Method for monitoring disturbance stress of opposite stope face of deep coal mine

Abstract

The invention discloses a deep coal mine opposite stope face disturbance stress monitoring method, and relates to the technical field of data processing. The method comprises the following steps that S1, geomechanical parameters and mining conditions are preprocessed; s2, constructing a multi-dimensional monitoring network; s3, synchronously collecting and preprocessing multi-source data; s4, recovering the three-dimensional structure and solving the residual stress of the paleotectonic structure; s5, performing space-time coupling modeling on the opposite stoping disturbance stress field; step S6, microseismic-stress collaborative inversion correction is carried out; step S7, predicting a disturbance stress evolution trend; and step S8, dynamically generating a multi-level early warning threshold and iteratively optimizing the system. According to the method, the three-dimensional structure recovery technology and the elastic mechanics superposition principle are fused, the paleotectonic stress residual amount is incorporated into the opposite stoping stress calculation model, the fracture influence coefficient and the spatialization influence radius are corrected through micro-seismic-stress collaborative inversion, accurate quantification of disturbance stress under the deep complex geological condition is achieved, and data errors are reduced.

Description

Technical Field

[0001] This invention belongs to the field of data processing technology, and in particular relates to a method for monitoring disturbance stress in deep coal mine facing coal face. Background Technology

[0002] Coal, as a crucial energy source in my country, has consistently held a dominant position in the primary energy consumption structure. With the gradual depletion of shallow coal resources, deep coal mining has become an inevitable trend. my country possesses abundant deep coal resources; however, deep mining faces numerous severe challenges, such as complex geological conditions including high ground stress, high gas content, and high ground temperature, significantly increasing safety risks during the mining process. During deep coal mining, the changes in mining-induced stress are extremely complex, which can not only cause deformation and damage to the coal and rock mass but also potentially lead to severe dynamic disasters such as rock bursts and coal and gas outbursts.

[0003] Currently, the main methods for monitoring mining-induced stress in deep coal mines include borehole stress gauge monitoring, microseismic monitoring, and electromagnetic radiation monitoring. Borehole stress gauge monitoring directly measures stress changes in the coal and rock mass by installing stress gauges in the borehole. While this method can accurately obtain local stress information, its monitoring range is limited, and it can only perform point measurements, making it difficult to comprehensively reflect the spatial distribution characteristics of mining-induced stress. Microseismic monitoring utilizes the microseismic signals generated when the coal and rock mass fractures due to mining to infer stress changes. It can achieve monitoring over a larger area, but it is easily affected by geological conditions, noise, and other factors, affecting the accuracy and reliability of the monitoring results. Electromagnetic radiation monitoring is based on the principle that coal and rock masses generate electromagnetic radiation under stress. It indirectly reflects stress changes by monitoring the intensity of electromagnetic radiation. This method is relatively simple to operate, but it suffers from problems such as susceptibility to external electromagnetic interference and difficulty in achieving quantitative monitoring.

[0004] Existing monitoring methods generally suffer from insufficient accuracy, limited monitoring range, and poor adaptability to complex geological conditions, making it difficult to meet the needs of comprehensive, real-time, and accurate monitoring of mining-induced stress in deep coal mine facing faces. In complex deep geological environments, the distribution and variation patterns of mining-induced stress are even more intricate. Traditional monitoring methods cannot effectively capture minute stress changes and complex spatiotemporal evolution characteristics, thus failing to provide timely and accurate decision-making basis for coal mine safety production. Therefore, the development of a novel method for monitoring disturbance stress in deep coal mine facing faces is urgently needed. Summary of the Invention

[0005] The purpose of this invention is to provide a method for monitoring disturbance stress in deep coal mine face-to-face mining. By integrating three-dimensional structural restoration technology with the superposition principle of elasticity, the residual stress of ancient structures is incorporated into the stress calculation model of face-to-face mining. The method uses microseismic-stress co-inversion to correct the fracture influence coefficient and spatialized influence radius, thus solving the problems of insufficient accuracy, limited monitoring range, and poor adaptability to complex geological conditions in existing stress monitoring.

[0006] To solve the above-mentioned technical problems, the present invention is achieved through the following technical solution: This invention relates to a method for monitoring disturbance stress in a deep coal mine facing coal face, comprising the following steps: Step S1, Geomechanical Parameter and Mining Condition Preprocessing: Obtain basic geological data of the target mine and key parameters of the opposite mining face, establish a geomechanical parameter database, and construct an initial geological model using 3D modeling software; Step S2: Construct a multi-dimensional monitoring network: Fiber optic stress sensors are selected to monitor static stress and deployed in the coal seam and deep roof of the two roadways of the working face. Microseismic detectors are configured to capture rock fracture signals, and additional sensors are deployed around the periphery of the mining area. A series of detectors form a three-dimensional monitoring network; a mine pressure sensor is installed to monitor the working resistance of the support and record dynamic load changes in real time. Step S3, Multi-source data synchronous acquisition and preprocessing: Develop a synchronous acquisition protocol to achieve timestamp alignment of fiber optic stress sensor, micro-vibration detector and mine pressure sensor through industrial Ethernet, and preprocess the acquired data; Step S4, Three-dimensional structural restoration and determination of residual paleotectonic stress: Based on the three-dimensional geological model, the strata are flattened to the initial sedimentary level using the dynamic relaxation method, the strata deformation before and after flattening is extracted, and the strain state of the coal and rock mass is calculated using the Lagrange strain formula; Step S5, Spatiotemporal coupling modeling of stress field of opposite mining: Based on the relative positions of the two working faces, the stress superposition correction coefficient is derived, and combined with the superposition principle of elasticity, the stress formula of single working face mining is extended to the opposite mining scenario. Step S6, Microseismic-Stress Co-inversion Correction: Based on the calculation of rock mass fracture influence coefficient by microseismic energy, the stress field is corrected by coupling the microseismic influence with tectonic residual stress; Step S7, Prediction of the evolution trend of disturbance stress: Input the newly added residual stress and elastic parameters of the feature structure into the improved LSTM model to predict the stress value in the next 24 hours; Step S8, Dynamic Generation and Iterative Optimization of Multi-Level Early Warning Thresholds: Combining the uniaxial compressive strength of the coal body and the distribution of tectonic stress, a three-level early warning mechanism is set up, and the early warning duration is dynamically adjusted.

