Coal mine pressure monitoring system and method based on deep learning

Abstract

The invention belongs to the technical field of coal mine pressure monitoring, and discloses a coal mine pressure monitoring system and method based on deep learning. The method comprises the following steps: establishing a coal mining geologic model, establishing a model of a three-zone geologic layer on a typical goaf, performing response calculation, and performing characteristic analysis on a coal mine pressure characteristic curve influenced by coal mine fracturing fluid resistivity and mining hole size in a three-zone profile on the goaf. And analyzing the characteristics of the coal mine pressure characteristic curve influenced by deformation and surrounding rock in the three-zone profile on the goaf. According to the method, coal mine pressure monitoring is performed in cooperative dynamic coupling monitoring of deformation processes of three zones (a bending sinking zone, a fissure zone and a caving zone) on the goaf, space-time complementarity and process collaboration of multiple means are utilized, dynamic monitoring of deformation under the mining hole of the coal mine goaf is realized, the monitoring precision is high, and the method is suitable for large-scale popularization and application. And technical support is provided for further coal mining.

Description

Technical Field

[0001] The present invention belongs to the technical field of coal mine pressure monitoring, and in particular relates to a coal mine pressure monitoring system and method based on deep learning. Background Art

[0002] In the process of coal mining, safe production of coal mines is very important; among them, the mine pressure monitoring technology of coal mines can effectively monitor the mine pressure in a timely manner, accurately grasp the law of mine pressure manifestation in the mining field, improve the mine pressure value of tunnel support, ensure the safe production of coal mines, and play an important role.

[0003] In the existing coal mining process, it is necessary to monitor the interior of the mine tunnel in real time through a mine pressure monitoring device. The conventional monitoring method uses an image camera and a pressure monitor to monitor the image and pressure inside the mine tunnel in real time, so as to ensure the safety of the mine tunnel during use. However, the image camera and the pressure monitor are easily affected by falling objects inside the mine tunnel during use. The impact of large falling objects can easily cause damage to the image camera and the pressure monitor, resulting in the image camera and the pressure detector needing regular maintenance and replacement, reducing the use value of the coal mine pressure monitoring device.

[0004] To solve this problem, an invention patent for a coal mine pressure monitoring device and method (publication number CN115575012A, publication date 2023.01.06) is disclosed, including a mounting column and a sleeve frame, a mounting rod is distributed in a ring on the top of the sleeve frame, and a protective component is provided on the outside of the multiple mounting rods, and the protective component includes two protective nets, an inner ring frame and an outer ring frame. By providing a protective component, during the use of the image capturer and the pressure monitor, they are protected by a double-layer protective net. When a small-volume falling object passes through the two layers of protective nets, its impact is weakened. When it reaches the top of the image capturer and the pressure monitor, its downward impact cannot cause any damage to them. When a large-volume falling object falls on the upper protective net, the protective net is squeezed and sunken, and the spring rod under the pressure-bearing fitting rod is passively compressed. The elastic potential energy of the spring rod's reset drives the upper protective net to bounce off the large-volume falling object, preventing it from accumulating on the protective net and causing damage to the protective net.

[0005] The technical defect of this invention is that it does not use relatively advanced deep learning technology to predict the coal mine pressure data in advance, and cannot eliminate hidden dangers at the starting stage.

[0006] Currently, safety issues have become a focal point and a challenge in coal mine production. Accidents caused by mine pressure primarily affect roadway wall stability, roof support load range, and hydraulic support stability. This is primarily due to the inability to accurately and timely collect, process, and provide early warnings for parameters such as internal roadway pressure, roadway wall pressure, and periodic pressure.

[0007] At present, the main method for preventing and controlling roof accidents in coal mines is mine pressure monitoring. Mine pressure monitoring has the following monitoring methods, including installing a comprehensive mining support pressure sensor on the hydraulic support of the comprehensive mining working face under the coal mine mining hole, installing a roof inner layer sensor on the roadway roof, installing a bearing stress sensor in the coal rock mass, and installing an anchor stress sensor on the roadway surrounding rock support anchor. The above monitoring methods are all independently arranged or manually read, so the degree of automation is low, data collection is inconvenient, and the processing is cumbersome, prone to errors, and the information feedback cycle is long. Therefore, it fails to fundamentally meet the requirements of timely guidance of production. To overcome the above shortcomings, there is also a mine pressure monitoring system. The main function of this mine pressure monitoring system is to collect and display mine pressure data, but it does not have a complete database, and cannot perform various statistical analyses and early warnings on the data. It cannot review historical data and reproduce historical status.

