Coal and gas outburst prediction method based on extension-fuzzy analytic hierarchy process theory

Abstract

The present application relates to coal and gas outburst disaster prediction technical field, especially based on the coal and gas outburst prediction method of fuzzy analytic hierarchy process theory of extension, including the construction of outburst risk prediction system;The construction of matter element extension prediction framework;Determine the weight of prediction index;Calculate the correlation and its bias of risk grade;The present application comprehensively considers coal seam depth, coal seam solidity coefficient, coal seam gas content, coal seam damage type, gas pressure, gas diffusion initial velocity and other key indexes of coal and gas outburst, and the objective weight and subjective weight of various prediction indexes can be calculated by clear mathematical method, and there is no non-convergence problem.

Description

Technical Field

[0001] This invention relates to the field of coal and gas outburst disaster prediction technology, and in particular to a coal and gas outburst prediction method based on extension-fuzzy hierarchical analysis theory. Background Technology

[0002] As coal mining depths gradually increase, ground stress, gas pressure, and gas content also rise, leading to a continuous increase in the frequency and intensity of coal and gas outbursts. Coal and gas outbursts involve the violent ejection of large amounts of coal seam gas and methane, posing a significant hazard. Furthermore, the main component of coal seam gas is methane, and the large amounts of methane emitted during outbursts exacerbate the greenhouse effect. Because coal and gas outbursts involve multi-field, multi-phase occurrence, gas adsorption and desorption, and spatiotemporal evolution of multiple physical fields, their mechanisms are extremely complex. Predicting coal and gas outbursts has become a global challenge for ensuring safe coal mine production and protecting the ecological environment.

[0003] In recent years, scholars both at home and abroad have conducted extensive research on coal and gas outburst prediction, proposing a variety of prediction methods and technologies, such as:

[0004] Chinese patent publication number CN111079978A discloses a method for predicting coal and gas outbursts based on logistic regression and reinforcement learning. This method integrates LR and ADABOOST reinforcement learning to design a coal and gas outburst prediction model, collects data samples of various influencing factors of coal and gas outbursts, and trains and corrects errors in the outburst prediction model based on the data samples.

[0005] Chinese patent CN107194524B discloses a coal and gas outburst prediction method based on RBF neural network. The prediction steps are as follows: performing dimensionality reduction and normalization on the outburst feature data, calculating the center of the radial basis function, introducing an adaptive differential evolution algorithm to determine the optimal expansion factor and optimal weight when the number of hidden layer neurons is determined, determining the prediction model of RBF neural network, and making predictions based on test data.

[0006] Chinese patent publication number CN109492816B discloses a dynamic prediction method for coal and gas outbursts based on hybrid intelligence. The prediction steps are as follows: data detection, data processing using the mean batch estimation fusion method, forming a new problem for prediction, and prediction model verification and correction.

[0007] Chinese patent publication number CN112183901A discloses a method for predicting the intensity of coal and gas outbursts based on deep learning. The prediction steps are as follows: data preparation, feature extraction, configuration of the learning process, training of the model, and verification of the model.

[0008] In summary, although existing methods and technologies for predicting coal and gas outbursts each have their own characteristics, these coal and gas outburst prediction models based on artificial intelligence and computer methods suffer from problems such as difficulty in determining the membership degree and weight of indicators, and difficulty in achieving convergence in the calculation process. Summary of the Invention

[0009] The purpose of this invention is to provide a coal and gas outburst prediction method based on extension-fuzzy hierarchical analysis theory, to solve the problems of difficulty in determining the membership degree and weight of indicators, and difficulty in convergence of the calculation process, in coal and gas outburst prediction models based on artificial intelligence and computer methods. To achieve the above objective, this invention provides the following technical solution:

[0010] This invention provides a method for predicting coal and gas outbursts based on extension-fuzzy hierarchical analysis theory, comprising the following steps:

[0011] Step 1: Construct a coal and gas outburst risk prediction system; First, determine the coal seam depth, coal seam firmness coefficient, coal seam gas content, coal seam failure type, gas pressure, and initial gas release velocity as indicators for coal and gas outburst prediction; then classify the outburst risk levels.

