Coal seam outburst risk multi-factor fusion quantitative region prediction method

Abstract

The invention provides a coal seam outburst risk multi-factor fusion quantitative area prediction method, and belongs to the field of coal and gas outburst area prediction.The method comprises the steps that the gas occurrence, the geological structure, the coal seam burial depth, the soft layering and the mining condition of a target coal seam are obtained, and basic information of the target coal seam is obtained; according to the basic information of the target coal seam, in combination with Rhino and FLAC3D software, simulating to obtain spatial distribution of vertical stress and horizontal stress of the target coal seam; obtaining spatial distribution of coal seam parameter values of the target coal seam based on a Kriging and random forest spatial combination interpolation method; based on the vertical stress, the horizontal stress, the gas pressure and the gas content of the target coal seam, the firmness coefficient of the coal and the initial gas diffusion speed of the coal, the outburst danger level of each area of the target coal seam is obtained in combination with the extension matter element model, and the method is of great significance to fine management and accurate prevention and control of area dangers of outburst disasters.

Description

Technical Field

[0001] The present invention relates to the field of regional prediction of coal and gas outbursts, and in particular to a quantitative regional prediction method for coal seam outburst hazards by integrating multiple factors. Background Art

[0002] As mining depth and intensity increase, coal seam gas pressure and ground stress continue to rise, making outburst disasters increasingly severe. Accurate prediction of outburst risk areas is fundamental to zoning management and effective early warning of outburst disasters. This accurate prediction of outburst risk allows for the implementation of differentiated outburst prevention measures for different outburst risk areas, reducing blindness in disaster prevention efforts, saving on outburst prevention project costs, and ensuring safe and efficient mine production.

[0003] Coal and gas outbursts are the result of the combined effects of geostress, gas, and coal properties. Studying the distribution characteristics of stress, gas, and coal property indicators is key to predicting areas at risk of outbursts. Currently, these indicators are primarily derived through spatial interpolation of discrete, manually measured data. This interpolation method is highly dependent on the measurement point data, resulting in poor precision and accuracy for the entire region, especially in areas far from the measurement points. Furthermore, current methods for predicting areas at risk of outbursts primarily rely on qualitative or single indicators and their critical values, lacking a quantitative, comprehensive indicator of outburst risk. Summary of the Invention

[0004] To address the above issues, the present invention provides a quantitative regional prediction method for coal seam outburst hazard by integrating multiple factors. This method uses machine learning to capture the complex nonlinear relationships among factors influencing indicators and improves the interpolation accuracy of indicator data. This method is of great significance for achieving precise regional risk management and prevention of coal seam outburst disasters. Specifically, it includes:

[0005] A quantitative regional prediction method for coal seam outburst hazard based on multi-factor integration, including:

[0006] Obtain the gas occurrence, geological structure, coal seam burial depth, soft layer and mining conditions of the target coal seam to obtain basic information of the target coal seam;

[0007] Based on the basic information of the target coal seam, combined with Rhino and FLAC 3D Software, simulating and obtaining the spatial distribution of vertical stress and horizontal stress of the target coal seam;

[0008] Based on the spatial combined interpolation method of Kriging and random forest, the spatial distribution of the coal seam parameter values of the target coal seam is obtained, and the coal seam parameter values include gas pressure, gas content, coal solidity coefficient f value and coal gas release initial velocity ΔP;

[0009] Based on the vertical stress, horizontal stress, gas pressure, gas content, coal strength coefficient f value and coal gas release initial velocity ΔP of the target coal seam, combined with the extension matter-element model, the outburst hazard level of each area of the target coal seam is obtained.

[0010] Optionally, according to the basic information of the target coal seam, Rhino and FLAC 3D Software, simulating the spatial distribution of vertical stress and horizontal stress of the target coal seam includes:

[0011] Based on the surface contours, coal seam depth, geological structure and mining conditions, the surface contours and coal seam contours were processed into a polysurface format using Rhino software to obtain a large numerical calculation model for coal seam stress that includes the main geological structures.

[0012] The surface contour lines, coal seam depth and elastic modulus of the coal seam are brought into Rhino software to obtain a polyhedron, and the polyhedron is brought into the large model of coal seam stress numerical calculation to obtain a FLAC 3D Large model for numerical calculation of coal seam stress based on unit cell grid;

[0013] Assign the model to the FLAC according to the coal rock mechanical parameters in the mine geological report 3D In the large numerical calculation model of coal seam stress of the unit grid, the spatial distribution of vertical stress and horizontal stress of the target coal seam is obtained.

[0014] Optionally, the spatial combined interpolation method based on Kriging and random forest to obtain the coal seam parameter value of the target coal seam includes:

[0015] S31, based on Kriging interpolation and residual calculation, obtaining the coal seam parameter values of the measuring points in the target coal seam and the corresponding training set residuals;

[0016] S32, calculating the factors influencing the prediction indicators;

[0017] S33. Based on a random forest regression model, the coal seam parameter value of the target coal seam is corrected according to the indicator influencing factors to obtain a corrected interpolation result;

[0018] S34, evaluating whether the interpolation result after correction based on the random forest regression model meets the preset requirements. If so, saving the random forest regression model and executing S35; if not, retraining the random forest regression model and returning to step S33;

[0019] S35: Correct the interpolation in the area to be interpolated.

[0020] Optionally, the step of obtaining the coal seam parameter values of the measuring points in the target coal seam and the corresponding training set residuals based on Kriging interpolation and residual calculation in S31 includes:

[0021] S311. Extracting gas pressure, gas content, coal solidity coefficient f, coal gas release initial velocity ΔP, and measuring point locations from the coal seam gas geological map based on Python to form an indicator data set S = (X, Y, Z) corresponding to the measuring point locations and the measured data.