[0008] To solve this problem, a utility model patent for a coal mine pressure analysis system (Announcement No. CN205297648U, Publication Date 2016.06.08) discloses that the data processing subsystem above the mining hole receives the original data collected by the data acquisition subsystem below the mining hole through a communication module. The data is analyzed by the data processing subsystem above the mining hole and then transmitted to the pressure management subsystem above the mining hole through the communication module. The coal mine pressure analysis system provided by the utility model can collect more comprehensive monitoring data for pressure analysis through the data acquisition subsystem below the mining hole. In addition, the data processing subsystem above the mining hole and the pressure management subsystem above the mining hole adopt a C / S structure to realize real-time information monitoring, search and early warning of the fully mechanized mining working face, storage and historical query of information, and can also analyze and review historical data.

[0009] However, the technical defects of this patent are: the analysis of coal mine pressure data is based on the current commonly used data processing method, and the accuracy of determining the data evolution information of complex geology and the deformation process of the three zones above the goaf (curved sinking zone, fracture zone and caving zone) needs to be improved. Summary of the Invention

[0010] In order to overcome the problems existing in the related art, the embodiments disclosed in the present invention provide a coal mine pressure monitoring system and method based on deep learning.

[0011] The technical solution is as follows: A coal mine pressure monitoring method based on deep learning, comprising the following steps:

[0012] S1: Establish a coal mining geological model. Based on the multi-source parameter coupling method for coal mine pressure, complete the coal mine pressure lidar magnetic field decomposition, lidar wave field progression, lidar magnetic vector solution, coal mine mining geological layer conductivity calculation, and coal mine pressure characteristic curve synthesis;

[0013] S2, model establishment and response calculation of the three geological zones above the typical goaf, and analysis of the original coal mine pressure characteristic curves of the three zones above the goaf in the mine pressure sensing area for uniform strata, strata with mining holes, intangible strata, and tangible strata, to provide the original coal mine pressure characteristic curves for synthesis processing;

[0014] S3, in the upper three zones of the goaf, analyze the characteristics of the coal mine pressure characteristic curve affected by the resistivity of the coal mine fracturing fluid and the mining hole size; analyze different stratum thicknesses and different coal mine fracturing fluid resistivities, evaluate the pressure values ​​of the coal mine pressure characteristic curve in deep and shallow detection, and provide the coal mine pressure characteristic curve for stratum evaluation work;

[0015] S4, characteristic analysis of coal mine pressure characteristic curves affected by deformation and surrounding rock in the upper three zones of goaf profile; for coal mines with mining holes and invisible deformation, characteristic analysis of coal mine pressure characteristic curves in the upper three zones of goaf affected by surrounding rock pressure induction of mining geological models, and selection of synthetic coal mine pressure characteristic curves to provide a basis for coal and shale identification work.

[0016] In step S1, based on the coal mine pressure multi-source parameter coupling method, the coal mine pressure lidar magnetic field decomposition, lidar wave field progression, lidar magnetic subvector solution and coal mining geological layer conductivity calculation, and coal mine pressure characteristic curve synthesis are completed, including:

[0017] A coal mining geological model with M horizontal interfaces is established. Each layer is radially non-uniform and consists of mining holes, curved subsidence zones, fracture zones, caving zones, deformation zones, and original strata. (f, χ, h) are cylindrical coordinates, where f is the layer thickness, χ is the angular frequency, and h is the mining hole axis diameter.

