Three-dimensional energy bending moment coupling fracturing point location determination method for hard roof mine earthquake prevention and control

Abstract

According to the three-dimensional energy bending moment coupling fracturing point location determination method for hard roof mine earthquake prevention and control, the optimal scheme that fracturing point locations are optimized through a three-dimensional energy bending moment coupling model in combination with a neural network algorithm, training is conducted through a multi-layer feedforward neural network, and the fracturing point locations are predicted is adopted. The input of the network comprises physical parameters of the ore bed and initial geologic features of bending moment energy distribution. The network output is the three-dimensional coordinates of the fracturing points. The problem that mine earthquakes cannot be accurately predicted through a traditional fracturing method is solved. By performing coupling analysis on the three-dimensional bending moment field and the energy field of the coal mine roof, the fracturing point location can be more accurately determined, so that mine earthquakes caused by fracture of the hard roof are effectively prevented and treated.

Description

Technical Field

[0001] The present invention relates to the field of prevention and control of mine dynamic disasters in underground coal mining, and particularly relates to a method for determining three-dimensional energy-bending moment coupling fracturing points for preventing and controlling mine tremors in hard roof coal mines. Background Art

[0002] The mine tremor disaster in coal mining is one of the main disasters that plague the safe and efficient mining of coal seams. The prevention and control of mine tremor disasters have the characteristics of uncertainty in broken rock strata, irregularity in the breaking step distance of hard roofs, and difficulty in actually measuring the energy and bending moment accumulation positions of hard roofs. When mine tremors are severe, they can lead to the occurrence of rock bursts, causing damage to working face equipment, threatening the safety of workers, and triggering secondary disasters such as water inrush and other serious problems. After a mine tremor occurs, the working face is shut down, resulting in a decrease in the production efficiency of the working face and a decline in coal production rate, causing losses to the national economy. Due to the rapidity and unpredictability of mine tremors caused by the breaking of hard roofs, it is difficult to predict mine tremor disasters during the coal production process, which increases the difficulty of preventing and controlling mine tremor disasters caused by the breaking of hard roofs.

[0003] The mine tremor disaster caused by the breaking of hard roofs greatly restricts the efficient mining of coal seams. Traditional control methods include technologies such as hydraulic fracturing in wells and deep-hole blasting. However, these methods are difficult to accurately determine the breaking position and energy accumulation position of hard roofs. The control technology guided by experience greatly increases the control cost of mine tremor disasters caused by the breaking of hard roofs, resulting in a decline in the economic benefits of coal extraction in mines and a general prevention and control effect. A large amount of control costs are consumed, reducing the economic value of coal. Therefore, accurately locating the fracturing points of hard roofs and reducing the prevention and control cost of mine tremors caused by hard roofs are important problems that urgently need to be solved in coal mines with hard roofs. Summary of the Invention

[0004] The present application provides a method for determining three-dimensional energy-bending moment coupling fracturing points for preventing and controlling mine tremors in hard roof coal mines to solve the problem of mine tremor disasters caused by the breaking of hard roofs in related technologies, accurately identify the bending moment and energy accumulation positions of hard roofs, and identify the three-dimensional energy and bending moment coupling points of hard roofs. Use ground fracturing technology to perform hydraulic fracturing on the fracturing points to solve the high-cost problem of preventing and controlling mine tremor disasters based on experience.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A method for determining three-dimensional energy-bending moment coupling fracturing points for preventing and controlling mine tremors in hard roof coal mines, comprising the following steps:

[0007] Step 1: Construct a three-dimensional space bending moment distribution model of mine tremors in hard roof coal mines; the bending moment field of the hard roof satisfies the following micro-element relationship:

[0008]

[0009] In the formula, M(x, y, z) is the distribution of the bending moment of the hard roof in three-dimensional space; x, y, and z are the three-dimensional coordinates of the mine, and their specific meanings are the coordinate along the working face advancing direction, the coordinate perpendicular to the working face advancing direction, and the coordinate in the working face burial depth direction respectively; α, β, and γ are constants related to material anisotropy, reflecting the non-linear stress-strain relationship of the bending moment field.

