes to the entire grid. A continuous three-dimensional crack distribution field is formed in the grid space. At the same time, topological correlation analysis is performed on the crack potential information after spatial interpolation. By judging the geometric proximity of nodes in the grid, comparing the similarity of crack potential indicators (such as opening tendency and propagation direction), combining the consistency of stress field gradient and crack direction, and using adjacency matrix or connected graph methods to analyze potential continuous paths, it is determined whether there is crack connectivity and mutual influence between nodes, and a crack network structure is established to ensure that the continuous three-dimensional crack propagation field maintains topological consistency and continuity in space. The interpolation processing and topological correlation analysis results are integrated to form a continuous three-dimensional crack propagation field.

[0075] S3.2. Perform dynamic simulation and time step iteration on the continuous three-dimensional crack propagation field to generate time series crack evolution data.

[0076] Furthermore, the continuous three-dimensional fracture propagation field is divided into time steps. Dynamic simulations are performed on the pressure, temperature, stress, pore structure, and gas content of each time step grid cell, integrating basic coal seam parameters and nodal fracture potential information. This simulates the evolution of fractures over time during CO2 high-pressure gas injection. In each time step, the fracture propagation amount, fracture connectivity, and fracture network evolution of the grid cell are calculated. Based on the fracture distribution results of the previous time step, time step iterations are performed to update the fracture length, direction, and potential information of each grid cell. Through continuous time step iterations and dynamic simulations, the fracture evolution results of each time step are sequentially integrated to generate complete time-series fracture evolution data.

[0077] S3.3 Perform real-time monitoring, feedback correction, and iterative updates on the time-series crack evolution data to form a crack evolution prediction dataset.

[0078] Furthermore, comprehensive coal seam basic parameters, including pressure, temperature, stress, pore structure, and gas content, as well as fracture observation information, are collected from real-time on-site monitoring. The observed values ​​are then compared point-by-point with the predicted values ​​of the corresponding time steps and grid cells in the time-series fracture evolution data to calculate the observation residuals. When calculating node corrections based on these residuals, the observation reliability of each node is determined by evaluating the measurement accuracy of the monitoring equipment, the stability of historical data, and the impact of environmental interference. Simultaneously, the spatial neighborhood weights are determined by analyzing the geometric distance between grid cells, spatial topological relationships, and the similarity of comprehensive coal seam basic parameters. The observation reliability and spatial neighborhood weights are then comprehensively applied to the node correction calculation to reflect the credibility of the observation data and the influence of adjacent nodes. The expression is as follows:

[0079] ;

[0080] in, Indicates the node correction amount. This represents the correction factor. Represents the observation residual. Represents the spatial neighborhood weight. Indicates the weight of observation reliability;

[0081] The node correction is used to correct the crack length, crack connectivity, and node crack potential information of the corresponding grid cells in the time series crack evolution data. After each node correction, a short-term dynamic simulation is performed on the corrected time series crack evolution data to generate an updated time step prediction, and the observation residuals are recalculated to evaluate the correction effect. Then, the node parameters and time series predictions are iteratively updated by residual correction and short-term dynamic simulation until the observation residuals reach the iteration limit. The time series crack evolution data that have passed the residual correction and been verified by short-term dynamic simulation are integrated and uniformly formatted according to time series and spatial location to form a crack evolution prediction dataset.

[0082] S4. Analyze the nodal crack characteristics and match the working conditions on the crack evolution prediction dataset to obtain an optimized injection scheme.

[0083] S4.1 Perform node crack characteristic analysis and working condition matching calculation on the crack evolution prediction dataset to obtain node injection optimization data, and perform crack-working condition coupling calculation to generate injection optimization index.