[0018] The laser radar transmission line structure is located at the ground level h s On the meridian plane, use a point (r,h s ) represents, r is the radius, h s is the distance from the ground surface; the laser radar transmission line structure generates an axisymmetric coal mine pressure laser radar magnetic field; the laser radar transmission line structure passes the alternating current A T =A0d iat , A0 is the initial value of the alternating current, d iat Because the magnetic field of the coal mine pressure lidar contains a time factor, based on the classical differential form of the equation group, we get:

[0019]

[0020] d 2 =a 2 β(η-iγ / a)

[0021] Where, is the differential deviation, Q is the laser radar magnetic field power, a is the laser radar transmission coefficient, β is the magnetic permeability of the laser radar transmission, i is the time in seconds, d is the time factor, η is the dielectric constant, γ is the conductivity, F T is the emission current density;

[0022] According to the axial symmetry, Q is independent of χ. For the passive region, Q in the mth layer is χ =Q m , Q χ is the laser radar magnetic field power at angular frequency, Q m is the magnetic field power of the mth layer lidar, then:

[0023]

[0024] Where a m is the transmission coefficient of the mth layer laser radar;

[0025] The wave number is:

[0026]

[0027] Where, β m (f) is the magnetic permeability of the mth layer, η m (f) is the dielectric constant of the mth layer, γ m (f) is the electrical conductivity of the coal mining geological layer at layer m;

[0028] Solving by separation of variables method, we get:

[0029]

[0030] Where, e m (f) is the electrical conductivity of the coal mining geological stratum separation of the mth layer, is the separation distance between the upper and lower layers of the mth layer under the diameter of the mining hole axis, γ m is an N-order diagonal matrix;

[0031] After establishing the boundary conditions that the basis function should satisfy, the above formula is transposed and transformed, e m (f) Only numerical solutions are available to solve the generalized complex eigenvalues ​​and calculate the electrical conductivity of the geological layer in coal mining;

[0032] The resistivity of the uniform infinite coal mining geological model is set to vary from 0.01Ω.m to 200Ω.m. The characteristics of the original coal mine pressure characteristic curves obtained from different sub-mine pressure sets are observed, and the applicable conditions of the lidar analysis equipment for the upper three zones of the mine pressure-sensing goaf are obtained.

[0033] In step S2, the characteristic analysis of the original coal mine pressure characteristic curves of the upper three zones of the mine pressure sensing goaf is performed, including:

[0034] (1) Establish an infinite-thickness coal mining geological model considering the extraction holes, and compare the original response signals of multiple sub-mining pressure sets at different frequencies under different coal mine fracturing fluid conditions.

[0035] (2) Establish a multi-layer coal mining geological model considering the extraction holes and surrounding rocks, and compare the original response signals of multiple sub-mining pressure sets at different frequencies under different layer thickness and surrounding rock conditions.

[0036] (3) Establish a multi-layer coal mining geological model considering the extraction holes, surrounding rocks, and deformation, and compare the original response signals of multiple sub-mining pressure sets at different frequencies at different radial deformation depths under high-pressure conditions.

[0037] (4) Establish a multi-layer coal mining geological model considering the extraction holes, surrounding rocks, and deformation, and compare the original response signals of multiple sub-mining pressure sets at different frequencies at different radial deformation depths under low-pressure conditions.

[0038] In steps (1) - (4), the coal mining geological model is used to handle the problem of non-miscible multi-zones. When simulating and calculating the two-zone problem, the free surface position of the deformation zone under mining pressure and the volume of the deformation zone under mining pressure are represented by processing the deformation volume fraction equation under mining pressure.

[0039] If the deformation volume fraction of a certain zone under mining pressure in each control volume is A, then A = 0 indicates that the control volume does not contain this zone, A = 1 indicates that the control volume only contains this zone, and 0 < A < 1 indicates that there is an interface of two-zone deformations in the control volume.

[0040] The coal mining geological model is used to handle transient problems, and the expression of its deformation volume fraction equation under mining pressure is as follows:

[0041]

[0042] In the formula, c q is the density of the q-th zone, is the deformation volume fraction of the q-th zone under mining pressure, l q is the deformation frequency of the q-th zone, is the momentum source term, is the mining pressure value transferred from the p-th zone to the q-th zone, is the mining pressure value transferred from the q-th zone to the p-th zone, t is the time, is the differential deviation.

[0043] Furthermore, the deformation volume fraction equation under mining pressure is:

[0044]

[0045] The momentum equation of the coal mining geological model:

[0046]

[0047] Where c is density, v is deformation angle, w is phase frequency deformation coefficient, is the phase frequency differential, for The transposed matrix, l is the phase frequency, Δp is the pressure difference, g is the gravity frequency, l T is the phase frequency transposed matrix, W is the body force;

[0048] The energy equation of the coal mining geological model is:

[0049]

[0050] Where D is energy, p is pressure, ΔT is temperature, k is conductivity, Γ h is the energy source term.