[0010] Step 2: Determine the non-linear coupling micro-element relation of the three-dimensional energy distribution of the hard roof in coal mines; considering the fracture and plastic deformation effects of the rock stratum, the energy concentration is expressed as:

[0011]

[0012] In the formula, E(x, y, z) represents the energy concentration, u(x, y, z) is the displacement field of the hard roof, λ, μ, and ν are the elastic constants of the material, where ν is the Poisson's ratio, and R nonlinear (x, y, z) is the non-linear part, considering the stress softening and plastic influence of the hard roof.

[0013]

[0014] Step 3: Determine the coupling stress-bending moment micro-element relation of the hard roof in coal mines;

[0015]

[0016] In the formula, σ x , σ y , σ z are the stresses in the x, y, and z directions respectively; u(x, y, z) is the displacement field of the hard roof; D is the elastic modulus of the hard roof.

[0017] Step 4: Through the neural network optimization algorithm, establish a neural network optimization model to identify the characteristics of the bending moment and energy distribution of the hard roof; the loss function of the neural network is designed in the following form:

[0018]

[0019] In the formula, L is the loss function, which is used to measure the sum of the error of the model prediction and the regularization constraint; are the loss accumulation terms respectively; is the predicted value; r i is the actual target value; N is the number of executions; f j is the variable output by the function or model, which is used when the input data is the bending moment; g i is the variable output by the model, which is used when r k is the variable output by the model, which is used when r iThis item is used when the input data is energy; i is the increment of the prediction times, increasing by 1 each time; j is the increment of the energy density prediction, increasing by 1 each time; k is the increment of the bending moment prediction, increasing by 1 each time, t is the number of bending moment cycles; λ1 is the regularization coefficient, controlling the contribution degree of this item to the loss function; λ2 is the regularization coefficient, controlling the influence of this item on the loss function; s is the number of energy cycles;

[0020] Step 5: The iterative form of the micro-element relation in the multi-loop solution process; in each iteration, using the results of the current calculated bending moment field and energy field, update the prediction results of the fracturing points; the equation for each iteration can be expressed in the following form:

[0021] M (n+1) = M (n) + ΔM (n) ,E (n+1) = E (n) + ΔE (n)

[0022] M (n) and E (n) respectively represent the bending moment and energy at the nth iteration; M (n+1) and E (n+1) respectively represent the bending moment and energy at the (n + 1)th iteration; ΔM (n) and ΔE (n) respectively represent the increments of the bending moment and energy updated in this iteration;

[0023] Step 6: Export the iterative data, generate the three-dimensional coupling diagram of the bending moment and energy according to the neural network discrimination; according to the physical and mechanical parameters of the hard roof overlying the coal seam, cycle through the processes of Step 1 to Step 5, export the distribution data of the bending moment and energy density of the hard roof, and generate the three-dimensional coupling diagram of the bending moment and energy of the hard roof according to the exported data; discriminate the distribution points of the bending moment and energy, and determine that the highest point of the bending moment and energy distribution is the fracturing point.

[0024] Furthermore, it also includes Step 7: Establish the coupling control equation considering damage evolution and optimize the fracturing point data; the optimization control relationship of its damage evolution is as follows:

[0025]

[0026] In the formula, F is the damage factor of the hard roof, ρ is the density of the hard roof, g is the acceleration due to gravity; the above relationship can optimize the fracturing points determined in Step 6 and normalize the fracturing points of the same hard roof layer.