[0022] S312: Divide the index data set into interpolation data sets S according to the ratio of 2:2:1. k =(X k ,Y k ,Z k ), training set S t =(X t ,Y t ,Z t ) and the test set S c =(X c ,Y c ,Z c );

[0023] S313, based on the interpolation data set S k =(X k ,Y k ,Z k ), the measurement point position P of the training set t =(X t ,Y t ) and the test point position P of the test set c =(X c ,Y c ), perform Kriging spatial interpolation on the measurement point positions of the training set and the test set to obtain the Kriging interpolation result Z' of the training set t And the Kriging interpolation result Z' of the test set c ;

[0024] S314, according to the measurement point data Z of the training set t And the Kriging interpolation result Z' t , and get the corresponding training set residual r t =Z t -Z' t .

[0025] Optionally, the calculation of the influencing factors of the prediction index in S32 includes:

[0026] The fault density is defined as follows:

[0027]

[0028] Among them, E fd is the fault density in the unit area; H is the fault drop, in m; L is the fault strike length, in m; α is the fault dip; β is the angle between the principal stress and the fault strike; S is the unit area, in m 2 ; n is the number of faults in the unit area; the unit area is a square area with a side length of a centered at the measuring point;

[0029] The degree of wrinkle deformation is defined as formula (2):

[0030]

[0031] Among them, F d is the degree of wrinkle deformation in the unit area; h is the elevation distance between adjacent contour lines, in meters; k is the actual length of the contour line with the largest curvature in the unit area, in meters; l is the secant length of the contour line with the largest curvature, in meters; d is the distance between the two contour lines with the largest curvature, in meters;

[0032] The coefficient of variation of coal thickness is defined as formula (3):

[0033]

[0034] Wherein, γ is the coefficient of variation of coal thickness, in %; m is the average coal thickness in the unit area, in m; s is the standard deviation of coal thickness variation;

[0035] Among them, the formula for the standard deviation of coal thickness change is formula (4):

[0036]

[0037] m i is the measured thickness of the coal points, in m; n is the number of coal points.

[0038] Optionally, in S33, based on the random forest regression model, the coal seam parameter value of the target coal seam is corrected according to the indicator influencing factors, and the corrected interpolation result obtained includes:

[0039] The coal seam depth, fault density, fold deformation degree, coal seam thickness, coal seam inclination, coal thickness variation coefficient, soft layer thickness and spatial position coordinates were input into the random forest regression model, and the training set residual r t As the output of the random forest regression model, the random forest regression model is trained to obtain a trained random forest regression model;

[0040] According to the test set S cThe trained random forest regression model is tested with the data to obtain the corrected residual r c ;

[0041] Correct the Kriging interpolation result to obtain the corrected interpolation result Z p , Z p =Z' c +r c .

[0042] Optionally, the evaluation in S34 is based on whether the interpolation result after the correction of the random forest regression model meets the preset requirements. If so, the random forest regression model is saved and S35 is executed. If not, the random forest regression model is retrained and the process returns to step S33, including:

[0043] Comparison of measurement data Z c And the interpolation result Z p , the evaluation indicators determination coefficient and Pearson correlation coefficient were used to evaluate the combined interpolation results;

[0044] If the coefficient of determination and the Pearson correlation coefficient are both higher than 0.85, the random forest regression model is saved and step S35 is executed;

[0045] Otherwise, retrain the random forest regression model and return to step S33.

[0046] Re-adjust the hyperparameters of the random forest regression model, such as n_estimators (number of trees), max_depth (maximum depth), min_samples_split (minimum sample split), etc., use RandomizedSearchCV to find the best parameter combination, and finally retrain and evaluate the model until the accuracy requirements are met.

[0047] Optionally, the interpolation correction in the to-be-interpolated area in S35 includes:

[0048] S351, divide the interpolation area into grids with a grid side length of b, and obtain the spatial position P of the interpolation point N ;

[0049] S352, based on the measured point data S = (X, Y, Z) calculate the Kriging interpolation result of each interpolation point as Z' N ;

[0050] S353, calculating the influencing factors of the point to be interpolated according to step S33, where the influencing factor calculation unit area is a square area with a side length of a=2b centered at the point to be interpolated;

[0051] S354, the influencing factors and spatial position P of the point to be interpolated are combinedN Input the saved random forest regression model and get the corrected residual r N ;

[0052] S355, apply the residual corrected Kriging interpolation result obtained by the random forest regression model to obtain the final interpolation result Z of the interpolation point N =Z' N +r N .

[0053] Optionally, the outburst hazard level of each area of the target coal seam is obtained based on the vertical stress, horizontal stress, gas pressure, gas content, coal solidity coefficient f value and coal gas release initial velocity ΔP of the target coal seam in combination with the extension matter-element model, including:

[0054] S41. Obtain the subjective weight of each indicator according to the hierarchical analysis method, obtain the objective weight of each indicator according to the entropy weight method, and obtain the comprehensive weight of each indicator based on the subjective weight and the objective weight of each indicator;

[0055] S42. Apply the matter-element extension model to integrate the prediction indicators of various indicators and calculate the comprehensive index Q of coal seam outburst hazard;

[0056] S43. Define the hazard level based on the range of the comprehensive outburst hazard index Q and divide the coal seam area into hazard areas. The hazard level definition includes:

[0057] When 0≤Q<50, the current area is predicted to be a low-risk area;

[0058] When 50≤Q<75, the current area is predicted to be a medium-risk area;

[0059] When 75≤Q≤100, the current area is predicted to be a high-risk area.

[0060] S44. Based on the division results of coal seam area danger zones, a coal seam outburst hazard comprehensive index distribution map and an outburst hazard zone prediction map are drawn.