[0051] Furthermore, in the coal mining geological model, the surface repulsion model is a continuous surface force model. When the surface repulsion is activated, the momentum equation adds a source term. When the surface repulsion is constant and only the interface normal force is considered, the pressure difference expression on both sides of the surface is:

[0052]

[0053] Where p2-p1 is the pressure difference on both sides of the surface, ζ is the surface repulsion coefficient; E1 and E2 are the curvature radii of the belt interface.

[0054] Furthermore, in the coal mining geological model, the deformation angle model between the belt surface and the deformation zone under the mine pressure is used in combination with the surface repulsion model. The deformation angle formed by the deformation zone under the mine pressure and the belt surface is used to adjust the belt interface normal near the belt surface, which leads to a change in the curvature of the belt interface near the belt surface.

[0055] Define v w is the deformation angle formed by the deformation band phase and the band surface under the mine pressure, and the normal expression of the unit band interface of adjacent band surfaces is:

[0056] X=X w cosv w +t w sinv w

[0057] Where X is the normal angle of the unit strip interface between adjacent strip faces, v w is the deformation angle formed by the deformation zone phase and the zone surface under the mine pressure, X w is the unit vector of the surface normal, t w is the unit vector tangent to the strip surface.

[0058] In step S3, the resistivity of different stratum thicknesses and different coal mine fracturing fluids is analyzed to evaluate the pressure values ​​of the coal mine pressure characteristic curve of deep and shallow detection, and provide the coal mine pressure characteristic curve for stratum evaluation, including:

[0059] (1) Establish a multi-layer coal mining geological model, set the mining hole radius, change the type of coal mine fracturing fluid from salt water coal mine fracturing fluid to clear water coal mine fracturing fluid, and change the stratum thickness from thin to thick, and determine the influence of the mining hole coal mine fracturing fluid on the mine pressure induction and the characteristics of the coal mine pressure characteristic curve;

[0060] (2) A uniform infinite coal mining geological model is established, and the characteristics of the coal mine pressure characteristic curves at different detection depths with the same resolution are determined under different coal mine fracturing fluid conditions when the mining hole changes.

[0061] In step S4, a synthetic coal mine pressure characteristic curve is selected to provide a basis for coal and shale identification, including:

[0062] (1) Establish a three-layer coal mining geological model with salt water slurry conditions, set the mining hole radius, and consider deformation under high-pressure surrounding rock conditions to determine the influence of mining hole fracturing fluid on the mine pressure induction synthetic coal mine pressure characteristic curve characteristics;

[0063] (2) Under the condition of freshwater slurry, the mining hole radius is set, and in the three-layer coal mining geological model considering deformation under low-pressure surrounding rock conditions, the characteristics of the coal mine pressure characteristic curve of the synthetic coal mine pressure induction are determined by the influence of the coal mine fracturing fluid in the mining hole.

[0064] Another object of the present invention is to provide a coal mine pressure monitoring system based on deep learning, which implements the coal mine pressure monitoring method based on deep learning. The system includes:

[0065] Establish a coal mining geological model module to establish a coal mining geological model. Based on the multi-source parameter coupling method of coal mine pressure, complete the coal mine pressure lidar magnetic field decomposition, lidar wave field progression, lidar magnetic vector solution, coal mining geological layer conductivity calculation, and coal mine pressure characteristic curve synthesis;

[0066] The model establishment and response calculation module of the three geological zones above the typical goaf is used for the model establishment and response calculation of the three geological zones above the typical goaf. It analyzes the characteristic curves of the original coal mine pressure in the three zones above the goaf in the mine pressure induction process for uniform strata, strata with mining holes, intangible strata, and tangible strata, and provides the original coal mine pressure characteristic curves for synthetic processing.