[0027] Further, it also includes step 8: fracturing and engineering evaluation of the determined hard roof fracturing points; after the fracturing points are implemented on-site, through the mine seismic monitoring data, monitor the decrease in the frequency of mine seismicity after fracturing; especially the change in the intensity of mine seismicity in the high-energy concentration area around the fracturing points. A method for determining the three-dimensional energy-bending moment coupling fracturing points for preventing and controlling mine seismicity of hard roofs provided by the present invention realizes the effective prevention and control of mine seismic disasters caused by the fracture of hard roofs through a series of processes including constructing a three-dimensional space bending moment distribution model of mine seismicity of hard roofs in coal mines, determining the non-linear coupling micro-element relationship of the three-dimensional energy distribution of hard roofs in coal mines, determining the coupling stress-bending moment micro-element relationship of hard roofs in coal mines, establishing a neural network optimization model through a neural network optimization algorithm to distinguish the bending moment and energy distribution characteristics of hard roofs, the iterative form of the micro-element relationship in the multiple-loop solution process, deriving iterative data, generating a three-dimensional map of the coupling of bending moment and energy based on neural network discrimination, establishing a coupling control equation considering damage evolution, optimizing the fracturing point data, and fracturing and engineering evaluation of the determined hard roof fracturing points. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 It is a flow chart of the specific implementation steps of the present invention.

[0029] Figure 2 It is a schematic diagram for illustrating an embodiment, 1 - coal seam, 2 - soft rock stratum, 3 - hard roof, 4 - fracturing well, 5 - fracturing point, 6 - hydraulic support.

[0030] Figure 3 In step 6 of the embodiment, a distribution map of the coupling of the first bending moment and energy is derived. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0031] The present invention is a method for determining the three-dimensional energy-bending moment coupling fracturing points for preventing and controlling mine seismicity of hard roofs. It realizes the effective prevention and control of mine seismic disasters caused by the fracture of hard roofs through a series of processes including constructing a three-dimensional space bending moment distribution model of mine seismicity of hard roofs in coal mines, determining the non-linear coupling micro-element relationship of the three-dimensional energy distribution of hard roofs in coal mines, determining the coupling stress-bending moment micro-element relationship of hard roofs in coal mines, establishing a neural network optimization model through a neural network optimization algorithm to distinguish the bending moment and energy distribution characteristics of hard roofs, the iterative form of the micro-element relationship in the multiple-loop solution process, deriving iterative data, generating a three-dimensional map of the coupling of bending moment and energy based on neural network discrimination, establishing a coupling control equation considering damage evolution, optimizing the fracturing point data, and fracturing and engineering evaluation of the determined hard roof fracturing points.

[0032] Specifically, it includes the following steps:

[0033] Step 1: Construction of the three-dimensional spatial bending moment distribution model of hard roof rock bursts in coal mines; During the underground coal mining process, the coupled distribution of bending moment and energy is the main cause of rock stratum fracture and the formation of rock bursts. In three-dimensional space, the distribution of bending moment and energy usually not only involves second-order differential equations but also needs to consider factors such as the inhomogeneity, anisotropy of the rock stratum, and stress-strain relationship. Therefore, more terms need to be added to the partial differential equation of the bending moment, specifically manifested as higher-order non-linear terms. The non-linear characteristic is a complex manifestation form of the bending moment and energy distribution. Considering the above factors, the bending moment field of the hard roof satisfies the following differential element relationship:

[0034]

[0035] In the formula, M(x, y, z) is the distribution of the bending moment of the hard roof in three-dimensional space; x, y, z are the three-dimensional coordinates of the mine, and their specific meanings are the coordinate along the working face advancing direction, the coordinate perpendicular to the working face advancing direction, and the coordinate in the direction of the working face burial depth respectively; α, β, γ are constants related to material anisotropy, reflecting the non-linear stress-strain relationship of the bending moment field.

[0036] Step 2: Determine the non-linear coupling differential element relationship of the three-dimensional energy distribution of the hard roof in coal mines; During the coal mining process, the release of energy caused by the fracture of the hard roof is the key factor for rock bursts. For the calculation of energy, not only the second-order differential element relationship of the bending moment needs to be considered, but also the non-linear concentration differential element relationship of energy needs to be introduced, and this non-linear concentration differential element relationship considers the non-linear characteristics of the rock stratum. Considering the fracture and plastic deformation effects of the rock stratum, the energy concentration can be expressed as:

[0037]

[0038] In the formula, E(x, y, z) represents the energy concentration, u(x, y, z) is the displacement field of the hard roof, λ, μ, ν are the elastic constants of the material, where v is the Poisson's ratio, and R nonlinear (x, y, z) is the non-linear part, considering the effects of stress softening and plasticity of the hard roof, etc.