[0007] As a preferred technical solution, in step S1, the basic geological data includes: coal seam burial depth (H), coal seam thickness (h), and roof lithology (elastic modulus). Poisson's ratio ), base lithology (elastic modulus) Poisson's ratio ), original geostress field ( , The key parameters of the opposing mining faces include: working face length (rock density), fault and fracture distribution characteristics; (The text also mentions rock density, fault and fracture distribution characteristics, and key parameters of the opposing mining faces.) , ), propulsion speed ( , ), initial spacing ( ) and the dip angle of the coal seam being mined ( ).

[0008] As a preferred technical solution, the multi-dimensional monitoring network construction process in step S2 is as follows: Step S21: Screen and calibrate the fiber optic stress sensor, microseismic monitor, and mine pressure sensor respectively; Step S22: The fiber optic stress sensor is deployed in spatial layers, and the microseismic monitoring instrument is arranged in a three-dimensional array. Step S23: Optimize the dynamic density of monitoring points, when the real-time distance between the two working surfaces... When, the monitoring point encryption mechanism is triggered; where, This is the initial spacing. To accelerate, For the time of recovery; The formula for calculating the dynamic density of the monitoring points is as follows: ; In the formula, The distance between the two working surfaces is The density of monitoring points at that time For the initial density, For correction factor, The maximum stress gradient in the stress superposition region. The initial stress gradient is given.

[0009] As a preferred technical solution, in step S4, based on the relative positions of the two working surfaces, the formula for calculating the stress superposition correction coefficient is derived as follows: ; In the formula, The radius of influence of the working face, For the time of recovery; Based on the superposition principle of elasticity, the formula for stress in a single working face is extended to a scenario of opposing mining operations, and a coupled model is established as follows: In the formula, and These are the stress fields when the two working faces are mined separately. The original ground stress is represented by the negative sign, indicating the stress release area.

[0010] As a preferred technical solution, the specific process for three-dimensional structural reconstruction and determination of paleotectonic stress residue in step S4 is as follows: Step S41: Import the geological data collected in step S1 into the 3D modeling software, and use tetrahedral tetrahedral mesh to mesh the model; Specifically, this includes: coordinates of the interfaces between the coal seam and the roof and floor strata. Data was obtained through mine geological drilling, with one control point set every 50m, and the density increased to 20m in fault areas; the elastic modulus of the roof and floor ( Poisson's ratio of top plate and bottom plate () Density of top and bottom plates Clearly define parameters such as fault strike (e.g., NE30°), dip angle (65°), and elevation difference (15m), and mark the range of fracture development zone; use tetrahedral mesh (accuracy level: mesh size ≤5m in structurally complex areas, ≤10m in homogeneous areas) to ensure that the number of mesh elements is ≥500,000, meeting the accuracy requirements for stress calculation.

[0011] Step S42: Set the stress field at the initial moment (initial deposition state) to 0, retain only the gravity field (generated by the weight of the rock strata), apply a fixed constraint to the bottom layer to simulate the immobile basement condition; from the top layer to the target coal seam, apply a flattening displacement layer by layer according to the deposition sequence, and calculate the vertical difference between the current interface and the initial deposition interface of each layer based on the geological profile. By using the software displacement loading function, each interface layer is moved along... Axial movement This enables the back-pushing of strata. After each layer is flattened, the internal strain increment of the rock layer is calculated. If the convergence is not achieved, the displacement loading rate is adjusted and the iteration is restarted until all rock layers are flattened to the initial depositional level. The displacement cloud map of the stratigraphic nodes before and after flattening is then output. Step S43: In the flattened model, select one feature point every 20m along the strike (x-axis) and dip (y-axis) of the coal seam, for a total of 100 feature points, and record the coordinates of each feature point; compare the coordinates of the same feature point in the current structural state with the coordinates in the initial depositional state, and calculate the three-dimensional displacement vector using the following formula: In the formula, The coordinates are the current construction state coordinates. The coordinates are the initial state coordinates of the deposition. They are respectively Displacement components in the direction; Step S44: Based on the displacement vector of the feature point, calculate the displacement gradient between adjacent feature points, then calculate the deformation gradient tensor, and substitute it into the Lagrange strain formula to calculate the strain tensor. Step S45: Construct the characteristic equation, solve the principal strain characteristic equation for the plane strain tensor, and calculate the residual paleotectonic stress.

[0012] As a preferred technical solution, in step S44, any two adjacent points and The displacement gradient in the x-direction is , The distance between feature points; similarly, the calculation is performed. ; Build Deformation gradient tensor The calculation formula is: ; In the formula, 1 represents the initial unit length. For the first The displacement of a feature point in the x-direction. For the first The displacement of a feature point in the x-direction; For displacement components Rate of change along the x-direction.

[0013] As a preferred technical solution, the formula for calculating the residual ancient tectonic stress in step S45 is as follows: ; In the formula, This represents the maximum horizontal residual paleotectonic stress. This represents the minimum horizontal residual paleotectonic stress. For the maximum horizontal principal strain, For the minimum horizontal principal strain, The spatialized elastic modulus of coal and rock mass. This represents the spatialized Poisson's ratio of the coal and rock mass.

[0014] As a preferred technical solution, the specific process of microseismic-stress co-inversion correction in step S6 is as follows: Step S61: Extract the features of the microseismic events from the microseismic signal preprocessed in step S3, specifically including the source coordinates ( Vibration duration Peak amplitude and micro-vibration energy And screen the characteristics of microseismic events; Step S62: Collect historical microseismic data (containing at least 500 valid microseismic events) from the past three longwall faces of the target mine, establish an energy statistical distribution model, and dynamically adjust the threshold by introducing a geological condition correction term. The specific formula is as follows: ; In the formula, This is a geological correction factor (calibrated through field tests, with a value range of 0.3-0.8, and a larger value is taken for areas with complex structures). The maximum horizontal paleotectonic stress residual at this location (from step S4). This refers to the maximum value of the residual horizontal paleotectonic stress within the monitoring area.

[0015] Step S63: For microseismic events of different energy levels, design a segmented influence coefficient calculation model to calculate the influence coefficient of rock mass fracture. Perform layered calculations; Step S64: Establish a spatialized influence radius model based on the coal and rock mass integrity coefficient obtained from in-situ borehole acoustic testing; Step S65: Introduce synergistic correction of pore pressure and rock anisotropy to construct a three-dimensional coupled correction model.