[0084] Furthermore, the geometric features of the fractures, fracture connectivity indices, fracture length growth rate, fracture potential information, local stress response, and pore structure response of each node are extracted from the fracture evolution prediction dataset. During the node fracture characteristic analysis stage, statistical feature calculations, time-series feature analysis, and sensitivity analysis are performed on the extracted results. Statistical feature calculations reveal the overall patterns of fracture characteristics through mean, variance, extreme values, and distribution patterns. Time-series feature analysis identifies the dynamic characteristics of fracture evolution by extracting trends, periodic features, and abrupt change points. Sensitivity analysis compares the changes in fracture response before and after disturbances, using operating condition variables such as perturbation pressure, stress, and porosity, to determine key influencing parameters. The output of the node fracture characteristic analysis is then compared with current operating condition information (e.g., injection pressure level, injection flow rate pattern, coal seam temperature). The system performs condition matching calculations based on the following parameters: stress conditions, original stress field distribution, gas concentration, and pore structure variation characteristics. This includes constructing condition feature vectors, similarity measurement, and weight allocation to evaluate the adaptability and response of each node under different injection conditions. Based on the results of the condition matching calculations, a fracture-condition coupling calculation is performed on each node. This calculation employs a fluid-rock coupling evaluation method, pore pressure difference transmission analysis, and energy balance calculations to obtain the pressure transmission efficiency, fracture opening probability, and fracture connectivity change rate of each node during injection. The node-level performance indicators output from the fracture-condition coupling calculations are normalized, weighted, and spatially sorted. Node injection optimization data is then generated by summarizing the data according to injection targets and priorities, and finally, injection optimization indicators are formed from the summarized node injection optimization data.

[0085] S4.2. Injection optimization indicators are spatially integrated and sorted to form an injection optimization scoring field, and dynamic matching calculations are performed on the pressure, flow rate and duration at each injection hole location to generate the optimal injection parameter combination.

[0086] Furthermore, based on the geographic coordinates and depth information of each node within the injection optimization index, the injection optimization index is mapped to the injection borehole location. The quantities reflecting injection priority, pressure sensitivity, flow capacity, and duration requirements within the injection optimization index are normalized and weighted to calculate the initial local score for each injection borehole location. Spatial smoothing and interpolation are performed on the initial local scores, and existing mathematical methods such as inverse distance weighting, Kriging interpolation, or multinomial regression are used to generate a continuous injection optimization score field. Subsequently, dynamic matching is performed on the injection optimization score field to calculate the pressure, flow rate, and duration for each injection borehole location, establishing injection borehole constraints and the injection pressure-flow rate transmission relationship, and identifying the mutual influence and connection of crack potential between adjacent nodes. In general, by judging the geometric proximity of nodes in the mesh, comparing crack potential indicators (such as crack opening tendency, propagation direction, crack initiation trend), combining the consistency between stress field gradient and crack direction, and using adjacency matrix or connected graph methods to analyze potential continuous paths, it is determined whether there is crack connectivity and mutual influence between nodes. Existing optimization algorithms such as iterative solution or constraint optimization and heuristic search are used to solve for the pressure, flow rate and duration combination that satisfies the injection hole constraints and takes into account the objective function of the injection optimization scoring field. The obtained injection hole parameters are spatially sorted and their feasibility is verified. The injection hole parameters that satisfy the injection hole constraints and score first in the injection optimization scoring field are combined according to the priority sequence to form the optimal injection parameter combination.

[0087] S4.3. Organize and summarize the optimal injection parameter combinations according to spatial location and time order to obtain the optimized injection scheme.

[0088] Furthermore, the optimal injection parameter combination is mapped to the specific injection hole location according to the injection hole location coordinates, and the injection tasks are time-ordered based on the injection start time and injection duration in the injection parameter combination. In the spatial dimension, the injection hole locations are spatially grouped based on the Euclidean distance between them to identify potential time or pressure conflicts between adjacent injection tasks. For each spatial group, the injection hole locations are prioritized according to the injection optimization scoring field, and a feasibility check is performed in conjunction with the injection hole constraints and the injection pressure, injection flow rate, and injection duration information in the injection parameter combination. When a conflict is found, the injection start time is adjusted and the injection duration is reallocated. Conflict resolution is carried out by intermittently extending the injection sequence. Conflict resolution adopts iterative optimization or heuristic scheduling methods until the injection hole constraints and safety constraints are met. After completing the sorting and conflict resolution within the spatial group, the injection timing and corresponding injection parameters (injection pressure, injection flow rate, injection duration) of each injection hole location are merged into an injection time table. The injection plan table is then generated by organizing the data according to spatial location and time order. The injection plan table records the injection hole location coordinates, injection start time, injection duration, injection pressure, injection flow rate, priority and feasibility verification results of each injection task according to a unified coding and field format. All injection plan table records are summarized and sorted to form an optimized injection scheme.