[0039]

[0040] Step 3: Determine the coupled stress-bending moment differential element relationship of the hard roof in coal mines; The bending moment is the main factor for the fracture of the hard roof in coal mines after coal seam mining. At the same time, the large-scale roof fracture caused by the coupled fracture of the bending moment and stress causes strong rock bursts. The coupling relationship between the bending moment field and the stress field in space can be described by a more complex elastic model, using a more accurate constitutive relationship:

[0041]

[0042] In the formula, σx , σ y , σ z are the stresses in the x, y, and z directions respectively; u(x, y, z) is the three-dimensional displacement (displacement field) of the hard roof; D is the elastic modulus of the hard roof.

[0043] Step 4: Establish a neural network optimization model through the neural network optimization algorithm to identify the bending moment and energy distribution characteristics of the hard roof; the optimization of the neural network model requires the introduction of a more complex objective function and more precise training in combination with the backpropagation algorithm. The fracturing of the hard roof in coal mines needs to consider multiple factors, and the loss function of the neural network is designed in the following form:

[0044]

[0045] In the formula, L is the loss function, which is used to measure the sum of the error of the model prediction and the regularization constraint; are the loss accumulation terms respectively; is the predicted value; r i is the actual target value; N is the number of executions; f j is the variable output by the function or model, which is used when r i the input data is the bending moment; g k is the variable output by the model, which is used when r i the input data is the energy; i is the prediction number increment, increasing by 1 each time; j is the energy prediction increment, increasing by 1 each time; k is the bending moment prediction increment, increasing by 1 each time, t is the number of bending moment cycles; λ1 is the regularization coefficient, controlling the contribution of this term to the loss function; λ2 is the regularization coefficient, controlling the influence of this term on the loss function; s is the number of energy cycles.

[0046] Step 5: The iterative form of the microelement relationship in the multiple-loop solution process; due to considering the plastic and softening characteristics of the hard roof, there may be blank phenomena in the calculation process of Step 4. Therefore, in each solution, the parameters can be updated through multiple loops to further optimize the model. In each iteration, we use the results of the current calculated bending moment field and energy field to update the prediction results of the fracturing points. The equation for each iteration can be expressed in the following form:

[0047] M (n+1) = M (n) + ΔM (n) , E (n+1) = E (n) + ΔE (n)

[0048] M (n) and E (n) represent the bending moment and energy at the nth iteration respectively; M (n+1) and E(n+1) represent the bending moment and energy of the (n + 1)-th iteration respectively; ΔM (n) and ΔE (n) represent the increments of the bending moment and energy updated in this iteration respectively.

[0049] Step 6: Export the iterative data, and generate a three-dimensional diagram of the coupling of the bending moment and energy according to the neural network discrimination; according to the physical and mechanical parameters of the hard roof overlying the coal seam, cycle through the processes of Step 1 to Step 5, export the distribution data of the bending moment and energy concentration of the hard roof, and generate a three-dimensional distribution diagram of the coupling of the bending moment and energy of the hard roof according to the exported data. Discriminate the distribution points of the bending moment and energy, and determine the highest point of the bending moment and energy distribution as the fracturing point.

[0050] Step 7: Establish a coupling control equation considering damage evolution and optimize the fracturing point data; as a rock material, the damage evolution of the hard roof under fracturing is different from that of other materials. Considering damage evolution can make the fracturing points more accurate, and the optimized control relationship of its damage evolution is as follows:

[0051]

[0052] In the formula, F is the damage factor of the hard roof, ρ is the density of the hard roof, and g is the acceleration of gravity. The above relational formula can optimize the fracturing points determined in Step 6 and normalize the fracturing points of the same layer of hard roof.