[0016] As a preferred technical solution, in step S7, the total stress output in step S6 is collected, and the historical stress characteristics, working face dynamic characteristics, and geomechanical characteristics at the corresponding time are collected simultaneously; the LSTM model outputs the stress prediction value for the next 24 hours. .

[0017] As a preferred technical solution, in step S8, a three-level early warning mechanism is set as follows: when At that time, it was a Level 1 warning; when At that time, it was a Level II warning; when At that time, it was a Level 3 warning; In the formula, This represents the corrected disturbance stress value for the opposite mining face in a deep coal mine. The uniaxial compressive strength of the coal body. To determine the maximum horizontal residual paleotectonic stress, uniaxial compressive strength tests were conducted using the MTS rock mechanics testing system to obtain data for each coal sample. Value, Establish Spatial distribution model: ; In the formula, The average uniaxial compressive strength of the region. This is the strength correction factor. This is the distance from the location to the fault. The radius of influence of the fault.

[0018] The present invention has the following beneficial effects: (1) This invention integrates three-dimensional structural restoration technology with the superposition principle of elasticity, incorporates the residual stress of ancient structures into the stress calculation model of opposite mining, and uses microseismic-stress synergistic inversion to correct the rupture influence coefficient and spatial influence radius, thereby achieving accurate quantification of disturbance stress under deep and complex geological conditions and reducing data errors. (2) By introducing stress gradient ratio and relative distance, the present invention monitors the adaptive matching of point density and stress disturbance intensity in real time. When the two working surfaces are close and the stress gradient increases, the density increases exponentially, ensuring the data resolution of key areas.

[0019] (3) This invention optimizes through a dual mechanism of spatialized threshold and dynamic correction. On the one hand, it establishes a three-level basic early warning threshold based on spatial distribution and structural features. On the other hand, it introduces a stress change rate correction term and fits the critical spacing, combines LSTM prediction and Kalman filter correction, predicts the risk of stress exceeding the threshold 24 hours in advance, and generates a graded response strategy.

[0020] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0021] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a flowchart of a method for monitoring disturbance stress in a deep coal mine facing coal face according to the present invention. Detailed Implementation

[0023] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] Furthermore, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0025] To make the purpose, technical solution, and advantages of this application clearer, the following description is provided in conjunction with the appendix. Figure 1 The present application will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the application.

[0026] Please see Figure 1 As shown, this invention provides a method for monitoring disturbance stress in a deep coal mine facing coal face, comprising the following steps: Step S1, Geomechanical Parameter and Mining Condition Preprocessing: Obtain basic geological data of the target mine and key parameters of the opposite mining face, establish a geomechanical parameter database, and construct an initial geological model using 3D modeling software; Step S2: Construct a multi-dimensional monitoring network: Fiber optic stress sensors are selected to monitor static stress and deployed in the coal seam and deep roof of the two roadways of the working face. Microseismic detectors are configured to capture rock fracture signals, and additional sensors are deployed around the periphery of the mining area. A series of detectors form a three-dimensional monitoring network; a mine pressure sensor is installed to monitor the working resistance of the support and record dynamic load changes in real time. Step S3, Multi-source data synchronous acquisition and preprocessing: Develop a synchronous acquisition protocol to achieve timestamp alignment of fiber optic stress sensor, micro-vibration detector and mine pressure sensor through industrial Ethernet, and preprocess the acquired data; Step S4, Three-dimensional structural restoration and determination of residual paleotectonic stress: Based on the three-dimensional geological model, the strata are flattened to the initial sedimentary level using the dynamic relaxation method, the strata deformation before and after flattening is extracted, and the strain state of the coal and rock mass is calculated using the Lagrange strain formula; Step S5, Spatiotemporal coupling modeling of stress field of opposite mining: Based on the relative positions of the two working faces, the stress superposition correction coefficient is derived, and combined with the superposition principle of elasticity, the stress formula of single working face mining is extended to the opposite mining scenario. Step S6, Microseismic-Stress Co-inversion Correction: Based on the calculation of rock mass fracture influence coefficient by microseismic energy, the stress field is corrected by coupling the microseismic influence with tectonic residual stress; Step S7, Prediction of the evolution trend of disturbance stress: Input the newly added residual stress and elastic parameters of the feature structure into the improved LSTM model to predict the stress value in the next 24 hours; Step S8, Dynamic Generation and Iterative Optimization of Multi-Level Early Warning Thresholds: Combining the uniaxial compressive strength of the coal body and the distribution of tectonic stress, a three-level early warning mechanism is set up, and the early warning duration is dynamically adjusted.

[0027] In step S1, the basic geological data includes: coal seam burial depth (H), coal seam thickness (h), and roof lithology (elastic modulus). Poisson's ratio ), base lithology (elastic modulus) Poisson's ratio ), original geostress field ( , The key parameters of the opposing mining faces include: rock density, fault and fracture distribution characteristics; and the length of the working face (the following parameters are also important). , ), propulsion speed ( , ), initial spacing ( ) and the dip angle of the coal seam being mined ( ).

[0028] Specifically, the physical and mechanical parameters of the coal seam and rock strata are obtained through core drilling (one set of cores every 50m depth), and uniaxial compressive strength, elastic modulus, and Poisson's ratio are tested in the laboratory; the rock mass integrity coefficient is inverted using sonic logging technology (such as sonic full-wavelength logging). This helps to verify the representativeness of lithological parameters.

[0029] The original in-situ stress field was determined using either hydraulic fracturing (suitable for deep mines ≥500m) or stress relief methods, with 3-5 test points set up in the return airway and transport roadway to obtain the maximum horizontal principal stress. Minimum horizontal principal stress ( ) and vertical stress ( ), and establish a model of the distribution of the geostress field.

[0030] Fault and fracture characteristics were combined with geological drilling reports, underground tunnel data, and seismic exploration profiles to mark fault strike, dip angle, elevation difference, and fracture density (number of fractures per unit area). Fractal theory was used to quantify the degree of fracture development (fractal dimension). ).

[0031] Key parameters of the opposing longwall face are obtained from the mine production design drawings, including the face length. , ), initial spacing ( Horizontal distance between the two working face cuts), coal seam dip angle ( ); The propulsion speed is calibrated through on-site measurements (such as laser rangefinders). , (Unit: m / d), taking the average advance over 7 consecutive days as the baseline value.