[0067] The curve characteristic analysis module is used to analyze the characteristics of the coal mine pressure characteristic curve affected by the resistivity of the coal mine fracturing fluid and the mining hole size in the upper three-zone profile of the goaf. It analyzes different stratum thicknesses and different coal mine fracturing fluid resistivities, evaluates the pressure values ​​of the coal mine pressure characteristic curve in deep and shallow detection, and provides the coal mine pressure characteristic curve for stratum evaluation work.

[0068] The synthetic coal mine pressure characteristic curve acquisition module is used for characteristic analysis of coal mine pressure characteristic curves affected by deformation and surrounding rock in the upper three zones of goaf profiles; for coal mines with mining holes and invisible deformation, characteristic analysis of coal mine pressure characteristic curves in the upper three zones of goaf affected by surrounding rock pressure in the mining geological model is performed, and synthetic coal mine pressure characteristic curves are selected to provide a basis for coal and shale identification.

[0069] Combined with all the above technical solutions, the beneficial effects of the present invention are as follows: the coal mine pressure monitoring system and method based on deep learning provided by the embodiment of the present invention performs coal mine pressure monitoring in the coordinated dynamic coupling monitoring of the deformation process of the three zones above the goaf (curved sinking zone, fracture zone and caving zone), and utilizes the spatiotemporal complementarity and process synergy of multiple means to realize dynamic monitoring of deformation under the mining hole in the coal mine goaf, with high monitoring accuracy, providing technical support for further coal mining. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure;

[0071] Figure 1 This is a flow chart of a coal mine pressure monitoring method based on deep learning provided by an embodiment of the present invention;

[0072] Figure 2 This is a characteristic effect diagram of the impact of the mining hole on different detection depth curves provided by the embodiment of the present invention;

[0073] Figure 3 Schematic diagram of a coal mine pressure monitoring system based on deep learning provided by an embodiment of the present invention;

[0074] In the figure: 1. Module for establishing a geological model for coal mining; 2. Module for establishing a model and calculating the response of the three geological layers above a typical goaf; 3. Module for analyzing curve characteristics; 4. Module for obtaining a synthetic coal mine pressure characteristic curve. DETAILED DESCRIPTION

[0075] To make the above-mentioned objects, features, and advantages of the present invention more readily apparent, specific embodiments of the present invention are described in detail below with reference to the accompanying drawings. The following description sets forth numerous specific details to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways than those described herein, and those skilled in the art may make similar modifications without departing from the scope of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0076] Example 1, as Figure 1 As shown, the coal mine pressure monitoring method based on deep learning provided by the embodiment of the present invention includes:

[0077] S1: Establish a coal mining geological model. Based on the multi-source parameter coupling method for coal mine pressure, complete the coal mine pressure lidar magnetic field decomposition, lidar wave field progression, lidar magnetic vector solution, coal mine mining geological layer conductivity calculation, and coal mine pressure characteristic curve synthesis;

[0078] S2, model establishment and response calculation of the three geological zones above the typical goaf, and analysis of the original coal mine pressure characteristic curves of the three zones above the goaf in the mine pressure sensing area for uniform strata, strata with mining holes, intangible strata, and tangible strata, to provide the original coal mine pressure characteristic curves for synthesis processing;

[0079] S3, in the upper three zones of the goaf, analyze the characteristics of the coal mine pressure characteristic curve affected by the resistivity of the coal mine fracturing fluid and the mining hole size; analyze different stratum thicknesses and different coal mine fracturing fluid resistivities, evaluate the pressure values ​​of the coal mine pressure characteristic curve in deep and shallow detection, and provide the coal mine pressure characteristic curve for stratum evaluation work;

[0080] S4, characteristic analysis of coal mine pressure characteristic curves affected by deformation and surrounding rock in the upper three zones of goaf profile; for coal mines with mining holes and invisible deformation, characteristic analysis of coal mine pressure characteristic curves in the upper three zones of goaf affected by surrounding rock pressure induction of mining geological models, and selection of synthetic coal mine pressure characteristic curves to provide a basis for coal and shale identification work.

[0081] For example, in step S1, a coal mining geological model is established. Based on the coal mine pressure multi-source parameter coupling method, the coal mine pressure lidar magnetic field decomposition, lidar wave field progression, lidar magnetic vector solution, coal mine geological layer conductivity calculation, and coal mine pressure characteristic curve synthesis are completed. The applicable scope of the lidar analysis equipment includes:

[0082] A coal mining geological model with M horizontal interfaces is established. Each layer is radially non-uniform and consists of mining holes, curved subsidence zones, fracture zones, caving zones, deformation zones, and original strata. (f, χ, h) are cylindrical coordinates, where f is the layer thickness, χ is the angular frequency, and h is the mining hole axis diameter.