[0053] Step 8: Conduct fracturing and engineering evaluation on the determined fracturing points of the hard roof. After the fracturing points are implemented on site, through the mine tremor monitoring data, monitor the decrease in the frequency of mine tremors after fracturing. Especially, monitor the change in the intensity of mine tremors in the high-energy concentration area around the fracturing points.

[0054] In the embodiment of the present invention, taking the specific numerical values of a certain underground coal mine as an example, through the verification of the previous mining history, there are long-term mine tremor disasters in the working face. The strike length of a working face in this mine is 1500 m, the dip length is 250 m, and there is a 20-m-thick hard roof occurring 90 m above the coal seam. The hardness of this roof is 150 MPa, the elastic modulus is 30 GPa, and the Poisson's ratio is 0.3. Through the calculation of all the method steps of the present invention, the points of the coupling of the energy and bending moment of the hard roof are accurately identified, and the points are fractured, eliminating the mine tremor disasters caused by the fracture of the hard roof and saving the prevention and control costs of the mine tremor disasters caused by the fracture of the hard roof.

[0055] (1) Construction of a 3D spatial bending moment distribution model for rock bursts in hard roofs of coal mines; During the underground coal mining process, the coupled distribution of bending moment and energy is the main cause of rock stratum fracture and the formation of rock bursts. In 3D space, the distribution of bending moment and energy usually not only involves second-order differential equations but also needs to consider factors such as the inhomogeneity, anisotropy of the rock stratum, and stress-strain relationships. Therefore, more terms need to be added to the partial differential equation of the bending moment, specifically manifested as higher-order non-linear terms. The non-linear characteristics are complex manifestations of the distribution of bending moment and energy. Considering the above factors, the bending moment field of the hard roof satisfies the following differential element relationship:

[0056]

[0057] In the formula, M(x, y, z) is the distribution of the bending moment of the hard roof in 3D space; x, y, z are the 3D coordinates of the mine, and their specific meanings are the coordinate along the working face advancing direction, the coordinate perpendicular to the working face advancing direction, and the coordinate in the direction of the working face burial depth respectively; α, β, γ are constants related to the anisotropy of the material, reflecting the non-linear stress-strain relationship of the bending moment field.

[0058] In the formula, for rock materials, α = 0.15, β = 0.002, γ = 0.23.

[0059] (2) Determination of the non-linear coupling differential element relationship of the 3D energy distribution of the hard roof in coal mines; During the coal mining process, the release of energy from the fracture of the hard roof is the key factor causing rock bursts. For the calculation of energy, not only the second-order differential element relationship of the bending moment needs to be considered, but also the non-linear intensity differential element relationship of energy needs to be introduced, and this non-linear intensity differential element relationship considers the non-linear characteristics of the rock stratum. Considering the fracture and plastic deformation effects of the rock stratum, the energy intensity can be expressed as:

[0060]

[0061] In the formula, E(x, y, z) represents the energy intensity, u(x, y, z) is the displacement field of the hard roof, λ, μ, ν are the elastic constants of the material, where v is the Poisson's ratio, and R nonlinear (x, y, z) is the non-linear part, considering the effects such as stress softening and plasticity of the hard roof.

[0062]

[0063] The elastic constants of the hard roof material can be looked up in the material constant table, λ = 0.33, μ = 0.21, ν = 0.3; Considering the effects of stress softening of the hard roof material and the plasticity of the hard roof, it is measured that R nonlinear (x, y, z) = 0.358.

[0064] (3) Determine the coupled stress-bending moment microelement relationship of the hard roof in coal mines; the bending moment is the main factor for the hard roof in coal mines to break after coal seam mining. At the same time, the large-scale roof breakage caused by the coupled breakage of the bending moment and stress results in strong mine tremors. The coupling relationship between the bending moment field and the stress field in space can be described by a more complex elastic model, using a more accurate constitutive relationship:

[0065]

[0066] where σ x , σ y , σ z are the stresses in the x, y, and z directions respectively; u(x, y, z) is the three-dimensional displacement of the hard roof; D is the elastic modulus of the hard roof.