[0012] Step 2: Construct a matter-element extension prediction framework and establish a dimensionless matter-element extension prediction model based on extension theory;

[0013] Step 3: Determine the weights of the predictive indicators; first, determine the objective weights through correlation analysis, then determine the subjective weights through fuzzy hierarchical analysis theory, and finally determine the comprehensive weights of the indicators through the comprehensive weight allocation method.

[0014] Step 4: Calculate the correlation and bias of risk levels; calculate the correlation between the predictive indicators and the prominent risk levels through the correlation function, and use the maximum correlation criterion to identify the final prominent risk level.

[0015] As a further technical solution, in step 1, the risk level is divided into four levels: no risk, low risk, medium risk, and high risk. Each prominent prediction indicator is divided into four risk levels according to the magnitude of the measured value.

[0016] As a further technical solution, in step 3, a fuzzy consistency judgment matrix is ​​first constructed through expert subjective analysis, and the judgment matrix is ​​adjusted until the difference between the elements in the first row and the corresponding elements in the other rows is a constant.

[0017] As a further technical solution, if the adjusted, consistent judgment matrix differs significantly from the subjective analysis, then the subjective analysis will ultimately be used as the standard to establish an inconsistent judgment matrix.

[0018] As a further technical solution, the least squares method is used to solve the constrained programming problem and obtain the subjective weights of the prediction indicators.

[0019] As a further technical solution, the constrained programming problem and the unconstrained programming problem are treated as equivalent using the Lagrange multiplier method.

[0020] As a further technical solution, in step 1, the larger the coal seam firmness coefficient, the lower the outburst risk; and the larger the other five outburst prediction indicators, the higher the outburst risk.

[0021] As a further technical solution, in step 2, when establishing a dimensionless matter-element extension prediction model based on extension theory, the prediction index values ​​are first processed to be dimensionless.

[0022] As a further technical solution, in step 2, the dimensionless matter-element extension prediction model established based on extension theory is as follows:

[0023]

[0024] All of the above predictive indices have the same cross-sectional area (v iP = (<0,1>), the classic field of the index v ij = ij ,b ij >(1≤i≤6, 1≤j≤4) corresponds to the range of each prediction level in the formula; where, C i (1≤i≤6) represents the prominent predictive indicator, N i (1≤i≤4) represents a prominent risk level.

[0025] As a further technical solution, when calculating the objective weight of a predictive indicator, when a certain prominent predictive indicator value 'a' is... i In the interval v ij = ij ,b ij >Inner time, a i With v ij The correlation between them can be obtained using a simple correlation function.

[0026] The beneficial effects of the present invention are as follows:

[0027] (1) This invention comprehensively considers key indicators of coal and gas outburst, such as coal seam burial depth, coal seam firmness coefficient, coal seam gas content, coal seam failure type, gas pressure, and initial gas release velocity, and the prediction results are highly accurate.

[0028] ​​(2) This invention first determines the objective weights through correlation analysis, then determines the subjective weights through fuzzy hierarchical analysis theory, and finally determines the comprehensive weights of the indicators through the comprehensive weight allocation method. Therefore, the objective and subjective weights of various predictive indicators can be obtained by calculating using explicit mathematical methods.