[0061] Optionally, the matter-element extension model applied in S42 is used to fuse the prediction indicators of various indicators to calculate the comprehensive index Q of coal seam outburst hazard, including:

[0062] S421. Construct a matter-element matrix. The matter-element matrix is expressed as formula (5):

[0063]

[0064] Among them, R is the coal seam outburst hazard element; R i is the element of R; N is the coal seam outburst hazard; C is the regional index reflecting the coal seam outburst hazard; V is the index value;

[0065] S422, determine the classical domain and the node domain, the classical domain matter element R j Expressed as formula (6):

[0066]

[0067] Among them, N j is the outburst hazard of the jth coal seam, v ij is the value range of the ith indicator of the risk level;

[0068] Section matter element R p Expressed as formula (7):

[0069]

[0070] Among them, v ip The matter element of this section is about the feature c i The value range of a ip is the lower limit of the value interval of the i-th indicator, b ip is the upper limit of the value range of the i-th indicator;

[0071] S423. Determine the matter-element to be evaluated and the correlation function, wherein the formula for determining the matter-element to be evaluated is formula (8):

[0072]

[0073] The formula for calculating the correlation function is formula (9):

[0074]

[0075] Among them, k xj (v i ) is the xth thing-element to be evaluated, the jth degree, and the correlation degree of the i-th indicator;

[0076] Among them, the distance from a certain point x to each classical domain interval X is formula (10):

[0077]

[0078] Where X is the length of the classical domain interval; a and b are the classical domain values, i.e., the lower and upper limits of each indicator; x is the data of the area to be predicted;

[0079] S424. Calculate the comprehensive correlation, which is formula (11):

[0080]

[0081] Among them, w i is the comprehensive weight value of the indicator; k j (v i) The calculated indicator correlation value;

[0082] S425. Based on the maximum membership principle, the comprehensive index of coal seam regional outburst hazard is calculated as formula (12):

[0083] Q=max(k j (N x ));(12).

[0084] Compared with the prior art, the above technical solution has at least the following beneficial effects:

[0085] The present invention first targets the three elements of coal outbursts: ground stress, gas, and the mechanical properties of coal body structure, and considers the influence of factors such as coal seam burial depth, geological structure, mining disturbance, gas occurrence, and soft stratification. It then constructs an indicator system for predicting coal outburst risk areas, including six indicators: vertical stress, horizontal stress, gas pressure, gas content, coal solidity coefficient f value, and coal gas release initial velocity ΔP. Subsequently, a precise numerical simulation of the spatial distribution of coal seam vertical stress and horizontal stress is performed. Then, based on the spatial combination interpolation method of Kriging and random forest, spatial interpolation of gas pressure, gas content, coal solidity coefficient f value, and coal gas release initial velocity ΔP is achieved respectively. Finally, an extensional matter-element model is applied to fuse regional prediction indicators, calculate the comprehensive index of coal outburst risk, draw a distribution map of the comprehensive index of coal outburst risk, and predict coal seams as high, medium, or low risk areas based on the range of the comprehensive index of coal outburst risk. The present invention can realize quantitative regional prediction of coal seam outburst hazard. Taking into account the influence of factors affecting regional prediction indicators, it creatively proposes a spatial combination interpolation method based on Kriging and random forest to improve the spatial interpolation accuracy of regional prediction indicators; it creatively integrates ground stress, gas and coal body structural mechanical property indicators to propose a quantitative outburst hazard comprehensive index, which is of great significance for achieving regional hazard fine management and precise prevention and control of outburst disasters. BRIEF DESCRIPTION OF THE DRAWINGS

[0086] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0087] Figure 1 This is the execution process of the multi-factor fusion quantitative regional prediction method for coal seam outburst hazard provided by the embodiment of the present invention;

[0088] Figure 2 A relationship diagram of the influencing factors, regional prediction indicators, three elements and outburst risk provided by an embodiment of the present invention;

[0089] Figure 3 Surface contour lines of a coal seam and a coal seam contour map provided by an embodiment of the present invention;

[0090] Figure 4 The execution process of the spatial combined interpolation method based on Kriging and random forest provided in the embodiment of the present invention;

[0091] Figure 5 Gas pressure interpolation results using different interpolation methods provided in the embodiments of the present invention;

[0092] Figure 6 A gas pressure distribution diagram of a coal seam provided in an embodiment of the present invention;

[0093] Figure 7 A distribution map of the comprehensive index of coal seam outburst hazard provided by an embodiment of the present invention;

[0094] Figure 8 A prediction map of a coal seam outburst hazard area provided by an embodiment of the present invention;

[0095] Figure 9 A graph showing changes in effectiveness inspection indicators and comprehensive outburst hazard index with working face length provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0096] To make the purpose, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the described embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0097] Unless otherwise defined, the technical or scientific terms used in the present invention shall have the usual meaning understood by persons of ordinary skill in the field to which the present invention belongs. The words "first", "second" and similar terms used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. Similarly, words such as "one", "an" or "the" do not indicate a quantity limitation, but rather indicate the presence of at least one. Words such as "include" or "comprise" mean that the elements or objects preceding the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. Words such as "connect" or "connected" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect.

[0098] It should be noted that the terms "up", "down", "left", "right", "front" and "back" used in the present invention are only used to indicate relative position relationships. When the absolute position of the object being described changes, the relative position relationship may also change accordingly.

[0099] Taking an existing coal seam as an example, the maximum buried depth of this coal seam is more than 1000m, which is representative of the quantitative regional prediction research of coal seam outburst hazard. Based on this, this embodiment takes this coal seam as the engineering background, collects relevant data, and provides a quantitative regional prediction method for coal seam outburst hazard by integrating multiple factors. The execution process of this method is as follows: Figure 1 The specific contents are as follows:

[0100] A quantitative regional prediction method for coal seam outburst hazard based on multi-factor integration, including:

[0101] S1. Obtain the gas occurrence, geological structure, coal seam burial depth, soft layer and mining conditions of the target coal seam to obtain basic information of the target coal seam;

[0102] S2. Based on the basic information of the target coal seam, combined with Rhino and FLAC 3D Software, simulating and obtaining the spatial distribution of vertical stress and horizontal stress of the target coal seam;

[0103] S3. Obtaining coal seam parameter values of the target coal seam based on a spatial combined interpolation method of Kriging and random forest, wherein the coal seam parameter values include gas pressure, gas content, coal solidity coefficient f value, and coal gas release initial velocity ΔP;

[0104] S4. Based on the vertical stress, horizontal stress, gas pressure, gas content, coal strength coefficient f value and coal gas release initial velocity ΔP of the target coal seam, combined with the extension matter-element model, the outburst hazard level of each area of the target coal seam is obtained.