[0067] The three-dimensional displacement of the hard roof is:

[0068]

[0069] Taking the elastic modulus of the hard roof as 20 GPa and substituting the relevant parameters, the stresses are obtained as σ x = 25 MPa, σ y = 20 MPa, σ z = 30 MPa.

[0070] (4) Through the neural network optimization algorithm, establish a neural network optimization model to identify the bending moment and energy distribution characteristics of the hard roof; the optimization of the neural network model requires the introduction of a more complex objective function and more accurate training in combination with the backpropagation algorithm. The fracturing of the hard roof in coal mines needs to consider various factors, and the loss function of the neural network is designed in the following form:

[0071]

[0072] where L is the loss function, which is used to measure the sum of the error of the model prediction and the regularization constraint; are the loss accumulation terms respectively; is the predicted value; r i is the actual target value; N is the number of executions; f j is the variable output by the function or model, which is used when the input data of r i is the bending moment; g k is the variable output by the model, which is used when the input data of r iThis item is used when the input data is energy; i is the increment of the prediction times, increasing by 1 each time; j is the increment of the energy prediction, increasing by 1 each time; k is the increment of the bending moment prediction, increasing by 1 each time, t is the number of bending moment cycles; λ1 is the regularization coefficient, controlling the contribution degree of this item to the loss function; λ2 is the regularization coefficient, controlling the influence of this item on the loss function; s is the number of energy cycles.

[0073] λ1 is the regularization coefficient, controlling the contribution degree of this item to the loss function, taking 0.12; λ2 is the regularization coefficient, controlling the influence of this item on the loss function, taking 0.15, and both the bending moment and energy cycle 100 times, that is, N = 1000, s = 1000.

[0074] (5) The iterative form of the micro - element relation in the multi - loop solution process; due to considering the plastic and softening characteristics of the hard roof, there may be blank phenomena in the calculation process of step 4. Therefore, in each solution, the parameters can be updated through multi - loops to further optimize the model. In each iteration, we use the results of the current calculated bending moment field and energy field to update the prediction results of the fracturing points. The equation for each iteration can be expressed in the following form:

[0075] M (n+1) =M (n) +ΔM (n) ,E (n+1) =E (n) +ΔE (n)

[0076] M (n) and E (n) respectively represent the bending moment and energy at the n - th iteration; M (n+1) and E (n+1) respectively represent the bending moment and energy at the (n + 1) - th iteration; ΔM (n) and ΔE (n) respectively represent the increments of the bending moment and energy updated in this iteration.

[0077] After optimizing the bending moment and energy algorithms for the hard roof, the number of loop iterations in the calculation process of the hard roof is set to 1000, and the data after the loop is saved.

[0078] (6) Export the iterative data, and generate the three - dimensional coupling diagram of the bending moment and energy according to the neural network discrimination; according to the physical and mechanical parameters of the hard roof overlying the coal seam, loop through the processes from step 1 to step 5, export the distribution data of the bending moment and energy density of the hard roof, and generate the three - dimensional coupling diagram of the bending moment and energy of the hard roof according to the exported data. Discriminate the distribution points of the bending moment and energy, and determine that the highest point of the bending moment and energy distribution is the fracturing point.

[0079] Through the above calculations, the distribution diagram of the initial bending moment and energy coupling is exported as Figure 3 shown.