[0029] (3) The prediction process of this invention does not rely on machine learning, has a fast calculation speed, and does not have a non-convergence problem, making it suitable for a variety of complex geological conditions. Attached Figure Description

[0030] The accompanying drawings, which form part of this specification, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute a limitation thereof. It should also be understood that these drawings are for simplicity and clarity and are not necessarily drawn to scale. The invention will now be described and explained with additional features and details using the drawings, wherein:

[0031] Figure 1 A flowchart for predicting coal and gas outbursts in an embodiment of the present invention is shown;

[0032] Figure 2 This illustrates the variation law of the outstanding coal and rock mass quality and the maximum correlation value in embodiments of the present invention;

[0033] Figure 3 This illustrates the impact of reduced gas content on the overall weight in an embodiment of the present invention;

[0034] Figure 4 This illustrates the impact of reduced gas pressure on the overall weight in an embodiment of the present invention;

[0035] Figure 5 This illustrates the impact of simultaneously reducing gas content and pressure on the overall weight in an embodiment of the present invention;

[0036] Figure 6 The impact of gas extraction on the risk level of outbursts is illustrated in an embodiment of the present invention. Detailed Implementation

[0037] The technical solutions in typical embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings.

[0038] Example

[0039] like Figures 1 to 6 As shown, this embodiment provides a coal and gas outburst prediction method based on extension-fuzzy hierarchical analysis theory, including the following steps:

[0040] Step 1: Emphasize the construction of a risk prediction system

[0041] Step 1.1: Selection of the indicator system

[0042] According to the "Detailed Rules for the Implementation of Prevention and Control of Coal and Gas Outbursts", and in accordance with the principles of importance, simplicity and quantification, the following six evaluation indicators were determined: coal seam burial depth, coal seam firmness coefficient, coal seam gas content, coal seam failure type, gas pressure, and initial gas emission velocity, which are represented by the symbols H, f, Q, Z, p, and Δp, respectively. Coal seam burial depth H refers to the depth of the coal seam from the surface; coal seam firmness coefficient f is a parameter representing the mechanical strength of the coal seam; the lower the mechanical strength of the coal seam, the lower the coal seam firmness coefficient; coal seam gas content Q refers to the amount of gas contained in a unit mass of coal under mine atmospheric conditions (ambient temperature 20℃, ambient atmospheric pressure 0.1MPa); coal seam failure type Z refers to the type classified according to the degree of coal fracturing, indicating the degree of coal seam fragmentation and wrinkling under tectonic stress; gas pressure p refers to the gas pressure within the coal seam; initial gas release velocity Δp refers to the velocity of coal seam gas outburst when the coal is initially exposed, reflecting not only the coal's ability to release gas but also the gas infiltration and flow patterns. Coal seam failure types are divided into five levels: simple structure, normal, few faults and folds, relatively developed faults and folds, and well-developed faults and folds, assigned values ​​from 1 to 5 respectively. The higher the coal seam firmness coefficient, the lower the risk of outburst; the higher the other five indicators, the higher the risk of outburst.

[0043] Step 1.2: Risk Level Classification

[0044] The risk levels are categorized into four levels: no risk (N1), low risk (N2), medium risk (N3), and high risk (N4). Six predictive indicators are further classified into four risk levels based on their measured values, thus constructing a hierarchical structure for coal and gas outburst risk level evaluation that meets the data requirements of extension theory and fuzzy hierarchical analysis theory. The classification standards for the predictive indicators are shown in Table 1.

[0045] Table 1 Quantitative Classification Standards for Coal and Gas Outburst Prediction Indicators

[0046]

[0047] Step 2: Constructing a Matter-Element Extension Prediction Framework

[0048] To couple the analysis of different prediction indicators, the prediction indicator values ​​are treated as dimensionless. The coal seam firmness coefficient is treated according to equation (1), and the other indicators are treated according to equation (2).

[0049]

[0050]

[0051] In the formula, a is the actual value of the predicted indicator. max and a minThe maximum and minimum values ​​that this indicator can reach in the coal mine are given respectively. ā is a dimensionless indicator value. The upper limits of the predicted indicators C1 to C6 are 1500, 2, 25, 5, 3, and 35, respectively, and the lower limits are all 0.