[0083] The laser radar transmission line structure is located at the ground level h s On the meridian plane, use a point (r,h s ) represents, r is the radius, h s is the distance from the ground surface; the laser radar transmission line structure generates an axisymmetric coal mine pressure laser radar magnetic field; the laser radar transmission line structure passes the alternating current A T =A0d iat , A0 is the initial value of the alternating current, d iat Because the magnetic field of the coal mine pressure lidar contains a time factor, based on the classical differential form of the equation group, we get:

[0084]

[0085] d 2 =a 2 β(η-iγ / a)

[0086] Where, is the differential deviation, Q is the laser radar magnetic field power, a is the laser radar transmission coefficient, β is the magnetic permeability of the laser radar transmission, i is the time in seconds, d is the time factor, η is the dielectric constant, γ is the conductivity, F T is the emission current density;

[0087] According to the axial symmetry, Q is independent of χ. For the passive region, Q in the mth layer is χ =Q m , Q χ is the laser radar magnetic field power at angular frequency, Q m is the magnetic field power of the mth layer lidar, then:

[0088]

[0089] Where a m is the transmission coefficient of the mth layer laser radar;

[0090] The wave number is:

[0091]

[0092] Where, β m (f) is the magnetic permeability of the mth layer, η m (f) is the dielectric constant of the mth layer, γ m (f) is the electrical conductivity of the coal mining geological layer at layer m;

[0093] Solving by separation of variables method, we get:

[0094]

[0095] where, e m (f) is the conductivity of the geological layer separation for coal mine mining in the m-th layer, is the separation distance between the upper and lower layers of the m-th layer under the diameter of the mining hole axis, γ m is an N-order diagonal matrix;

[0096] After establishing the boundary conditions that the basis function should satisfy, perform a transpose transformation on the above formula, e m (f) has only a numerical solution. Solve the generalized complex eigenvalue to calculate the conductivity of the geological layer for coal mine mining;

[0097] Set the resistivity of the homogeneous infinite coal mine mining geological model to vary from \({{0.01}}\Omega\cdot m\) to \({{200}}\Omega\cdot m\). Observe the characteristics of the original coal mine strata pressure characteristic curve obtained from different sub-mining pressure sets, and obtain the applicable conditions of the lidar analysis equipment for the three zones above the mined-out area with strata pressure induction.

[0098] In step S2, analyze the characteristics of the original coal mine strata pressure characteristic curve for the three zones above the mined-out area with strata pressure induction, including:

[0099] (1) Establish an infinitely thick coal mine mining geological model considering the mining hole. Under different coal mine fracturing fluid conditions, compare the original response signals of multiple sub-mining pressure sets at different frequencies;

[0100] (2) Establish a multi-layer coal mine mining geological model considering the mining hole and surrounding rock. Under different surrounding rock thickness conditions, compare the original response signals of multiple sub-mining pressure sets at different frequencies;

[0101] (3) Establish a multi-layer coal mine mining geological model considering the mining hole, surrounding rock, and deformation. Under high-pressure conditions, compare the original response signals of multiple sub-mining pressure sets at different frequencies at different radial deformation depths;

[0102] (4) Establish a multi-layer coal mine mining geological model considering the mining hole, surrounding rock, and deformation. Under low-pressure conditions, compare the original response signals of multiple sub-mining pressure sets at different frequencies at different radial deformation depths.