[0080] According to the figure, the coordinates of the initial fracturing sites derived for the first time are (150, 20, 100) and (150, 175, 105); the coordinates of the subsequent fracturing sites derived are (200, 20, 100), (200, 175, 105), (250, 20, 100), (250, 175, 105), (300, 20, 100), (300, 175, 105), (350, 20, 100), (350, 175, 105), (400, 20, 100), (400, 175, 105), (450, 20, 100), (450, 175, 105), (500, 20, 100), (500, 175, 105), (550, 20, 100), (550, 175, 105), (600, 20, 100), (600, 175, 105), (650, 20, 100), (650, 175, 105), (700, 20, 100), (700, 175, 105), (750, 20, 100), (750, 175, 105), (800, 20, 100), (800, 175, 105), (850, 20, 100), (850, 175, 105), (900, 20, 100), (900, 175, 105), (1000, 20, 100), (1000, 175, 105), (1050, 20, 100), (1050, 175, 105), (1100, 20, 100), (1100, 175, 105), (1150, 20, 100), (1150, 175, 105), (1200, 20, 100), (1200, 175, 105), (1250, 20, 100), (1250, 175, 105), (1300, 20, 100), (1300, 175, 105), (1350, 20, 100), (1350, 175, 105), (1400, 20, 100), (1400, 175, 105), (1450, 20, 100), (1450, 175, 105), (1500, 20, 100), (1500, 175, 105).

[0081] Due to the non-uniformity of the data in the z-direction and the existence of two maximum regions of bending moment and energy concentration in the y-direction, it is necessary to consider the damage evolution law during the fracturing of the hard roof and normalize the fracturing sites and fracturing parameters of the hard roof.

[0082] (7) Establish a coupled control equation considering damage evolution and optimize the fracturing point data; as a rock material, the damage evolution of the hard roof under fracturing is different from that of other materials. Considering damage evolution can make the fracturing points more accurate, and the optimized control relationship of its damage evolution is as follows:

[0083]

[0084] In the formula, F is the damage factor of the hard roof, ρ is the density of the hard roof, and g is the acceleration due to gravity. The above relational formula can optimize the fracturing points determined in (6) and normalize the fracturing points of the same layer of hard roof.

[0085] In the formula, the damage factor of the hard roof is 0.88, the density of the hard roof is 2.7×10 3 kg / m 3 , and the acceleration due to gravity is 10 m / s 2 . Through the above micro-element relationship, the fracturing points of the hard roof are normalized. The coordinate in the z direction is 102.5, the coordinate in the y direction is 125, and the coordinate in the x direction remains unchanged. Therefore, the derived fracturing point coordinates are: (150, 125, 102.5), (200, 125, 102.5), (250, 125, 102.5), (300, 125, 102.5), (350, 125, 102.5), (400, 125, 102.5), (450, 125, 102.5), (500, 125, 102.5), (550, 125, 102.5), (600, 125, 102.5), (650, 125, 102.5), (700, 125, 102.5), (750, 125, 102.5), (800, 125, 102.5), (850, 125, 102.5), (900, 125, 102.5), (950, 125, 102.5), (1000, 125, 102.5), (1050, 125, 102.5), (1100, 125, 102.5), (1150, 125, 102.5), (1200, 125, 102.5), (1250, 125, 102.5), (1300, 125, 102.5), (1350, 125, 102.5), (1400, 125, 102.5), (1450, 125, 102.5), (1500, 125, 102.5).

[0086] (8) Conduct fracturing and engineering evaluation on the determined hard roof fracturing sites. After the fracturing points are implemented on-site, through the mine seismic monitoring data, monitor the decline of the mine seismic frequency after fracturing. Especially the change of the mine seismic intensity in the high-energy density area around the fracturing points.

[0087] After the fracturing points are implemented on site, the post-fracturing microseismic situation is monitored through microseismic monitoring data, and the frequency of microseismic events decreases significantly. In particular, the microseismic intensity changes greatly in the high-energy density area around the fracturing points, and the microseismicity is reduced by 40%.