[0052] Matter-element theory and extended set theory are the two theoretical foundations of extension theory. The former constructs material elements and their transformations, while the latter quantifies the results of extension theory. Extension theory regards the objective world as a material world, and contradictory problems can be transformed into material factors. Extension theory has very important guiding significance for the prediction of coal and gas outburst hazards. Introducing extension theory into the prediction of coal and gas outbursts can open up new ideas and directions for the prevention and control of coal and gas outbursts. The ordered triple R = (N, c, ν) is used as the basic element (called material element) to describe the object, where N represents matter, c is a characteristic, and ν is the measure of matter N on characteristic c. The expression ν = c(N) describes the relationship between the qualitative and quantitative characteristics of matter. Matter with multiple characteristic elements can be described by n-dimensional material elements, as shown in equation (3).

[0053]

[0054] Extension theory has been used to solve evaluation problems. The work on extension evaluation methods mainly includes three parts: the construction of the extension evaluation system, the correlation analysis between the evaluation object and the evaluation level, and the identification of the evaluation level of the evaluation object.

[0055] (1) Cross-sectional domain, classical domain and elements with the same characteristic features

[0056] Suppose we use indices c1, c2, ..., c n The evaluation object P, where P is the collective term for all problems to be evaluated in extension theory. i The value of (1≤i≤n) is in ν iP = iP b iP Within the range of >. R P It is an element of the evaluation object P and can be represented by equation (4).

[0057]

[0058] In the formula, the value range ν iP = iP b iP This is called the cross-sectional region. ij = ij b ij This is called a classical domain. According to the above definition, a classical domain is contained within a cross-sectional domain, that is...

[0059] ​​​The dimensionless matter-element extension prediction model based on extension theory is shown in equations (5) to (6).

[0060]

[0061] According to equation (5)R P The assessment objects and the coal and gas outburst prediction indicators shown in Table 1 will be used to predict R. P Instantiated as a dimensionless matter-element extension prediction model R that specifically characterizes the risk of coal and gas outbursts F As shown in Equation 6.

[0062]

[0063] All of the above predictive indices have the same cross-sectional area (v iP = (<0,1>), the classic field of the index v ij = ij ,b ij >(1≤i≤6,1≤j≤4) corresponds to the range of each prediction level in equation (6).

[0064] Step 3: Determining the weights of the predictive indicators

[0065] The weights of the coal and gas outburst prediction indicators are determined using the comprehensive weight allocation method shown in Equation (7).

[0066] w i =ψ o w io +ψ s w is ψ o +ψ s =1 (7)

[0067] In the formula, w i It is the predictive indicator C i The overall weight, w is and w io These are subjective weights and objective weights, respectively; Ψ s and Ψ o These represent the proportions of subjective weight and objective weight, respectively, both of which are set to 0.5.

[0068] Step 3.1: Determining Objective Weights

[0069] Objective weights are determined through correlation analysis. In the prominent risk prediction system established above, the prediction indicator values ​​for different risk levels do not overlap. Therefore, when a certain prominent prediction indicator value 'a'... i In the interval v ij = ij ,b ij >Inner time, a i With v​​ij The correlation between them can be obtained by a simple correlation function as shown in equation (8).

[0070]

[0071] In the formula, K ij (a i v ij The expression satisfies 1≤i≤6 and 1≤j≤4.

[0072] Then, the prediction index a is obtained through equations (9) to (10). i Objective weighting.

[0073]

[0074]

[0075] In the formula, w io It is the objective weight of the i-th predictive indicator.

[0076] Step 3.2: Determining Subjective Weights

[0077] Subjective weights are determined by fuzzy hierarchical analysis theory. First, a fuzzy consistency judgment matrix R is constructed through expert subjective analysis to represent the relative importance among different prominent prediction indicators. The matrix R can be expressed as equation (11).

[0078]

[0079] In the formula, r ij Representing element a i and element a j The fuzzy membership degree between them is quantitatively represented using a scale of 0.1 to 0.9, and the specific meanings are shown in Table 2.

[0080] Table 2 Quantitative Scales

[0081]

[0082] The fuzzy consistency judgment matrix R should satisfy equation (12).