[0105] In a specific implementation method, step S1 specifically includes: considering the influence of factors such as coal seam depth, geological structure, mining disturbance, gas occurrence and soft stratification, and constructing a regional prediction index system for outburst danger, including six indicators: vertical stress, horizontal stress, gas pressure, gas content, coal strength coefficient f value and coal gas release initial velocity ΔP. The relationship between influencing factors, regional prediction indicators, three elements and outburst danger is shown in the figure below: Figure 2 shown.

[0106] In a specific implementation manner, step S2 is mainly used to complete a fine numerical simulation of the spatial distribution of vertical stress and horizontal stress in the coal seam, specifically including:

[0107] S21. Based on the surface contours, coal seam depth, geological structure and mining conditions, the surface contours and coal seam contours were processed into a polyhedral format using Rhino software to obtain a large numerical model for coal seam stress calculation that includes the main geological structures.

[0108] In step S21, the surface contour lines and coal seam contour lines are vertically translated according to the actual assigned heights based on the actual geological features such as the surface contour lines, coal seam burial depth, geological structure and mining disturbance. The above contour lines are processed into a polysurface format using Rhino software to restore the actual undulations and occurrence of the coal seam surface and coal seam, and a large model for numerical calculation of coal seam stress including the main geological structures is established. Figure 3 shown.

[0109] S22, bringing the surface contour lines, coal seam depth and elastic modulus of the coal seam into Rhino software to obtain a polyhedron, bringing the polyhedron into the coal seam stress numerical calculation model to obtain a FLAC 3D Large model for numerical calculation of coal seam stress based on unit cell grid.

[0110] In this step, according to the surface and coal seam parameters, modeling is performed in Rhino software to generate polysurfaces, and the polysurfaces are converted into FLAC 3D Unit cell grid. The model's X-axis length corresponds to the strike length of the mine, and its Y-axis length corresponds to the dip length of the mine. Grid measurement points are placed in the coal seam to record changes in coal seam stress.

[0111] S23, assigning the model to the FLAC according to the coal rock mechanical parameters in the mine geological report 3D In the large numerical calculation model of coal seam stress of the unit grid, the spatial distribution of vertical stress and horizontal stress of the target coal seam is obtained.

[0112] In this step, the model is assigned values according to the coal rock mechanical parameters in the mine geological report, and the gravity acceleration is set to 9.8m / s 2 Fixed boundaries are applied to the model's perimeter to restrict movement around the perimeter and at the bottom. The top of the model represents the ground surface, where a free boundary is set. Fixed boundaries are used at the bottom of the model to prevent node movement in space, while rolling boundaries are used on the remaining surfaces, restricting node movement to two-dimensional motion along the boundaries. The constitutive model is a strain-softening model.

[0113] In a specific implementation, step S3 is mainly based on the spatial combination interpolation method of Kriging and random forest to realize the spatial interpolation of gas pressure, gas content, coal solidity coefficient f value and coal gas release initial velocity ΔP respectively. The specific implementation is as follows:

[0114] S31, based on Kriging interpolation and residual calculation, obtaining the coal seam parameter values of the measuring points in the target coal seam and the corresponding training set residuals;

[0115] S311. Extracting gas pressure, gas content, coal solidity coefficient f, coal gas release initial velocity ΔP, and measuring point locations from the coal seam gas geological map based on Python to form an indicator data set S = (X, Y, Z) corresponding to the measuring point locations and the measured data.

[0116] In this example, Python was used to extract gas pressure, gas content, coal solidity coefficient f, and coal gas initial velocity ΔP from a coal seam gas geological map. The measured data, P = (X, Y), and Z, were used to generate an index dataset S = (X, Y, Z) corresponding to the measured data. 274 sets of gas pressure data, 254 sets of gas content data, and 495 sets of coal solidity coefficient f and coal gas initial velocity ΔP data were extracted.

[0117] S312: Divide the index data set into interpolation data sets S according to the ratio of 2:2:1. k =(X k ,Y k ,Z k ), training set S t =(X t ,Y t ,Z t ) and the test set S c =(X c ,Y c ,Z c );

[0118] S313, based on the interpolation data set S k =(X k ,Y k ,Z k ), the measurement point position P of the training set t =(X t ,Y t ) and the test point position P of the test set c =(X c ,Y c ), perform Kriging spatial interpolation on the measurement point positions of the training set and the test set to obtain the Kriging interpolation result Z' of the training set t And the Kriging interpolation result Z' of the test set c ;

[0119] S314, according to the measurement point data Z of the training set t And the Kriging interpolation result Z' t , and get the corresponding training set residual r t =Zt -Z' t .

[0120] S32. Calculate the factors influencing the forecast indicators, including:

[0121] The fault density is defined as follows:

[0122]

[0123] Among them, E fd is the fault density in the unit area; H is the fault drop, in m; L is the fault strike length, in m; α is the fault dip; β is the angle between the principal stress and the fault strike; S is the unit area, in m 2 ; n is the number of faults in the unit area; the unit area is a square area with a side length of a centered at the measuring point;

[0124] The degree of wrinkle deformation is defined as formula (2):

[0125]

[0126] Among them, F d is the degree of wrinkle deformation in the unit area; h is the elevation distance between adjacent contour lines, in meters; k is the actual length of the contour line with the largest curvature in the unit area, in meters; l is the secant length of the contour line with the largest curvature, in meters; d is the distance between the two contour lines with the largest curvature, in meters;

[0127] The coefficient of variation of coal thickness is defined as formula (3):

[0128]

[0129] Wherein, γ is the coefficient of variation of coal thickness, in %; m is the average coal thickness in the unit area, in m; s is the standard deviation of coal thickness variation;

[0130] Among them, the formula for the standard deviation of coal thickness change is formula (4):

[0131]

[0132] m i is the measured thickness of the coal points, in m; n is the number of coal points.