[0089] S5. Perform node feasibility verification, risk assessment and parameter iterative optimization on the optimized injection scheme, and generate an injection scheme verification dataset.

[0090] S5.1 Perform feasibility calculations and risk analysis on the node injection parameters of the optimized injection scheme to obtain the injection verification index dataset.

[0091] Furthermore, the injection schedule in the optimized injection scheme is mapped line by line to the structured grid data of the digital twin coal seam model to determine the grid cell and time step corresponding to each injection hole location in the injection schedule. For each injection hole location, transient seepage calculations are performed based on the injection pressure, injection flow rate, and injection duration in the injection schedule. The spatiotemporal distribution of pressure propagation after injection is solved based on Darcy's law and pore pressure difference transmission analysis, and the surrounding pore pressure difference, seepage velocity, and expected pressure influence radius are calculated accordingly. The pore pressure change obtained from the seepage calculation is superimposed on the original effective stress state of the grid cell. The triaxial principal stress is corrected according to the effective stress principle, and the stress path is constructed using the stress disturbance caused by the pore pressure increment. The results are then compared with mechanical stability criteria (e.g., Mohr-Coulomb). The stability margin is calculated based on the relationship between shear stress and shear strength in the strength criterion. If the corrected stress state point is close to or exceeds the failure envelope, it is determined that there is a risk of crack reactivation or local instability. High-risk areas can be identified through spatial distribution maps. Mechanical stability criteria (such as the Mohr-Coulomb strength criterion) are used to assess the possibility of crack reactivation and local instability risk caused by injection. At the wellbore integrity level, the wellbore feasibility of the injection parameters is verified by comparing the injection pressure with the wellbore bearing capacity at the injection hole location. Neighborhood interference analysis is performed by combining the spatial spacing between injection hole locations and the injection optimization score field to assess the probability of cross-influence between injection holes. For leakage and gas migration risks, connectivity analysis is carried out by identifying low-permeability channels in the pressure propagation range and structured grid data, and leakage probability and risk level are estimated (for example, when the pressure propagation front intersects with a low-permeability channel or a known weak surface, the pressure gradient and seepage flow at the intersection are calculated. If it exceeds the critical bearing capacity of the channel, it is determined to be a high leakage probability, and the risk level is classified according to the extent of the exceedance). The risk level is adjusted by combining gas concentration and channel connectivity to obtain a leakage risk assessment result that better reflects the actual working conditions. The stress change amplitude caused by the injection process, including shear stress increment and normal stress disturbance, is extracted from the seepage and pore pressure-stress coupling calculation, and time series analysis is performed to identify stress concentration and rapidly changing areas. The stress change amplitude is compared with the empirical statistical relationship between the frequency and magnitude distribution of historical earthquake events. By judging whether the stress disturbance exceeds the known event triggering threshold, the probability of events of different intensities is estimated. The probability results are spatially corrected by combining the fracture structure characteristics and fracture connectivity of the injection area to obtain a graded induced earthquake risk score. The injection feasibility index, pressure propagation index, fracture reactivation probability, leakage probability, induced earthquake risk score, wellbore integrity identifier, and neighborhood interference factor at the injection hole location are summarized in a unified field format to form an injection verification index dataset.

[0092] S5.2 Perform node-level calculations and neighborhood coupling analysis on the injected verification index dataset to form node verification indices.

[0093] Furthermore, based on the injection well location and time step identified in each record of the injection verification index dataset, the corresponding grid cell is located in the structured grid data. Node-level calculations are performed for each grid cell, including pore pressure-stress coupling response, local seepage velocity, surrounding pore pressure differential, fracture reactivation probability, wellbore integrity indicator, leakage probability, and induced earthquake risk score. Subsequently, based on the spatial neighborhood definition, neighborhood factors are extracted from the injection verification index dataset, and neighborhood coupling analysis is performed. Pressure transmission, pore pressure differential superposition, and fracture reactivation between adjacent grid cells are quantified using weighting functions (e.g., inverse distance weighting or topological adjacency weighting). The purpose of neighborhood coupling analysis, which considers the influence of fracture connectivity, is to avoid isolated misjudgments in the calculation results of a single grid cell. By taking into account the pressure transmission between adjacent cells, the superposition of pore pressure difference, and the fracture connectivity effect, it more realistically reflects the spatial linkage characteristics under the injection action. The analysis results in the local response value of each injection hole location after correction under the influence of the neighborhood, and quantifies the neighborhood interference factor. After the neighborhood coupling analysis, the node-level indicators are normalized and weighted, and the timing consistency is checked and the short-term response is simulated in combination with the injection start time and injection duration in the injection plan table. The node verification indicators corresponding to each injection hole location are output.