[0032] The geological model construction process is as follows: Based on the above parameters, a three-dimensional geological model is built using FLAC3D or 3DEC. The model boundary conditions are set as follows: fixed constraint at the bottom, normal constraint around the perimeter, and an equivalent overlying rock load applied at the top. ); The coal seam, roof (subdivided into immediate roof and basic roof), and floor are grouped according to lithology and assigned physical and mechanical parameters. A weakening model is used for fault areas (reducing the elastic modulus to 30%-50% of the surrounding rock mass). Set the grid size of the working surface and the surrounding area (within a radius of 50m) to 1m×1m×1m, and gradually transition the outer area to 5m×5m×5m to balance calculation accuracy and efficiency.

[0033] The specific calculation formula is as follows: Vertical stress is generated by the weight of the overlying rock, as shown in the following formula: In the formula, The average density of the rock (taken as 25-27 kN / m³, determined by actual measurement through rock core sampling); The coal seam burial depth (m, obtained from borehole depth measurement data); In deep mines, horizontal stress is usually greater than vertical stress. An empirical formula is used to fit the maximum horizontal principal stress. ); In the formula, The horizontal stress coefficient is obtained through regression analysis of hydraulic fracturing test data; for deep mines, k is typically 1.2-2.5. It is the tectonic stress constant (MPa, reflecting the influence of regional tectonic activity, determined by statistics from the mining area geostress database).

[0034] The fractal dimension of fractures is calculated to quantify the degree of fracture development. The formula is as follows: In the formula, The number of cracks is the length of the crack. The crack length threshold, It is a proportionality constant. For fractal dimension, ( The larger the value, the more developed the fractures), through double logarithmic coordinates ( ) Calculation of the absolute value of the slope of the linear fitting.

[0035] The formula for the radius of influence of mining activities in a single working face is as follows: In the formula, h is the coal seam thickness (m). This is the Protodyakonov coefficient.

[0036] In step S2, the multi-dimensional monitoring network construction process is as follows: Step S21: Screen and calibrate the fiber optic stress sensor, microseismic monitor, and mine pressure sensor respectively; The specific deployment method is as follows: During the screening of fiber optic grating stress sensors (FBGs), considering the high stress (≥60MPa), high humidity (relative humidity ≥95%), and strong electromagnetic interference environment in deep coal mines, FBG sensors with a center wavelength of 1550nm±5nm were selected. Their stress sensitivity (≥1.2pm / MPa) and temperature cross-sensitivity (≤0.5pm / ℃) were verified through high-temperature and high-pressure calibration experiments (simulating a 3000m burial depth environment), ensuring their effectiveness. Measurement error within the range is ≤1%; the selected fiber optic stress sensor is armored with 316L stainless steel (2mm thickness) and has built-in polyimide-coated optical fiber, improving its impact resistance to 1500g (exceeding the original requirement of 1000g, suitable for blasting impact at the working face); an 8-channel microseismic monitoring instrument (sampling rate 1MHz, dynamic range 120dB) is selected, equipped with a three-component detector (natural frequency 10Hz, sensitivity 100mV / g), and verified by vibration table experiments: Signal-to-noise ratio (SNR) in the z-band Ensure the capture of tiny rock fractures (energy) The mine pressure sensor calibration uses a column pressure sensor (range). (Accuracy ±0.5%FS) Before installation, perform 5-point calibration (0, 15, 30, 45, 60MPa) using a hydraulic calibration bench, and fit the pressure-voltage conversion curve (linearity). This eliminates nonlinear errors.

[0037] Step S22: The fiber optic stress sensor is deployed in spatial layers, and the microseismic monitoring instrument is arranged in a three-dimensional array. Specifically, when deploying fiber optic stress sensors in a spatially layered manner, horizontally, boreholes (50mm in diameter) are drilled along the advancing direction in the coal seam of the transport roadway and return airway of the opposing working faces, at depths of 5m (shallow), 10m (medium), and 15m (deep), to implant FBG sensors, forming a "shallow-medium-deep" three-layer monitoring system to capture the stress transmission patterns at different depths of the coal seam. Vertically, boreholes are drilled in the roof (2m and 5m from the coal seam roof) and floor (2m and 5m from the coal seam floor) at the working face cut and stop line, and FBG sensors are implanted to monitor the stress response of the roof and floor rock masses. The microseismic monitoring instrument is centered on the "recovery unit" formed by the two working faces, and its periphery... Within the specified range, eight geophones are deployed according to the vertices of a regular hexahedron (increased to twelve if the working face length > 150m). Four of these are located in rock tunnels 10m above the coal seam roof, and four are located in rock tunnels 10m below the coal seam floor, forming a three-dimensional spatial positioning network to ensure that the seismic source positioning error is ≤ 5m. One mine pressure sensor is installed at each of the front and rear columns of the hydraulic support and the shield beam of the working face, with a sampling frequency of 1Hz, to record changes in support resistance in real time and reflect the dynamic load on the working face.

[0038] Step S23: Optimize the dynamic density of monitoring points, when the real-time distance between the two working surfaces... When, the monitoring point encryption mechanism is triggered; where, This is the initial spacing. To accelerate, For the time of recovery; The formula for calculating the dynamic density of monitoring points is as follows: ; In the formula, The distance between the two working surfaces is The density of monitoring points at that time For the initial density ( At that time, take 0.1 units / m, that is, a spacing of 10m). The correction coefficient (based on fitting of actual mine measurement data, with values ​​ranging from...) ), The maximum stress gradient (MPa / m) in the stress superposition zone is calculated through numerical simulation. The initial stress gradient (MPa / m, the gradient value of the original geostress field).

[0039] Before encryption ( ), and deploy one group (including shallow, medium, and deep FBG sensors) every 10m along the direction of propulsion; after densification ( The spacing between the sensors was reduced to 5m, and a temporary monitoring point (only shallow 5m) was added between the two sets of sensors to capture stress abrupt changes. Based on the geological model identified in step S1, dangerous areas such as faults (drop > 5m) and fracture zones (density > 3 fractures / m) were identified. Within their influence range (20m to both sides of the fault strike, 30m in the direction of fracture zone extension), additional sensors were deployed. One FBG sensor (10m depth) and two microseismic detectors form a dual guarantee for core monitoring and redundant verification.