[0133] S33. Based on a random forest regression model, the coal seam parameter value of the target coal seam is corrected according to the indicator influencing factors to obtain a corrected interpolation result;

[0134] The coal seam depth, fault density, fold deformation degree, coal seam thickness, coal seam inclination, coal thickness variation coefficient, soft layer thickness and spatial position coordinates were input into the random forest regression model, and the training set residual r t As the output of the random forest regression model, the random forest regression model is trained to obtain a trained random forest regression model;

[0135] According to the test set S c The trained random forest regression model is tested with the data to obtain the corrected residual r c ;

[0136] Correct the Kriging interpolation result to obtain the corrected interpolation result Z p , Z p =Z' c +r c .

[0137] In this embodiment, the training set data is used for model training. For gas pressure and gas content, the coal seam depth (A1), fault density (A2), fold deformation degree (A3), coal seam thickness (A4), coal seam inclination (A5), coal thickness variation coefficient (A6), soft layer thickness (A7), and spatial position coordinates are used as model inputs. For the coal solidity coefficient f value and the coal gas release initial velocity ΔP, the fault density (A2), fold deformation degree (A3), soft layer thickness (A7), and spatial position coordinates are used as model inputs. The training set residual r t As the model output, the random forest regression model is trained to obtain the trained random forest regression model. The trained model is tested using the test set data. The influencing factors and spatial positions of gas pressure, gas content, coal solidity coefficient f value and coal gas release initial velocity ΔP are input, and the corrected residual r is output. c , correct the Kriging interpolation result, that is, the final interpolation result is Z p =Z' c +r c .

[0138] S34, evaluating whether the interpolation result after correction based on the random forest regression model meets the preset requirements. If so, saving the random forest regression model and executing S35; if not, retraining the random forest regression model and returning to step S33;

[0139] Comparison of measurement data Z c And the interpolation result Z p , the evaluation indicators determination coefficient and Pearson correlation coefficient were used to evaluate the combined interpolation results;

[0140] If the coefficient of determination and the Pearson correlation coefficient are both higher than 0.85, the random forest regression model is saved and step S35 is executed;

[0141] Otherwise, retrain the random forest regression model and return to step S33. Readjust the hyperparameters of the random forest regression model, such as n_estimators (number of trees), max_depth (maximum depth), min_samples_split (minimum sample split), etc., and use randomized search (RandomizedSearchCV) to find the best parameter combination. Finally, retrain the model and evaluate it until the accuracy requirements are met.

[0142] In this embodiment, the comparison measurement data Z c And the interpolation result Z p , the determination coefficient of the application evaluation index (R 2 ) and the Pearson correlation coefficient (R) are used to evaluate the combined interpolation results. If the model evaluation meets the requirements, the random forest regression model is saved for use in correcting the interpolation results. Otherwise, the model is retrained.

[0143] S35: Correct the interpolation in the area to be interpolated.

[0144] The Kriging interpolation values, machine learning interpolation values, and combined interpolation values are compared with the measured values, and the residuals between the interpolation values and the measured values are calculated. Figure 5 The interpolation results and residual analysis for coal gas pressure using different interpolation methods are shown. These results are based on Kriging, the Random Forest Regression Model (RFR Model), and a combination of Kriging and Random Forest Regression (Kriging-RFR Model). This example demonstrates the feasibility and applicability of the proposed spatial combination interpolation method based on Kriging and Random Forest for spatial interpolation of prediction indicators.

[0145] Depend on Figure 5 It can be seen that the Kriging-RFR Model interpolation and the determination coefficient R of gas pressure 2 The correlation coefficient (R) is 0.89, and the correlation coefficient (R) is 0.95, which is higher than Kriging's 0.81 and 0.90, and the RFR Model's 0.84 and 0.93. The residuals of the interpolated and measured gas pressures are normally distributed and concentrated near 0, with low means and standard deviations. Therefore, the proposed spatial combined interpolation method based on kriging and random forests can be used for spatial interpolation of prediction indicators and can improve spatial interpolation accuracy compared to traditional kriging interpolation methods.

[0146] Step S35 specifically includes:

[0147] S351, divide the interpolation area into grids with a grid side length of b = 10m, and obtain the spatial position P of the interpolation point N ;

[0148] S352, based on the measured point data S = (X, Y, Z) calculate the Kriging interpolation result of each interpolation point as Z' N ;

[0149] S353, calculating the influencing factors of the point to be interpolated according to step S33, where the influencing factor calculation unit area is a square area with a side length of a=2b=20m centered on the point to be interpolated;

[0150] S354, the influencing factors and spatial position P of the point to be interpolated are combined N Input the saved random forest regression model and get the corrected residual r N ;

[0151] S355, apply the residual corrected Kriging interpolation result obtained by the random forest regression model to obtain the final interpolation result Z of the interpolation point N =Z' N +r N .

[0152] The spatial combined interpolation method based on Kriging and random forest is applied to obtain the coal seam gas pressure distribution map, such as Figure 6 .