[0083]

[0084] In practical decision analysis, due to the complexity of coal and gas outburst problems and the one-sidedness of subjective understanding, the constructed judgment matrix R often does not satisfy the above properties and needs to be adjusted according to the following steps:

[0085] 1) Select the element C1 with the highest confidence level for analysis to determine r. 11 r 12 ,...,r 16value;

[0086] 2) If the difference between the elements in the first row and the corresponding elements in the second row of the matrix is not a constant, adjust the elements in the second row until the differences are all constants;

[0087] 3) If the difference between the elements in the first row and the corresponding elements in the third row of the matrix is not a constant, adjust the elements in the third row until the differences are all constants;

[0088] 4) Adjust sequentially until the differences between the elements in the first row and the corresponding elements in the remaining rows are all constants.

[0089] When the judgment matrix R has consistency, there is a relationship shown in formula (13) between the subjective weight values w 1s , w 2s , …, w 6s .

[0090] r ij = 0.5 + x(w is - w js ) + 0.5 i, j = 1, 2, ..., 6 (13)

[0091] In the formula, x is a measure of the difference degree of the prediction index, related to the number and difference degree of the indexes, and the value range is 0 < x ≤  0.5. The larger the number or difference degree of the elements, the larger the value of x.

[0092] If there is a large difference between the judgment matrix with consistency after adjustment and the subjective analysis, then finally, based on the subjective analysis, a judgment matrix without consistency is established. At this time, formula (13) is no longer applicable, and the least squares method shown in formula (14) needs to be used to solve the constrained programming problem to obtain the subjective weights of the prediction indexes.

[0093]

[0094] It can be known from the Lagrange multiplier method that the constrained programming problem and the unconstrained programming problem can be equivalently processed through formula (15).

[0095]

[0096] In the formula, λ is the Lagrange multiplier.

[0097] By taking the partial derivative of L(w, λ) with respect to w is and setting it to zero, a system of equations shown in formulas (16) - (17) can be obtained.

[0098]

[0099]

[0100] Combine equations (16) to (17) with w 1s +w 2s +…+w 6s =1 By simultaneously establishing the final system of equations, the subjective weight vector W = [w 1s ,w 2s ,…,w 6s ] T The values ​​of each element.

[0101] Step 4: Calculation of risk level correlation and its bias

[0102] If the value of a certain prominent predictive indicator is a i In the interval v ij = ij ,b ij Within this range, the correlation K between the predictive indicators and the four prominent risk levels is... j (X) can be calculated using the correlation functions shown in equations (18) to (22).

[0103]

[0104]

[0105]

[0106] ∣ν ij |=|b ij -a ij | (21)

[0107]

[0108] In the formula, w i These are the weights of the predictive indicators.

[0109] The maximum relevance criterion shown in equation (23) is used to identify the final salient risk level, i.e., the maximum K. j The risk level corresponding to (X) is the final predicted risk level for an outburst under this geological parameter.

[0110]

[0111] Bias in the final predicted risk level j * Calculated using equations (24) to (25).

[0112]

[0113]

[0114] in and ​These are the maximum and minimum correlations, respectively.

[0115] Test case

[0116] This experimental example uses a publicly available method for predicting coal and gas outbursts based on extension-fuzzy hierarchical analysis theory to predict coal and gas outbursts in 12 high-gas mines.

[0117] Based on the prediction model established above, the measured values ​​of prediction indicators for 8 groups of mines that experienced coal and gas outburst accidents and 4 groups of high-gas mines that did not experience outburst accidents were obtained through extensive surveys, as shown in Table 3.

[0118] Table 3 Actual values ​​of indicators in the coal and gas outburst prediction model

[0119]

[0120] First, the predicted index values ​​of the 12 mines were processed dimensionlessly using equations (1) to (2), and the processing results are shown in Table 4.

[0121] Table 4 Dimensionless values ​​of indicators for coal and gas outburst prediction model

[0122]

[0123]

[0124] The objective weights of the prominent prediction indicators shown in Table 5 are obtained by formulas (8) to (10). It can be seen that the objective weights of prediction indicators for different mines fluctuate greatly, among which the objective weights of gas pressure, initial gas release velocity and coal body firmness coefficient are relatively large.