[0040] The specific implementation is as follows: The FBG sensor uses armored packaging and has a stress range. Temperature correction factor ( The microseismic monitoring instrument uses 48 channels, with a detector natural frequency of 10Hz and a positioning error ≤3m; the mine pressure sensor has a range of... The linearity R² = 0.9995; three layers (5m, 10m, and 15m deep) of FBG sensors are arranged in each of the W1 and W2 roadways, with an initial spacing of 10m, for a total of 60 sensors; eight microseismic monitoring instruments (four in the roof rock roadway and four in the floor rock roadway) are set up 30m outside the working face, with 50 supports in each of W1 and W2, each with three sensors, for a total of 300 sensors.

[0041] For example, on the tenth day of the re-mining, This triggers encryption. The density is calculated using the formula: 30 new FBG sensors were added, and 5 additional FBG sensors and 2 detectors were installed in the fault-affected zone (20m on both sides).

[0042] In step S3, temperature drift correction is performed on the FBG data: ( This is a temperature correction factor. (This refers to the measured temperature difference). Noise filtering of the microseismic signal is performed using wavelet thresholding denoising: ,in, For wavelet transform, For threshold function; Remove abnormal values ​​from the mine pressure sensor (such as instantaneous pulses caused by mechanical vibration) and retain the effective load data.

[0043] In step S4, the specific process for three-dimensional structural reconstruction and determination of paleotectonic stress residuals is as follows: Step S41: Import the geological data collected in step S1 into the 3D modeling software, and use tetrahedral tetrahedral mesh to mesh the model; Specifically, this includes: coordinates of the interfaces between the coal seam and the roof and floor strata. : Obtained through mine geological drilling data, with one control point set every 50m, and the density increased to 20m in fault areas; elastic modulus of the roof and floor ( Poisson's ratio of top plate and bottom plate () Density of top and bottom plates Clearly define parameters such as fault strike (e.g., NE30°), dip angle (65°), and elevation difference (15m), and mark the range of fracture development zone; use tetrahedral mesh (accuracy level: mesh size ≤5m in structurally complex areas, ≤10m in homogeneous areas) to ensure that the number of mesh elements is ≥500,000, meeting the accuracy requirements for stress calculation.

[0044] Step S42: Set the stress field at the initial moment (initial deposition state) to 0, retain only the gravity field (generated by the weight of the rock strata), apply a fixed constraint to the bottom layer to simulate the immobile basement condition; from the top layer to the target coal seam, apply a flattening displacement layer by layer according to the deposition sequence, and calculate the vertical difference between the current interface and the initial deposition interface of each layer based on the geological profile. By using the software displacement loading function, each interface layer is moved along... Axial movement This enables the back-pushing of strata. After each layer is flattened, the internal strain increment of the rock layer is calculated. If the convergence is not achieved, the displacement loading rate is adjusted and the iteration is restarted until all rock layers are flattened to the initial depositional level. The displacement cloud map of the stratigraphic nodes before and after flattening is then output. Step S43: In the flattened model, select one feature point every 20m along the strike (x-axis) and dip (y-axis) of the coal seam, for a total of 100 feature points, and record the coordinates of each feature point; compare the coordinates of the same feature point in the current structural state with the coordinates in the initial depositional state, and calculate the three-dimensional displacement vector using the following formula: In the formula, The coordinates are the current construction state coordinates. The coordinates are the initial state coordinates of the deposition. They are respectively Displacement components in the direction; Step S44: Based on the displacement vector of the feature point, calculate the displacement gradient between adjacent feature points, then calculate the deformation gradient tensor, and substitute it into the Lagrange strain formula to calculate the strain tensor. Specifically, for any two adjacent points and The displacement gradient in the x-direction is , The distance between feature points; similarly, the calculation is performed. ; Build Deformation gradient tensor: In the formula, 1 represents the initial unit length; Substituting into the Lagrange strain formula yields the strain tensor. : In the formula, for The transpose of the matrix, for The identity matrix; Step S45: Construct the characteristic equation, solve the principal strain characteristic equation for the plane strain tensor, and calculate the residual paleotectonic stress. Specifically, for plane strain tensor Solving for the principal strain requires satisfying the characteristic equation: In the formula, Candidate values ​​for principal strain; The discriminant is calculated as follows: Substituting these values ​​into the quadratic formula yields the maximum and minimum horizontal principal strains, as shown in the following formulas: ; The direction of the principal strain force passes through Calculate the principal strain direction angle This ensures that subsequent stress calculations are consistent with the actual structural stress direction (e.g., the direction of maximum principal strain is parallel to the fault strike). The relationship between strain and stress is established using the generalized Hooke's law, as shown in the following formula: ; In the formula, , The residual normal stresses of paleotectonic stress are in the x and y directions, respectively. for In-plane paleotectonic stress and residual shear stress , These are the engineering normal strains in the x and y directions calculated in step S44, respectively. The xy-direction engineering shear strain is calculated in step S44. The spatial elastic modulus of coal and rock mass. Poisson's ratio for the spatialized coal and rock mass; Since step S45 has already obtained the maximum horizontal principal strain through strain tensor decomposition, and minimum horizontal principal strain (There is no shear strain in the principal strain direction, only normal strain), therefore, the above formula needs to be simplified to the correspondence between principal strain and principal stress, and finally the formula for calculating the residual stress of paleotectonic structures is obtained: ; In the formula, This represents the maximum horizontal residual paleotectonic stress. This represents the minimum horizontal residual paleotectonic stress. For the maximum horizontal principal strain, Minimum horizontal principal strain; through spatialized mechanical parameters and Instead of traditional homogenization parameters, it ensures that stress calculations dynamically change with geological conditions (such as faults and fissures), which is consistent with the actual working conditions of deep coal mines; it also ensures the reliability of the formula output, providing accurate input for the stress field coupling modeling in the subsequent step S5.

[0045] In specific implementation, the basic conditions of the mine are as follows: coal seam burial depth Coal seam thickness Coal seam dip angle (Near-horizontal coal seam); Geological structure as follows: There is one NE-trending normal fault (strike 35°, dip 60°, elevation drop 12m), the affected area of ​​the fault is 50m on both sides of the fault; Coal and rock mass parameters as follows: (25 GPa near the fault, 32 GPa away from the fault). .