[0153] In a specific implementation method, step S4 applies an extension matter-element model to integrate regional prediction indicators, calculate a comprehensive index of outburst hazard, draw a distribution map of the comprehensive index of outburst hazard, and predict the coal seam as a high, medium, or low hazard zone based on the range of the comprehensive index of outburst hazard. Specifically, the method includes:

[0154] S41. Obtain the subjective weight of each indicator according to the hierarchical analysis method, obtain the objective weight of each indicator according to the entropy weight method, and obtain the comprehensive weight of each indicator based on the subjective weight and the objective weight of each indicator;

[0155] One specific implementation method uses the analytic hierarchy process (AHP) to calculate the subjective weights of prediction indicators. The principal component analysis hierarchical structure model uses coal seam outburst hazard as the target layer and includes six prediction indicators: vertical stress, horizontal stress, gas pressure, gas content, coal strength coefficient f, and coal gas release initial velocity ΔP. The importance of each of the six prediction indicators to outburst hazard is compared pairwise to obtain a prediction indicator judgment matrix:

[0156]

[0157] For the judgment matrix A, use Python to calculate the maximum eigenvalue λ of the judgment matrix A max=6.10, calculate the eigenvector corresponding to the maximum eigenvalue, and normalize it to get the weight vector:

[0158] W=[0.3686,0.1387,0.2115,0.1387,0.0875,0.0550] T

[0159] Calculate the consistency index C I =0.02, consistency ratio C R =0.016<0.1. Therefore, the inconsistency of the judgment matrix is within the allowable range. After passing the consistency test, the normalized eigenvector corresponding to its maximum eigenvalue can be used as the weight vector.

[0160] The entropy weight method was used to calculate the objective weights of the prediction indicators. Vertical and horizontal stress data were obtained through numerical simulation. Gas pressure, gas content, coal strength coefficient f, and coal gas release initial velocity ΔP were obtained through spatial interpolation of measured data. The entropy weights of each indicator were calculated and used as the objective weights of the prediction indicators.

[0161] Applying the analytic hierarchy process and entropy weight method, the subjective weight and objective weight of each prediction indicator are calculated. The subjective weight obtained by the analytic hierarchy process and the objective weight obtained by the entropy weight method are combined to obtain the comprehensive weight w i :

[0162]

[0163] The subjective weight, objective weight and comprehensive weight of each prediction indicator are shown in Table 1.

[0164] Table-1

[0165]

[0166] S42. Apply the matter-element extension model to integrate the prediction indicators of various indicators and calculate the comprehensive index Q of coal seam outburst hazard.

[0167] S421. Construct a matter-element matrix. The matter-element matrix is expressed as formula (5):

[0168]

[0169] Among them, R is the coal seam outburst hazard element; R i is the element of R; N is the coal seam outburst hazard; C is the regional index reflecting the coal seam outburst hazard; V is the index value;

[0170] S422, determine the classical domain and the node domain, the classical domain matter element R j It is expressed as formula (6):

[0171]

[0172] Among them, N j is the outburst hazard of the jth coal seam, v ij is the value range of the ith indicator of the risk level;

[0173] Section matter element R p Expressed as formula (7):

[0174]

[0175] Among them, v ip The matter element of this section is about the feature c i The value range of a ip is the lower limit of the value interval of the i-th indicator, b ip is the upper limit of the value range of the i-th indicator;

[0176] In this example, the coal seam outburst hazard is used as the object N in the matter-element matrix, the six prediction indicators are used as the features C, and the indicator data is used as the quantity Y. According to industry standards such as the "Coal Mine Safety Regulations" and the "Detailed Rules for Preventing and Controlling Coal and Gas Outbursts," the classical domain and section domain of each prediction indicator are determined, as shown in Table 2.

[0177] Table-2

[0178]

[0179] S423. Determine the matter-element to be evaluated and the correlation function, wherein the formula for determining the matter-element to be evaluated is formula (8):

[0180]

[0181] The formula for calculating the correlation function is formula (9):

[0182]

[0183] Among them, k xj (v i ) is the xth thing-element to be evaluated, the jth degree, and the correlation degree of the i-th indicator;

[0184] Among them, the distance from a certain point x to each classical domain interval X is formula (10):

[0185]

[0186] Where X is the length of the classical domain interval; a and b are the classical domain values, i.e., the lower and upper limits of each indicator; x is the data of the area to be predicted;

[0187] S424. Calculate the comprehensive correlation, which is formula (11):

[0188]

[0189] Among them, w i is the comprehensive weight value of the indicator; k j (v i ) The calculated indicator correlation value;

[0190] S425. Based on the maximum membership principle, the comprehensive index of coal seam regional outburst hazard is calculated as formula (12):

[0191] S43. Define the hazard level based on the range of the comprehensive outburst hazard index Q and divide the coal seam area into hazard areas. The hazard level definition includes:

[0192] When 0≤Q<50, the current area is predicted to be a low-risk area;

[0193] When 50≤Q<75, the current area is predicted to be a medium-risk area;

[0194] When 75≤Q≤100, the current area is predicted to be a high-risk area.

[0195] S44, based on the division results of coal seam area danger zones, draw a coal seam outburst hazard comprehensive index distribution map and an outburst hazard zone prediction map, such as Figure 7 、 Figure 8 shown.

[0196] In order to illustrate the consistency of the change trend of the efficiency inspection index and the comprehensive index of outburst danger Q during the mining process of the working face, the prediction results of the outburst danger area are verified. Taking a certain mining face as an example, based on the daily advance of the working face, the initial velocity of gas outburst q value and the drill cuttings S value, the following is drawn: Figure 9 The graph shows the changes of the inspection index and the comprehensive index of the protruding hazard with the length of the working face.

[0197] Depend on Figure 9As can be seen, the outburst hazard index within the working face length range of 0 to 400 m is high, indicating a high-risk zone. Correspondingly, the q and S values are both higher than in other areas. The q value exceeded its critical value of 3.5 L / min three times, at 6.02 L / min, 4.87 L / min, and 4.54 L / min, respectively, while the S value did not exceed its critical value of 5 kg / m. Furthermore, during mining in this area, coal discharge from the working face led to multiple gas anomalies, with the maximum gas concentration in the upper corner reaching 0.75%. The outburst hazard index within the working face length range of 1000 to 1400 m is relatively low, but still above 65, indicating a medium-risk zone. Compared to the high-value areas mentioned above, the q and S values are lower, neither exceeding their critical values. The S value is relatively low and stable. However, fault structures exist in this area, resulting in significant fluctuations in the q value. In summary, the quantitative regional prediction method of protruding hazards based on multi-factor integration proposed in this paper has certain rationality and applicability.