[0103] Exemplarily, the coal mine mining geological model is used to handle the problem of immiscible multi-zones. When simulating and calculating the two-zone problem, the free surface position of the deformation zone under strata pressure and the volume of the deformation zone under strata pressure are represented by processing the deformation volume fraction equation under strata pressure. The basic idea is that if the deformation volume fraction of a certain zone under strata pressure in each control volume is A, then A = 0 means that the control volume does not contain this zone, A = 1 means that the control volume only contains this zone, and 0 < A < 1 means that there is an interface of two-zone deformation in the control volume;

[0104] The coal mine mining geological model has the advantages of less memory occupation, less calculation amount, and high efficiency. The coal mine mining geological model is used to handle transient problems, and the expression of its deformation volume fraction equation under strata pressure is as follows:

[0105]

[0106] Where c q is the qth band density, is the volume fraction of the qth zone under pressure, l q is the qth band deformation frequency, is the momentum source term, is the rock pressure value transmitted from the pth zone to the qth zone, is the pressure value of the mine pressure transmitted from the qth zone to the pth zone, t is the time, is the differential deviation.

[0107] The equation for the volume fraction of deformation under mine pressure is:

[0108]

[0109] Momentum equation for coal mining geological model:

[0110]

[0111] Where c is density, v is deformation angle, w is phase frequency deformation coefficient, is the phase frequency differential, for The transposed matrix, l is the phase frequency, Δp is the pressure difference, g is the gravity frequency, l T is the phase frequency transposed matrix, W is the body force;

[0112] The energy equation of the coal mining geological model is:

[0113]

[0114] Where D is energy, p is pressure, ΔT is temperature, k is conductivity, Γ h is the energy source term.

[0115] For example, to accurately represent the two-belt dynamics, coal mining geological models consider both the surface repulsion effect between the two-belt interface and the influence of the deformation angle between the belt surface and the deformation zone under the mine pressure. The surface repulsion can be specified as a constant. The governing equation includes an additional shear stress term due to the variation in the surface repulsion coefficient, which, when solved, yields the interaction term. Under conditions of high or low mine pressure, the surface repulsion effect often has a significant impact on the two-belt dynamics.

[0116] In the coal mining geological model, the surface repulsion model is a continuous surface force model. When the surface repulsion is activated, the momentum equation needs to add a source term. When the surface repulsion is constant and only the interface normal force is considered, the pressure difference on both sides of the surface is expressed as follows:

[0117]

[0118] Where p2-p1 is the pressure difference on both sides of the surface, ξ is the surface repulsion coefficient; E1 and E2 are the curvature radii of the belt interface.

[0119] Deformation angle model between the belt surface and the deformation zone under mine pressure.

[0120] In the coal mining geological model, the deformation angle model between the belt surface and the deformation zone under mine pressure is used in combination with the surface repulsion model, and the deformation angle formed by the deformation zone under mine pressure and the belt surface is used to adjust the belt interface normal near the belt surface, which leads to a certain change in the curvature of the belt interface near the belt surface.

[0121] Define v w is the deformation angle formed by the deformation band phase and the band surface under the mine pressure, and the normal expression of the unit band interface of adjacent band surfaces is:

[0122] X=X w cosv w +t w sinv w

[0123] Where X is the normal angle of the unit strip interface between adjacent strip faces, v w is the deformation angle formed by the deformation zone phase and the zone surface under the mine pressure, X w is the unit vector of the surface normal, t w is the unit vector tangent to the plane

[0124] For example, in step S3, different stratum thicknesses and different coal mine fracturing fluid resistivities are analyzed to evaluate the pressure values ​​of the coal mine pressure characteristic curve of deep and shallow detection, and provide the coal mine pressure characteristic curve for stratum evaluation, including:

[0125] (1) Establish a multi-layer coal mining geological model, set the mining hole radius, change the type of coal mine fracturing fluid from salt water coal mine fracturing fluid to clear water coal mine fracturing fluid, and change the stratum thickness from thin to thick, and determine the influence of the mining hole coal mine fracturing fluid on the mine pressure induction and the characteristics of the coal mine pressure characteristic curve;

[0126] (2) A uniform infinite coal mining geological model is established, and the characteristics of the coal mine pressure characteristic curves at different detection depths with the same resolution are determined under different coal mine fracturing fluid conditions when the mining hole changes.

[0127] For example, in step S4, a synthetic coal mine pressure characteristic curve is selected to provide a basis for coal and shale identification, including:

[0128] (1) Establish a three-layer coal mining geological model with salt water slurry conditions, set the mining hole radius, and consider deformation under high-pressure surrounding rock conditions to determine the influence of mining hole fracturing fluid on the mine pressure induction synthetic coal mine pressure characteristic curve characteristics;

[0129] (2) Under the condition of freshwater slurry, the mining hole radius is set, and in the three-layer coal mining geological model considering deformation under low-pressure surrounding rock conditions, the characteristics of the coal mine pressure characteristic curve of the synthetic coal mine pressure induction are determined by the influence of the coal mine fracturing fluid in the mining hole.