[0046] The formation was leveled using a dynamic relaxation method, and a relaxation coefficient was set. After 120 iterations, the algorithm converged. 100 feature points were selected, including 30 points in the fault-affected area, and the maximum displacement vector was measured. Among them, the x-direction is stretching, the y-direction is compressing, and the z-direction is sinking; Calculate the strain tensor at a characteristic point in a homogeneous region: ; Discriminant Then the main strain: ; ; The residual ancient tectonic stress is calculated as follows: ; ; In step S5, based on the relative positions of the two working surfaces, the formula for calculating the stress superposition correction coefficient is derived as follows: ; In the formula, The radius of influence of the working face, For the time of recovery; Based on the superposition principle of elasticity, the formula for stress in a single working face is extended to a scenario of opposing mining operations, and a coupled model is established as follows: In the formula, and These are the stress fields when the two working faces are mined separately. The original ground stress is represented by the negative sign, indicating the stress release area.

[0047] Specifically, based on the Coulomb-Mohr strength criterion, the stress field distribution caused by a single longwall face is calculated. , The formula is as follows: Solve Similarly; In the formula, The coordinates of the monitoring points (with the initial position of the working face as the origin) (The axis is the direction of propulsion). The advancing speed of working face 1, For the recovery time, The horizontal distance between the monitoring point and the current position of working face 1. The dip angle of the coal seam. Used to correct the effect of tilt angle on horizontal stress. This represents the original geostress.

[0048] The initial distance between the two working surfaces is (Horizontal distance along the advancing direction), real-time spacing as the mining progresses. It changes dynamically over time, when A decrease indicates that the two working surfaces are closer together, and the stress superposition effect is enhanced. This indicates that the two working surfaces meet, and the superposition effect reaches its peak.

[0049] The traditional superposition principle assumes that stresses are directly and linearly added together. However, in practice, when the distance between the two working surfaces is different, the superposition effect is non-linear (the superposition is enhanced when the distance is small, and negligible when the distance is too large). Therefore, Must meet: when (The two working faces are far apart): (Approximately single working face independent action); when (Middle position): The local peak value is reached (significant superposition effect). when (About to meet): It tends to a stable value (avoiding meaningless infinite increases).

[0050] Constructing a coupled form of the exponential function and the sine function: ; In the formula, The radius of influence of the working face, Ensure that the superposition effect decays to zero when the spacing is too large. exist When the time is positive, it reflects the superposition enhancement when the spacing decreases. At that time (Peak point).

[0051] After correcting the stress field of a single working face using a superposition factor, and subtracting the original ground stress (reflecting the stress increment or release caused by mining), the coupled stress field is obtained: ; when This indicates that stress concentration has occurred at this point due to mining activity; when This indicates that stress release has occurred at this point due to mining activities (such as the stress relief zone behind the working face).

[0052] In practice, when a deep coal mine has opposite mining faces, Initial spacing The propulsion speed is Original geostress (burial depth) Coal seam dip angle (Approximately horizontal); the coordinates of the monitoring point are ( (Located on the centerline of the advancing direction of the two working faces, 50m away from the initial position of working face 1).

[0053] When the advance distance of working face 1 is The horizontal distance between the monitoring point and working face 1 is ; The advancing distance of working face 2 is: The horizontal distance from the monitoring point is ; .

[0054] Real-time spacing is Superposition coefficient for ; The coupled stress field is: .

[0055] In step S6, the specific process of microseismic-stress co-inversion correction is as follows: Step S61: Extract the features of the microseismic events from the microseismic signal preprocessed in step S3, specifically including the source coordinates ( Vibration duration Peak amplitude and micro-vibration energy And screen the characteristics of microseismic events; Specifically, establish effectiveness screening criteria: Excluding mechanical interference events: Vibration duration And peak amplitude The signals (mostly vibrations from equipment such as coal mining machines and conveyors); Excluding long-distance interference events: distance from the epicenter to the monitoring area ( Events with a microseismic influence radius; Effective microseismic events are classified by energy: Low-energy events: (Development of micro-fractures in coal and rock mass); Medium-energy events: (Medium-scale fracturing of coal and rock mass); High-energy events: (Large-scale shear or tensile failure of coal and rock mass).

[0056] Step S62: Collect historical microseismic data (containing at least 500 valid microseismic events) from the past three longwall faces of the target mine, establish an energy statistical distribution model, and dynamically adjust the threshold by introducing a geological condition correction term. The specific formula is as follows: ; In the formula, This is a geological correction factor (calibrated through field tests, with a value range of 0.3-0.8, and a larger value is taken for areas with complex structures). The maximum horizontal paleotectonic stress residual at this location (from step S4). This refers to the maximum value of the residual horizontal paleotectonic stress within the monitoring area.

[0057] Step S63: For microseismic events of different energy levels, design a segmented influence coefficient calculation model to calculate the influence coefficient of rock mass fracture. Layered calculations are performed to avoid the problems of traditional single formulas being insufficiently sensitive to low-energy events and over-amplifying high-energy events. Specifically, for low-energy events ( A linear growth model is used to highlight the slight influence of microcracks on local stress. ; Medium energy events ( A tangent function model is used to match the stress disturbance law of moderate fracture:

[0058] High-energy events ( A saturated growth model is used to avoid unbounded increases in coefficients due to high-energy events. .

[0059] Step S64: Establish a spatialized influence radius model based on the coal and rock mass integrity coefficient obtained from in-situ borehole acoustic testing; Specifically, traditional fixed value Unable to adapt to different geological conditions, based on the coal and rock mass integrity coefficient (Acquired from on-site borehole acoustic testing) A spatialized influence radius model is established, with the following formula: ; In the formula, For integrity correction (value range 0.4-1.2), The fracture zone is taken as 1.2. The complete region is taken as 0.4). The integrity coefficient of the coal and rock mass at this location ( , For the longitudinal wave velocity of the rock, (This refers to the longitudinal wave velocity of the coal body).

[0060] Step S65: Introduce synergistic correction of pore pressure and rock anisotropy to construct a three-dimensional coupled correction model; The specific formula for the three-dimensional coupling correction model is as follows: ; In the formula, This represents the final perturbation stress field after co-correction of microseismic and pore pressure. This represents the initial total disturbance stress field after correction for rock strata anisotropy. For monitoring points To the epicenter The Euclidean distance is calculated using the following formula: ; The weighting of pore pressure influence (calibrated through field tests, with a value range of 0.1-0.4, and a larger value in the high pore pressure region). The pore pressure at that location, This refers to the maximum pore pressure within the monitoring area.