Claims

1. A three-dimensional energy-bending-moment coupling fracturing point determination method for hard roof mine earthquake prevention and control, characterized by: The following steps are involved: Step 1: Construct a three-dimensional spatial bending moment distribution model of hard roof earthquake in coal mines; the bending moment field of the hard roof satisfies the following microelement relationship: Where M(x, y, z) is the distribution of the bending moment of the hard roof in three-dimensional space; x, y, z are the three-dimensional coordinates of the mine, and their specific meanings are the coordinates along the advancing direction of the working face, the coordinates perpendicular to the advancing direction of the working face, and the coordinates in the buried depth direction of the working face; α, β, γ are constants related to the anisotropy of the material, reflecting the nonlinear stress-strain relationship of the bending moment field; Step 2: Determine the nonlinear coupled microelement relationship of the three-dimensional energy distribution of the hard roof of the coal mine; considering the fracture and plastic deformation effects of the rock formation, the energy concentration is expressed as: Where E(x,y,z) represents the energy concentration, u(x,y,z) is the displacement field of the rigid top plate, λ,μ,ν are the elastic constants of the material, v is the Poisson's ratio, and R nonlinear (x,y,z) is the nonlinear part, which takes into account the plasticity of stress softening and the hard top plate. Step 3: Determine the coupled stress-bending moment microelement relationship of the hard roof of the coal mine; In the formula, σ x ,σ y ,σ z is the stress in the x, y, and z directions; u(x, y, z) is the displacement field of the hard top plate; D is the elastic modulus of the hard top plate; Step 4: Establish a neural network optimization model through the neural network optimization algorithm to identify the bending moment and energy distribution characteristics of the hard roof; the loss function of the neural network is designed as follows: Where L is the loss function, which measures the sum of the model prediction error and the regularization constraint; are the accumulated loss items respectively; is the predicted value; r i is the actual target value; N is the number of executions; f j is the variable output by the function or model. i This item is used when the input data is bending moment; g k is the variable output by the model. i This item is used when the input data is energy; i is the increment of prediction times, which increases by 1 each time; j is the increment of energy concentration prediction, which increases by 1 each time; k is the increment of bending moment prediction, which increases by 1 each time, and t is the number of bending moment cycles; λ1 is the regularization coefficient, which controls the contribution of this item to the loss function; λ2 is the regularization coefficient, which controls the influence of this item on the loss function; s is the number of energy cycles; Step 5: Iterative form of the differential relationship in the multi-cycle solution process; in each iteration, the prediction results of the fracturing point are updated using the currently calculated bending moment field and energy field results; the equation of each iteration can be expressed as follows: M (n+1) =M (n) +ΔM (n) ,E (n+1) =E (n) +ΔE (n) M (n) and E (n) Respectively represent the bending moment and energy at the nth iteration; M (n+1) and E (n+1) Represent the bending moment and energy of the n+1th iteration respectively; ΔM (n) and ΔE (n) Respectively represent the bending moment and energy increment of this iterative update; Step 6: Export the iterative data and generate a three-dimensional diagram of bending moment and energy coupling based on the neural network judgment; According to the physical and mechanical parameters of the hard roof overlying the coal seam, the process from step 1 to step 5 is repeated to derive the bending moment and energy concentration distribution data of the hard roof, and a three-dimensional distribution diagram of the bending moment energy coupling of the hard roof is generated according to the derived data; Determine the distribution points of bending moment and energy, and determine the highest point of bending moment and energy distribution as the fracturing point.

2. The three-dimensional energy-bending-moment coupling fracturing point determination method for hard roof mine earthquake prevention and control according to claim 1 is characterized in that: It also includes step 7: establishing a coupling control equation considering damage evolution and optimizing the fracturing point data; the optimization control relationship of damage evolution is as follows: Wherein, F is the damage factor of the hard roof, ρ is the density of the hard roof, and g is the gravitational acceleration. The above relationship can optimize the fracturing points determined in step 6 and normalize the fracturing points of the same layer of hard roof.

3. The method for determining the three-dimensional energy-bending-moment coupling fracturing point for preventing and controlling hard roof mine earthquakes according to claim 2, characterized in that: It also includes step 8: fracturing and engineering evaluation of the determined hard roof fracturing points; after the fracturing points are implemented on site, monitoring the decrease in the frequency of mine earthquakes after fracturing through mine earthquake monitoring data; especially the changes in the intensity of mine earthquakes in the high energy concentration areas around the fracturing points.

Images