[0094] S5.3. Spatial integration and sorting of node verification indicators are performed to generate a verification score field. Iterative adjustments and parameter optimization operations are then performed to generate an injection scheme verification dataset.

[0095] Furthermore, based on the distribution of each injection hole location in three-dimensional space within the structured grid data, spatial integration operations are performed on the node verification indicators. This includes weighted averaging or cumulative calculation of multiple node verification indicators within the same spatial unit to quantify local injection feasibility, pressure propagation capability, fracture reactivation probability, leakage probability, induced seismic risk score, wellbore integrity identifier, and neighborhood interference factor. Subsequently, the node verification indicators are sorted according to spatial location and risk priority to generate a verification score field, which reflects the comprehensive injection risk and feasibility distribution of the entire injection area. After the verification score field is generated, the node verification indicators are iteratively adjusted based on the injection parameter range, injection sequence, and historical node responses. By gradually correcting parameters such as injection pressure, flow rate, and duration, the local response of the nodes and the neighborhood coupling effect are optimized, and multiple rounds of iterative calculations are performed to finally generate the injection scheme verification dataset.

[0096] S6. Perform multi-dimensional evaluation and comprehensive judgment on the injection scheme verification dataset to obtain the crack effect evaluation results.

[0097] S6.1 Standardize and preprocess the injection scheme verification dataset to obtain multidimensional input data, and perform node-level nonlinear interaction analysis and mapping operations to generate node mapping results.

[0098] Furthermore, the injection pressure, flow rate, duration, and neighborhood coupling correction values ​​for each node are standardized and preprocessed, including unifying dimensions, normalization, and outlier removal, to generate multidimensional input data. After the multidimensional input data is generated, nonlinear interaction analysis is performed on each node, considering the pressure interaction between nodes, crack induction probability, neighborhood interference, and local stress influence. The node input parameters are mapped to the comprehensive injection response index through mapping operations. The mapping operation refers to using mathematical functions or matrix methods to map the standardized multidimensional input vector to the node-level comprehensive injection response index. At the same time, the correspondence between input and output is established by combining nonlinear interaction and neighborhood coupling correction values ​​to obtain the mapping relationship value of each node, and finally form the node mapping result.

[0099] S6.2 Perform node-level comprehensive judgment on the node mapping results, generate node crack effects, and perform spatial integration and time series synthesis operations to generate a continuous three-dimensional crack expansion effect field.

[0100] Furthermore, a comprehensive assessment is performed on each node, including evaluating the correspondence between node injection parameters and crack response indices, calculating crack propagation potential, considering neighborhood interactions and local stress distribution, and generating node crack effects. After the node crack effects are generated, the crack information of each node is integrated according to its spatial location to form a spatially continuous structure. At the same time, it is combined with time series data for synthesis, considering the evolution and propagation trend of cracks with injection time, and finally obtaining a continuous three-dimensional crack propagation effect field.

[0101] S6.3. The continuous three-dimensional crack propagation effect field is combined with the crack evolution prediction dataset to obtain the crack deviation value through node deviation calculation. The crack effect index is formed by iterative correction and spatial-temporal integration.

[0102] Furthermore, the actual crack geometry features (such as crack length, opening amount, and connectivity direction) of each node in the continuous three-dimensional crack propagation effect field are compared point by point with the predicted values ​​of the corresponding nodes in the crack evolution prediction dataset, and the node deviation value is calculated. The expression is as follows:

[0103] ;

[0104] in, Represents a node The crack deviation value, This represents the deviation adjustment factor. Represents a node The continuous three-dimensional crack propagation effect field value, Represents a node The crack evolution prediction dataset values, Represents a node Spatial integration weight, Represents a node Time series weights, Indicates the node index;

[0105] The node deviation value can be quantified by methods such as vector difference, relative rate of change, or geometric distance to differentiate between actual and predicted cracks. Based on the calculated node deviation value, combined with spatial neighborhood relationships and time series characteristics, the crack response characteristics of each node are adjusted. For example, the deviation is distributed to adjacent nodes according to neighborhood weights through a weighted correction method, and the predicted values ​​of the time steps are updated iteratively to make the node crack characteristics gradually converge to the actual observed trend, thereby generating a node crack effect index after spatial-temporal integration optimization.