[0061] In specific implementation, step S62: Calculate the microseismic rupture energy threshold at a point in the fault-affected zone: ; Step S63, a certain energy microseismic event ( Influence coefficient: ; .

[0062] In step S64, the radius of influence of the microseismic event at this point is: .

[0063] In step S65, the distance from the monitoring point to the seismic source is... Corrected stress: .

[0064] In step S7, the total stress output in step S6 is collected, and the historical stress characteristics, dynamic characteristics of the working face, and geomechanical characteristics at the corresponding time are collected simultaneously; the LSTM model outputs the stress prediction values ​​for the next 24 hours. .

[0065] Specifically, historical stress characteristics include (Stress values ​​from the first 3 hours, reflecting the temporal correlation of stress); Dynamic characteristics of the working face include real-time spacing and advance speed; Geomechanical characteristics include the maximum horizontal residual paleotectonic stress. Young's modulus of coal and rock mass Output the predicted stress values ​​for the next 24 hours. .

[0066] The LSTM model structure is designed as follows: The input layer is designed with a feature grouping and embedding structure: the eight input features are divided into three groups according to their physical meaning (historical stress group, mining dynamics group, and geomechanical group). Each group is mapped through an independent fully connected layer (with ReLU activation function), as shown in the formula: ; In the formula, , The weight matrix and bias vector for the corresponding group ensure that features with different physical meanings are extracted in a targeted manner.

[0067] A dual LSTM unit is used. The first LSTM unit (64 nodes) processes historical stress features with strong temporal correlation and outputs a temporal feature vector. The second LSTM unit (32 nodes) incorporates attention-weighted data. Combined with other feature groups, the final time-series features are output. The structure is the same as the first LSTM unit (parameters independent); a Dropout layer (probability 0.2) is added between the two LSTM units, as shown in the formula: In the formula, To discard the probability and avoid model overfitting, [the following is used]. The mapping is to a 24-dimensional output (corresponding to stress in the next 24 hours), and the activation function is Linear (because stress prediction is a regression task). The formula is: In the formula, The weight matrix for the fully connected layer is (32×24). This is the bias vector.

[0068] The loss function uses "weighted mean square error (WMSE)" for the high stress range ( The prediction error is given a higher weight, as shown in the formula: In the formula, For the sample size, For the output dimension, when hour, In other cases, The maximum number of training rounds is 100. If the validation set loss does not decrease for 10 consecutive rounds or the test set prediction error is ≤5%, training will be terminated early.

[0069] In step S8, a three-level early warning mechanism is set up as follows: when When the situation is critical, it is classified as a Level 1 warning (attention); the response measures are to generate a stress change report every 2 hours and notify the work face foreman to strengthen on-site inspections. when At that time, it is a Level II warning (alert); the response measures are to generate a report every hour, activate the microseismic monitoring encryption mode (sampling rate increased to 2MHz), and adjust the working face advance speed (reduced by 20%). when At this time, it is a Level 3 warning (danger); the response measures are to push the warning information to the mine manager and chief engineer in real time, suspend mining at the working face, organize personnel to evacuate to a safe area, and activate the emergency support plan.

[0070] In the formula, This represents the corrected disturbance stress value for the opposite mining face in a deep coal mine. The uniaxial compressive strength of the coal body. To determine the maximum horizontal residual paleotectonic stress, uniaxial compressive strength tests were conducted using the MTS rock mechanics testing system to obtain data for each coal sample. Value, Establish Spatial distribution model: ; In the formula, The average uniaxial compressive strength of the region. This is the strength correction factor. This is the distance from the location to the fault. The radius of influence of the fault.

[0071] It is worth noting that the various units included in the above system embodiments are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of each functional unit are only for easy differentiation and are not used to limit the scope of protection of the present invention.

[0072] Furthermore, those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware, and the corresponding program can be stored in a computer-readable storage medium.

[0073] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for monitoring disturbance stress in a deep coal mine facing coal face, characterized in that, Includes the following steps: Step S1, Geomechanical Parameter and Mining Condition Preprocessing: Obtain basic geological data of the target mine and key parameters of the opposite mining face, establish a geomechanical parameter database, and construct an initial geological model using 3D modeling software; Step S2: Construct a multi-dimensional monitoring network: Fiber optic stress sensors are selected to monitor static stress and deployed in the coal seam and deep roof of the two roadways of the working face. Microseismic detectors are configured to capture rock fracture signals, and additional sensors are deployed around the periphery of the mining area. A series of detectors form a three-dimensional monitoring network; a mine pressure sensor is installed to monitor the working resistance of the support and record dynamic load changes in real time. Step S3, Multi-source data synchronous acquisition and preprocessing: Develop a synchronous acquisition protocol to achieve timestamp alignment of fiber optic stress sensor, micro-vibration detector and mine pressure sensor through industrial Ethernet, and preprocess the acquired data; Step S4, Three-dimensional structural restoration and determination of residual paleotectonic stress: Based on the three-dimensional geological model, the strata are flattened to the initial sedimentary level using the dynamic relaxation method, the strata deformation before and after flattening is extracted, and the strain state of the coal and rock mass is calculated using the Lagrange strain formula; Step S5, Spatiotemporal coupling modeling of stress field of opposite mining: Based on the relative positions of the two working faces, the stress superposition correction coefficient is derived, and combined with the superposition principle of elasticity, the stress formula of single working face mining is extended to the opposite mining scenario. Step S6, Microseismic-Stress Co-inversion Correction: Based on the calculation of rock mass fracture influence coefficient by microseismic energy, the stress field is corrected by coupling the microseismic influence with tectonic residual stress; Step S7, Prediction of the evolution trend of disturbance stress: Input the newly added residual stress and elastic parameters of the feature structure into the improved LSTM model to predict the stress value in the next 24 hours; Step S8, Dynamic Generation and Iterative Optimization of Multi-Level Early Warning Thresholds: Combining the uniaxial compressive strength of the coal body and the distribution of tectonic stress, a three-level early warning mechanism is set up, and the early warning duration is dynamically adjusted.