[0125] Table 5 Objective Weights of Indicators in the Coal and Gas Outburst Prediction Model

[0126]

[0127] Through expert subjective evaluation and consistency adjustment, the fuzzy consistency judgment matrix shown in equation (26) is obtained.

[0128]

[0129] Equation (26) satisfies the consistency requirement. The measure of the degree of difference in perceived objects, a, is set to 0.5. Equation (13) is then compared with w. 1s +w 2s +…+w 6s =1. A system of equations was established, and the subjective weights of the prediction indicators were finally obtained as shown in Table 6.

[0130] Table 6 Subjective Weights of Coal and Gas Outburst Prediction Indicators

[0131]

[0132] The comprehensive weights of the predictive indicators are obtained by formula (7) as shown in Table 7. Based on the average value of the comprehensive weights, the importance of the predictive indicators is ranked as follows: p > Δp > Z > f > Q > H.

[0133] Table 7. Comprehensive Weights of Indicators in the Coal and Gas Outburst Prediction Model

[0134]

[0135] The correlation K between the predictive indicators and the risk level, as shown in Table 8, is obtained by using equations (18) to (25). j (X) and the final predicted risk level and its bias. Of the 8 groups of mines that experienced coal and gas outburst accidents, Yangquan No. 5 Mine is classified as N3 (medium risk), and the remaining 7 groups are classified as N4 (high risk). Of the 4 groups of mines that did not experience outburst accidents, 3 are classified as N2 (low risk), and 1 group is classified as N3 (medium risk). The predicted outburst risk level results show good agreement with the actual occurrence of outburst disasters, verifying the feasibility of the coal and gas outburst prediction model based on extension-fuzzy hierarchical analysis theory.

[0136] Table 8. Prediction Results of Coal and Gas Outbursts

[0137]

[0138]

[0139] Further analysis was conducted on the correlation between the mass of outburst-prone coal and rock mass and the maximum risk level in 12 groups of mines, with the mass of outburst-prone coal and rock mass in mines where no outbursts occurred being set to 0. Figure 2 It can be seen that the variation law of the correlation value between the quality of the coal and rock mass and the maximum risk level is basically consistent.

[0140] Currently, the main outburst prevention measure adopted in high-gas mines is gas drainage. Gas drainage primarily reduces the gas pressure and content in the coal seam. To analyze the outburst prevention effect of gas drainage, an outburst prediction model was applied to quantitatively study the impact of changes in gas pressure and gas content during gas drainage in the outburst-prone coal seam of Wangfenggang Mine on the outburst risk level.

[0141] The analysis process sets the reduction rate of gas pressure and gas content during gas extraction into nine levels: 10%, 20%, ..., 90%. Figure 3 It can be seen that the decrease in gas pressure and gas content has a relatively small impact on the comprehensive weight of the prediction indicators, and the effect of gas pressure is greater than that of gas content. This is because the comprehensive weight is calculated by combining objective and subjective weights. The decrease in gas pressure and gas content only changes the objective weight of the prediction indicators, and has a relatively small impact on the final result of the comprehensive weight.

[0142] Depend on Figure 4 Therefore, if gas drainage only reduces the gas content of the coal seam, even if the gas content is reduced by 90%, the outburst risk level remains N4 (high risk). If gas drainage only reduces the gas pressure of the coal seam, when the gas pressure is reduced by 50% (to 1.25 MPa), the outburst risk level drops to N3 (medium risk); when the gas pressure is reduced by 70% (to 0.75 MPa), the outburst risk level drops to N2 (low risk); and when the gas pressure is reduced by 90% (to 0.25 MPa), the outburst risk level drops to N1 (no risk). When both gas content and gas pressure are reduced simultaneously, the change pattern of the coal seam risk level is basically the same as when only gas pressure is reduced, the only difference being that when both gas content and gas pressure are reduced by 60%, the outburst risk level becomes N2 (low risk). Therefore, the main reason why gas drainage reduces the outburst risk level is by reducing the coal seam gas pressure.