[0106] S6.4 Perform spatial interpolation and time series synthesis operations on the nodal crack effect index to form a continuous three-dimensional crack effect evaluation field, and generate crack effect evaluation results through deviation correction and iterative adjustment of injection parameters.

[0107] Furthermore, spatial interpolation is performed on the crack effects at the nodes in spatial location, and the crack effects at each time step are integrated through time series synthesis to form a continuous three-dimensional crack effect evaluation field, which is used to describe the continuous distribution characteristics of cracks in space and time. Based on the deviation between the continuous three-dimensional crack effect evaluation field and the expected crack development trend, a deviation correction operation is performed, and the response characteristics of each node are optimized by iterative adjustment of the injected parameters to generate crack effect evaluation results.

[0108] This embodiment also provides a computer device applicable to the design method of CO2 high-pressure gas fracturing in high-gas coal roadways, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the design method of CO2 high-pressure gas fracturing in high-gas coal roadways as proposed in the above embodiment.

[0109] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0110] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the design method for CO2 high-pressure gas fracturing in high-gas coal roadways as proposed in the above embodiments. The 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 Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0111] In summary, this invention achieves accurate reproduction and dynamic updating of complex coal seam structures by: cleaning, structuring, and dynamically modeling coal seam environmental parameters to generate a digital twin coal seam model that reflects the stress and fracture distribution characteristics of the coal seam in real time; and by inputting high-pressure CO2 gas into the digital twin coal seam model for stress-fracture interaction simulation and iterative evolution prediction, generating a fracture evolution prediction dataset, thereby realizing visualization and trend control of the fracture network formation process. This provides a reliable basis for parameter optimization and scheme verification, ultimately achieving the beneficial effect of improving the accuracy and adaptability of coal roadway fracturing design.

[0112] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A design method for CO2 high-pressure gas fracturing in high-gas coal roadways, characterized in that: include, The environmental parameters of the target coal seam are collected, preprocessed, and comprehensively analyzed, and then summarized and organized to form a basic parameter dataset of the coal seam. The coal seam basic parameter dataset is structured, stress field and fracture distribution are analyzed, and virtual simulation and dynamic updates are performed to generate a digital twin coal seam model. High-pressure CO2 gas is input into a digital twin coal seam model to simulate nodal response and calculate fracture propagation, resulting in a fracture evolution prediction dataset. The nodal crack characteristics of the crack evolution prediction dataset are analyzed and the working conditions are matched to obtain an optimized injection scheme. The node feasibility verification, risk assessment and parameter iterative optimization of the optimized injection scheme are carried out, and an injection scheme verification dataset is generated. A multi-dimensional evaluation and comprehensive judgment were performed on the injection scheme verification dataset to obtain the crack effect assessment results.

2. The design method for CO2 high-pressure gas fracturing in high-gas coal roadways as described in claim 1, characterized in that: The collected coal seam environmental parameters are preprocessed and comprehensively analyzed, and then summarized to form a coal seam basic parameter dataset. The specific steps are as follows: The coal seam environmental parameters are cleaned, denoised, and smoothed to obtain a standardized dataset; The standardized dataset is processed through unified normalization, comprehensive analysis, and interactive parameter calculation to generate comprehensive coal seam basic parameter indicators. The comprehensive coal seam basic parameter indicators are summarized, structured, and uniformly formatted to form a coal seam basic parameter dataset.