[0198] The following points need to be explained:

[0199] (1) The drawings of the embodiments of the present invention only relate to the structures related to the embodiments of the present invention. Other structures may refer to conventional designs.

[0200] (2) For the sake of clarity, the thickness of layers or regions in the drawings used to describe the embodiments of the present invention are exaggerated or reduced, that is, these drawings are not drawn to scale. It is understood that when an element such as a layer, film, region, or substrate is referred to as being "on" or "under" another element, the element may be "directly" "on" or "under" the other element or intervening elements may be present.

[0201] (3) In the absence of conflict, the embodiments of the present invention and the features therein may be combined with each other to form new embodiments.

[0202] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. The protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A quantitative regional prediction method for coal seam outburst hazard by integrating multiple factors, characterized in that: include: Obtain the gas occurrence, geological structure, coal seam burial depth, soft layer and mining conditions of the target coal seam to obtain basic information of the target coal seam; Based on the basic information of the target coal seam, combined with Rhino and FLAC 3D Software, simulating and obtaining the spatial distribution of vertical stress and horizontal stress of the target coal seam; Based on the spatial combined interpolation method of Kriging and random forest, the spatial distribution of the coal seam parameter values of the target coal seam is obtained, and the coal seam parameter values include gas pressure, gas content, coal solidity coefficient f value and coal gas release initial velocity ΔP; Based on the vertical stress, horizontal stress, gas pressure, gas content, coal strength coefficient f value and coal gas release initial velocity ΔP of the target coal seam, combined with the extension matter-element model, the outburst hazard level of each area of the target coal seam is obtained.

2. The method for quantitative regional prediction of coal seam outburst hazard based on multi-factor fusion according to claim 1 is characterized in that: According to the basic information of the target coal seam, combined with Rhino and FLAC 3D Software, simulating the spatial distribution of vertical stress and horizontal stress of the target coal seam includes: Based on the surface contours, coal seam depth, geological structure and mining conditions, the surface contours and coal seam contours were processed into a polysurface format using Rhino software to obtain a large numerical calculation model for coal seam stress that includes the main geological structures. The surface contour lines, coal seam depth and elastic modulus of the coal seam are brought into Rhino software to obtain a polyhedron, and the polyhedron is brought into the large model of coal seam stress numerical calculation to obtain a FLAC 3D Large model for numerical calculation of coal seam stress based on unit cell grid; Assign the model to the FLAC according to the coal rock mechanical parameters in the mine geological report 3D In the large numerical calculation model of coal seam stress of the unit grid, the spatial distribution of vertical stress and horizontal stress of the target coal seam is obtained.

3. According to the multi-factor fusion quantitative regional prediction method for coal seam outburst hazard of claim 2, the spatial combined interpolation method based on Kriging and random forest is used to obtain the coal seam parameter values of the target coal seam, including: S31, based on Kriging interpolation and residual calculation, obtaining the coal seam parameter values of the measuring points in the target coal seam and the corresponding training set residuals; S32, calculating the factors influencing the prediction indicators; S33. Based on a random forest regression model, the coal seam parameter value of the target coal seam is corrected according to the indicator influencing factors to obtain a corrected interpolation result; S34, evaluating whether the interpolation result after correction based on the random forest regression model meets the preset requirements. If so, saving the random forest regression model and executing S35; if not, retraining the random forest regression model and returning to step S33; S35: Correct the interpolation in the area to be interpolated.

4. The method for quantitative regional prediction of coal seam outburst hazard based on multi-factor fusion according to claim 3 is characterized in that: The coal seam parameter values of the measuring points in the target coal seam and the corresponding training set residuals obtained in S31 based on Kriging interpolation and residual calculation include: S311. Extracting gas pressure, gas content, coal solidity coefficient f, coal gas release initial velocity ΔP, and measuring point locations from the coal seam gas geological map based on Python to form an indicator data set S = (X, Y, Z) corresponding to the measuring point locations and the measured data. S312: Divide the index data set into interpolation data sets S according to the ratio of 2:2:

1. k =(X k ,Y k ,Z k ), training set S t =(X t ,Y t ,Z t ) and the test set S c =(X c ,Y c ,Z c ); S313, based on the interpolation data set S k =(X k ,Y k ,Z k ), the measurement point position P of the training set t =(X t ,Y t ) and the test point position P of the test set c =(X c ,Y c ), perform Kriging spatial interpolation on the measurement point positions of the training set and the test set to obtain the Kriging interpolation result Z' of the training set t And the Kriging interpolation result Z' of the test set c ; S314, according to the measurement point data Z of the training set t And the Kriging interpolation result Z' t , and get the corresponding training set residual r t =Z t -Z' t .

5. The method for quantitative regional prediction of coal seam outburst hazard based on multi-factor fusion according to claim 4 is characterized in that: The calculation of the influencing factors of the prediction index in S32 includes: The fault density is defined as follows: Among them, E fd is the fault density in the unit area; H is the fault drop, in m; L is the fault strike length, in m; α is the fault dip; β is the angle between the principal stress and the fault strike; S is the unit area, in m 2 ; n is the number of faults in the unit area; the unit area is a square area with a side length of a centered at the measuring point; The degree of wrinkle deformation is defined as formula (2): Among them, F d is the degree of wrinkle deformation in the unit area; h is the elevation distance between adjacent contour lines, in meters; k is the actual length of the contour line with the largest curvature in the unit area, in meters; l is the secant length of the contour line with the largest curvature, in meters; d is the distance between the two contour lines with the largest curvature, in meters; The coefficient of variation of coal thickness is defined as formula (3): Wherein, γ is the coefficient of variation of coal thickness, in %; is the average coal thickness in the unit area, in meters; s is the standard deviation of coal thickness variation; Among them, the formula for the standard deviation of coal thickness change is formula (4): m i is the measured thickness of the coal points, in m; n is the number of coal points.