[0143] Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make possible changes and modifications to the technical solutions of the present invention by utilizing the methods and techniques disclosed above without departing from the spirit and scope of the present invention. Therefore, any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solutions of the present invention shall fall within the protection scope of the technical solutions of the present invention.

Claims

1. A method for predicting coal and gas outbursts based on extension-fuzzy hierarchical analysis theory, characterized in that, Includes the following steps: Construct a risk prediction system for coal and gas outbursts; first, determine the coal seam depth, coal seam firmness coefficient, coal seam gas content, coal seam failure type, gas pressure, and initial gas release velocity as indicators for predicting coal and gas outbursts; then, classify outburst risk levels. Constructing a matter-element extension prediction framework; establishing a dimensionless matter-element extension prediction model based on extension theory; The dimensionless matter-element extension prediction model based on extension theory takes the following form: All of the above predictive indicators have the same cross-sectional area. v iP =<0,1>, the classic field of the index v ij =< a ij , b ij >,1≤ i ≤6,1≤ j ≤4, corresponding to the range of each prediction level in the formula; where, C i， 1≤ i ≤6 indicates a prominent predictive indicator. N i， 1≤ i ≤4 indicates a high level of risk; Determine the weights of the predictive indicators; first, determine the objective weights through correlation analysis, then determine the subjective weights through fuzzy hierarchical analysis theory, and finally determine the comprehensive weights of the indicators through the comprehensive weight allocation method. Calculate the correlation and bias of risk levels; calculate the correlation between predictive indicators and prominent risk levels through correlation functions, and use the maximum correlation criterion to identify the final prominent risk level.

2. The coal and gas outburst prediction method based on extension-fuzzy hierarchical analysis theory as described in claim 1, characterized in that, The risk levels are divided into four categories: no risk, low risk, medium risk, and high risk. Each key predictive indicator is also divided into four risk levels based on the magnitude of its measured value.

3. The coal and gas outburst prediction method based on extension-fuzzy hierarchical analysis theory as described in claim 1, characterized in that, First, construct a fuzzy consistency judgment matrix, and then adjust the judgment matrix until the difference between the elements in the first row and the corresponding elements in the other rows is a constant.

4. The coal and gas outburst prediction method based on extension-fuzzy hierarchical analysis theory as described in claim 3, characterized in that, If the adjusted, consistent judgment matrix differs significantly from the subjective analysis, then the subjective analysis will be used as the standard to establish an inconsistent judgment matrix.

5. The coal and gas outburst prediction method based on extension-fuzzy hierarchical analysis theory as described in claim 4, characterized in that, By solving constrained programming problems using the least squares method, we can obtain the subjective weights of the prediction indicators.

6. The coal and gas outburst prediction method based on extension-fuzzy hierarchical analysis theory as described in claim 5, characterized in that, The constrained programming problem and the unconstrained programming problem are treated as equivalent using the Lagrange multiplier method.

7. The coal and gas outburst prediction method based on extension-fuzzy hierarchical analysis theory as described in claim 1, characterized in that, The higher the coal seam firmness coefficient, the lower the risk of outburst; the higher the other five outburst prediction indicators, the higher the risk of outburst.

8. The coal and gas outburst prediction method based on extension-fuzzy hierarchical analysis theory as described in claim 1, characterized in that, When establishing a dimensionless matter-element extension prediction model based on extension theory, the prediction index values ​​are first processed to be dimensionless.

9. The coal and gas outburst prediction method based on extension-fuzzy hierarchical analysis theory as described in claim 1, characterized in that, When calculating the objective weight of a predictive indicator, when a certain prominent predictive indicator value a i In the interval v ij =< a ij ,b ij >Inner time, a i and v ij The correlation between them can be obtained using a simple correlation function.

Images