3. The design method for CO2 high-pressure gas fracturing in high-gas coal roadways as described in claim 2, characterized in that: The specific steps for structuring the coal seam basic parameter dataset, analyzing stress field and fracture distribution, performing virtual simulation and dynamic updates, and generating a digital twin coal seam model are as follows. The coal seam basic parameter dataset is mapped to a three-dimensional grid, and structured grid data is generated through classification, normalization and spatial interpolation. Stress field and potential crack distribution analysis were performed on structured mesh data, and a preliminary digital twin model was generated by combining virtual simulation calculations and dynamic update processing. The initial digital twin model is updated through time series simulation, dynamic correction, and real-time data feedback to generate a digital twin coal seam model.

4. The design method for CO2 high-pressure gas fracturing in high-gas coal roadways as described in claim 3, characterized in that: The process of inputting high-pressure CO2 gas into a digital twin coal seam model to simulate nodal responses and calculate fracture propagation yields a fracture evolution prediction dataset. The specific steps are as follows: High-pressure CO2 gas is input into the digital twin coal seam model. Through stress-fracture interaction analysis and dynamic simulation, nodal fracture potential information is generated. Spatial interpolation and topological correlation analysis are then performed to generate a continuous three-dimensional fracture propagation field. Dynamic simulation and time-step iteration of the continuous three-dimensional crack propagation field are performed to generate time-series crack evolution data; Real-time monitoring, feedback correction, and iterative updates are performed on time-series crack evolution data to form a crack evolution prediction dataset.

5. The design method for CO2 high-pressure gas fracturing in high-gas coal roadways as described in claim 4, characterized in that: The optimization injection scheme is obtained by analyzing the nodal crack characteristics and matching the working conditions of the crack evolution prediction dataset. The specific steps are as follows. The nodal crack characteristics of the crack evolution prediction dataset are analyzed and the working condition matching calculation is performed to obtain the nodal injection optimization data. Crack-working condition coupling calculation is then performed to generate injection optimization index. The injection optimization indicators are spatially integrated and sorted to form an injection optimization scoring field, and the pressure, flow rate and duration of each injection hole location are dynamically matched and calculated to generate the optimal combination of injection parameters. The optimal injection parameter combinations are organized and summarized according to spatial location and time order to obtain the optimized injection scheme.

6. The design method for CO2 high-pressure gas fracturing in high-gas coal roadways as described in claim 5, characterized in that: The steps for performing node feasibility verification, risk assessment, and parameter iterative optimization on the optimized injection scheme, and generating an injection scheme verification dataset, are as follows: Feasibility calculations and risk analyses of node injection parameters were performed on the optimized injection scheme to obtain an injection verification index dataset. Node-level computation and neighborhood coupling analysis are performed on the injected verification index dataset to form node verification indices; Spatial integration and sorting of node verification indicators are performed to generate a verification score field. Iterative adjustments and parameter optimizations are then performed to generate an injection scheme verification dataset.

7. The design method for CO2 high-pressure gas fracturing in high-gas coal roadways as described in claim 6, characterized in that: The process of performing multi-dimensional evaluation and comprehensive judgment on the injection scheme verification dataset to obtain the crack effect evaluation result is as follows: The injection scheme verification dataset is standardized and preprocessed to obtain multidimensional input data, and node-level nonlinear interaction analysis and mapping operations are performed to generate node mapping results. The node mapping results are comprehensively judged at the node level to generate node crack effects. Spatial integration and temporal series synthesis operations are then performed to generate a continuous three-dimensional crack propagation effect field. Deviation correction and iterative updates are performed on the continuous three-dimensional crack propagation effect field to obtain crack effect evaluation results.

8. The design method for CO2 high-pressure gas fracturing in high-gas coal roadways as described in claim 7, characterized in that: The specific steps for performing deviation correction and iterative updates on the continuous three-dimensional crack propagation effect field to obtain the crack effect evaluation result are as follows. By combining the continuous three-dimensional crack propagation effect field with the crack evolution prediction dataset, the crack deviation value is obtained through node deviation calculation, and then formed into a node crack effect index through iterative correction and spatial-temporal integration. Spatial interpolation and time series synthesis operations are performed on the nodal crack effect index to form a continuous three-dimensional crack effect evaluation field. Crack effect evaluation results are generated by iterative adjustment of deviation correction and injection parameters.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the design method for CO2 high-pressure gas fracturing in high-gas coal roadways as described in any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the design method for CO2 high-pressure gas fracturing in high-gas coal roadways as described in any one of claims 1 to 8.

Images