6. According to the multi-factor fusion quantitative regional prediction method for coal seam outburst hazard of claim 5, the random forest regression model in S33 is used to correct the coal seam parameter values of the target coal seam according to the index influencing factors, and the corrected interpolation results obtained include: The coal seam depth, fault density, fold deformation degree, coal seam thickness, coal seam inclination, coal thickness variation coefficient, soft layer thickness and spatial position coordinates were input into the random forest regression model, and the training set residual r t As the output of the random forest regression model, the random forest regression model is trained to obtain a trained random forest regression model; According to the test set S c The trained random forest regression model is tested with the data to obtain the corrected residual r c ; Correct the Kriging interpolation result to obtain the corrected interpolation result Z p , Z p =Z' c +r c .

7. The method for quantitative regional prediction of coal seam outburst hazard based on multi-factor fusion according to claim 6 is characterized in that: The evaluation in S34 is based on whether the interpolation result after the correction of the random forest regression model meets the preset requirements. If so, the random forest regression model is saved and S35 is executed. If not, the random forest regression model is retrained and the process returns to step S33, including: Comparison of measurement data Z c And the interpolation result Z p , the evaluation indicators determination coefficient and Pearson correlation coefficient were used to evaluate the combined interpolation results; If the coefficient of determination and the Pearson correlation coefficient are both higher than 0.85, the random forest regression model is saved and step S35 is executed; Otherwise, retrain the random forest regression model and return to step S33; The retraining of the random forest regression model includes: re-adjusting the hyperparameters of the random forest regression model and using random search to find the optimal parameter combination.

8. The method for quantitative regional prediction of coal seam outburst hazard based on multi-factor fusion according to claim 7 is characterized in that: The correction of interpolation to the area to be interpolated in S35 includes: S351, divide the interpolation area into grids with a grid side length of b, and obtain the spatial position P of the interpolation point N ; S352, based on the measured point data S = (X, Y, Z) calculate the Kriging interpolation result of each interpolation point as Z' N ; S353, calculating the influencing factors of the point to be interpolated according to step S33, where the influencing factor calculation unit area is a square area with a side length of a=2b centered at the point to be interpolated; S354, the influencing factors and spatial position P of the point to be interpolated are combined N Input the saved random forest regression model and get the corrected residual r N ; S355, apply the residual corrected Kriging interpolation result obtained by the random forest regression model to obtain the final interpolation result Z of the interpolation point N =Z' N +r N .

9. The method for quantitative regional prediction of coal seam outburst hazard based on multi-factor fusion according to claim 8, characterized in that: Based on the vertical stress, horizontal stress, gas pressure, gas content, coal solidity coefficient f value and coal gas release initial velocity ΔP of the target coal seam, combined with the extension matter-element model, the outburst hazard level of each area of the target coal seam is obtained, including: S41. Obtain the subjective weight of each indicator according to the hierarchical analysis method, obtain the objective weight of each indicator according to the entropy weight method, and obtain the comprehensive weight of each indicator based on the subjective weight and the objective weight of each indicator; S42. Apply the matter-element extension model to integrate the prediction indicators of various indicators and calculate the comprehensive index Q of coal seam outburst hazard; S43. Define the hazard level based on the range of the comprehensive outburst hazard index Q and divide the coal seam area into hazard areas. The hazard level definition includes: When 0≤Q<50, the current area is predicted to be a low-risk area; When 50≤Q<75, the current area is predicted to be a medium-risk area; When 75≤Q≤100, the current area is predicted to be a high-risk area; S44. Based on the division results of coal seam area danger zones, a coal seam outburst hazard comprehensive index distribution map and an outburst hazard zone prediction map are drawn.

10. The method for quantitative regional prediction of coal seam outburst hazard based on multi-factor fusion according to claim 9, characterized in that: The matter-element extension model used in S42 is used to integrate the prediction indicators of various indicators and calculate the comprehensive index Q of coal seam outburst hazard, including: S421. Construct a matter-element matrix. The matter-element matrix is expressed as formula (5): Among them, R is the coal seam outburst hazard element; R i is the element of R; N is the coal seam outburst hazard; C is the regional index reflecting the coal seam outburst hazard; V is the index value; S422, determine the classical domain and the node domain, the classical domain matter element R j It is expressed as formula (6): Among them, N j is the outburst hazard of the jth coal seam, v ij is the value range of the ith indicator of the risk level; Section matter element R p Expressed as formula (7): Among them, v ip The matter element of this section is about the feature c i The value range of a ip is the lower limit of the value interval of the i-th indicator, b ip is the upper limit of the value range of the i-th indicator; S423. Determine the matter-element to be evaluated and the correlation function, wherein the formula for determining the matter-element to be evaluated is formula (8): The formula for calculating the correlation function is formula (9): Among them, k xj (v i ) is the xth thing-element to be evaluated, the jth degree, and the correlation degree of the i-th indicator; Among them, the distance from a certain point x to each classical domain interval X is formula (10): Where X is the length of the classic domain interval; a is the classic domain value, which is the lower limit of each indicator; b is the classic domain value, which is the upper limit of each indicator; x is the data of the area to be predicted; S424. Calculate the comprehensive correlation, which is formula (11): Among them, w i is the comprehensive weight value of the indicator; k j (v i ) The calculated indicator correlation value; S425. Based on the maximum membership principle, the comprehensive index of coal seam regional outburst hazard is calculated as formula (12): Q=max(k j (Nx));(12)。

Images