2. The method for monitoring disturbance stress in a deep coal mine facing coal face according to claim 1, characterized in that, In step S1, the basic geological data includes: coal seam burial depth, coal seam thickness, roof lithology, floor lithology, original geostress field, and fault and fracture distribution characteristics; the key parameters of the opposing mining faces include: working face length, advance speed, initial spacing, and mining coal seam dip angle.

3. The method for monitoring disturbance stress in a deep coal mine facing coal face according to claim 1, characterized in that, In step S2, the multi-dimensional monitoring network construction process is as follows: Step S21: Screen and calibrate the fiber optic stress sensor, microseismic monitor, and mine pressure sensor respectively; Step S22: The fiber optic stress sensor is deployed in spatial layers, and the microseismic monitoring instrument is arranged in a three-dimensional array. Step S23: Optimize the dynamic density of monitoring points, when the real-time distance between the two working surfaces... When, the monitoring point encryption mechanism is triggered; where, This is the initial spacing. To accelerate, For the time of recovery; The formula for calculating the dynamic density of the monitoring points is as follows: ； In the formula, The distance between the two working surfaces is The density of monitoring points at that time For the initial density, For correction factor, The maximum stress gradient in the stress superposition region. The initial stress gradient is given.

4. The method for monitoring disturbance stress in a deep coal mine facing coal face according to claim 1, characterized in that, In step S4, the specific process for three-dimensional structural reconstruction and determination of paleotectonic stress residuals is as follows: Step S41: Import the geological data collected in step S1 into the 3D modeling software, and use tetrahedral tetrahedral mesh to mesh the model; Step S42: Set the initial stress field to 0, retain only the gravity field, apply a fixed constraint to the bottom layer, and apply flattening displacement layer by layer from the top layer to the target coal seam in the depositional order. After each layer is flattened, calculate the internal strain increment of the rock layer. If it does not converge, adjust the displacement loading rate and iterate again until all rock layers are flattened to the initial depositional level. Output the displacement cloud map of the strata nodes before and after flattening. Step S43: In the flattened model, select one feature point every 20m along the strike and dip of the coal seam, for a total of 100 feature points. Compare the coordinates of the same feature point in the current structural state with the coordinates in the initial depositional state, and calculate the three-dimensional displacement vector using the following formula: In the formula, The coordinates are the current construction state coordinates. The coordinates are the initial state coordinates of the deposition. They are respectively Displacement components in the direction; Step S44: Based on the displacement vector of the feature point, calculate the displacement gradient between adjacent feature points, then calculate the deformation gradient tensor, and substitute it into the Lagrange strain formula to calculate the strain tensor. Step S45: Construct the characteristic equation, solve the principal strain characteristic equation for the plane strain tensor, and calculate the residual paleotectonic stress.

5. A method for monitoring disturbance stress in a deep coal mine facing coal face according to claim 4, characterized in that, In step S44, any two adjacent points and The displacement gradient in the x-direction is , The distance between feature points; similarly, the calculation is performed. ; Build Deformation gradient tensor The calculation formula is: In the formula, 1 represents the initial unit length. For the first The displacement of a feature point in the x-direction. For the first The displacement of a feature point in the x-direction; For displacement components Rate of change along the x-direction; Substituting into the Lagrange strain formula yields the strain tensor. : In the formula, for The transpose of the matrix, for The identity matrix.

6. The method for monitoring disturbance stress in a deep coal mine facing coal face according to claim 4, characterized in that, In step S45, the formula for calculating the residual paleotectonic stress is as follows: ； In the formula, This represents the maximum horizontal residual paleotectonic stress. The minimum horizontal residual paleotectonic stress, For the maximum horizontal principal strain, For the minimum horizontal principal strain, The spatialized elastic modulus of coal and rock mass. This represents the spatialized Poisson's ratio of the coal and rock mass.

7. The method for monitoring disturbance stress in a deep coal mine facing coal face according to claim 1, characterized in that, In step S5, based on the relative positions of the two working surfaces, the formula for calculating the stress superposition correction coefficient is derived as follows: ； In the formula, The radius of influence of the working face, For the time of recovery; Based on the superposition principle of elasticity, the formula for stress in a single working face is extended to a scenario of opposing mining operations, and a coupled model is established as follows: In the formula, and These are the stress fields when the two working faces are mined separately. The original ground stress is represented by the negative sign, indicating the stress release area.

8. A method for monitoring disturbance stress in a deep coal mine facing coal face according to claim 1, characterized in that, In step S6, the specific process of microseismic-stress co-inversion correction is as follows: Step S61: Extract the features of the microseismic events from the microseismic signal preprocessed in step S3, specifically including source coordinates, vibration duration, peak amplitude and microseismic energy, and filter the features of the microseismic events; Step S62: Collect historical microseismic data from the past three mining faces of the target mine, establish an energy statistical distribution model, and introduce a geological condition correction term to dynamically adjust the threshold. Step S63: For microseismic events of different energy levels, design a segmented influence coefficient calculation model and perform layered calculation of the rock mass fracture influence coefficient; Step S64: Establish a spatialized influence radius model based on the coal and rock mass integrity coefficient obtained from in-situ borehole acoustic testing; Step S65: Introduce synergistic correction of pore pressure and rock anisotropy to construct a three-dimensional coupled correction model.

9. A method for monitoring disturbance stress in a deep coal mine facing coal face according to claim 1, characterized in that, In step S7, the total stress output in step S6 is collected, and the historical stress characteristics, working face dynamic characteristics, and geomechanical characteristics at the corresponding time are collected simultaneously; the LSTM model outputs the stress prediction values ​​for the next 24 hours. .

10. A method for monitoring disturbance stress in a deep coal mine facing coal face according to claim 1, characterized in that, In step S8, a three-level early warning mechanism is set as follows: when At that time, it was a Level 1 warning; when At that time, it was a Level II warning; when At that time, it was a Level 3 warning; In the formula, This represents the corrected disturbance stress value for the opposite mining face in a deep coal mine. The uniaxial compressive strength of the coal body. To determine the maximum horizontal residual paleotectonic stress, uniaxial compressive strength tests were conducted using the MTS rock mechanics testing system to obtain data for each coal sample. Value, Establish Spatial distribution model: ； In the formula, The average uniaxial compressive strength of the region. This is the strength correction factor. This is the distance from the location to the fault. The radius of influence of the fault.

Images