[0130] The present invention adopts the multi-source parameter coupling method of coal mine pressure to establish a two-dimensional coal mining geological model, conducts coal mine pressure laser radar magnetic field decomposition, laser radar wave field progression, laser radar magnetic vector solution and coal mine mining geological layer conductivity calculation, and systematically analyzes the following main environmental factors that affect the response of the three zones above the mine pressure-induced goaf: (1) formation resistivity; (2) coal mine fracturing fluid resistivity; (3) coal mine fracturing fluid deformation; (4) mining hole size; (5) formation thickness.

[0131] A uniform infinite-scale coal mining geological model was established to analyze the response characteristics of the three zones above the goaf in response to pressure induction in strata with different conductive properties, determine the stratigraphic conditions affected by the trend effect, and standardize the instrument's applicable scope. A coal mining geological model with a mining hole and infinite thickness was established to determine the variation characteristics of the coal mine pressure characteristic curve affected by the mining hole and the resistivity of the coal mine fracturing fluid. The original signals at multiple transmission frequencies of the instrument were compared to analyze the variation characteristics of the coal mine pressure characteristic curve affected by the trend effect and identify the type of original coal mine pressure characteristic curve with good pressure values. A coal mining geological model was established in which the deformation radius of the coal mine fracturing fluid changes from shallow to deep. The deformation relationship of the coal mine fracturing fluid can be divided into high-pressure and low-pressure states. The variation characteristics of the synthetic coal mine pressure characteristic curve were analyzed and synthetic coal mine pressure characteristic curves with good pressure values ​​were classified, laying the foundation for deformation zone inversion. A geological model of coal mining with different layer thicknesses is established, and the deformation of coal fracturing fluid is considered to be divided into high pressure and low pressure conditions. The pressure values ​​of the original coal mine pressure characteristic curve of each sub-pressure set are analyzed to avoid the intersection of coal mine pressure characteristic curves with different source distances.

[0132] The present invention utilizes a multi-source parameter coupling method for coal mine pressure to systematically analyze the main environmental factors affecting the pressure characteristic curves of the upper three zones of the pressure-induced goaf. Different coal mining geological models are designed for complex strata and mining hole conditions and simulated calculations are performed. Through coal mine pressure laser radar magnetic field decomposition, laser radar wave field progression, laser radar magnetic vector solution, coal mining geological layer conductivity calculation, and resolution matching, the pressure characteristic curves of the upper three zones of the pressure-induced goaf under different coal mining geological models are obtained. Furthermore, the characteristic patterns of the pressure characteristic curves under the influence of a single factor among coal mine fracturing fluid resistivity, coal mine fracturing fluid deformation, mining hole size, stratum thickness, and stratum resistivity are analyzed and summarized. The present invention can improve the accuracy of the interpretation results of the pressure characteristic curves of the upper three zones of the pressure-induced goaf, providing a theoretical basis for stratum evaluation and coal seam identification.

[0133] The size of the production hole is also an important factor affecting the quality of the curve. The following is a formation model and its synthetic processing results when the production hole changes. Figure 2 The influence of a 12-inch production hole on the 120-inch detection depth curve (M1RX) and the 10-inch detection depth curve (M1R1) is shown. The horizontal axis is the true resistivity Rt of the formation, and the vertical axis is the ratio of the synthetic apparent resistivity to the true resistivity, Ra / Rt.

[0134] The following conclusions can be drawn:

[0135] (1) The larger the production hole, the more unfavorable the measurement results are. In particular, large production holes with salt water fracturing fluid conditions are extremely unfavorable for HDIL measurement, which significantly reduces the application range of HDIL instruments.

[0136] (2) The magnitude of Rt / Rm (i.e., the resistivity contrast between the formation and the fracturing fluid) has a significant impact on the measurement results. If the production hole is enlarged, the impact is more significant.

[0137] (3) The deep detection curve has a wider range of applications.

[0138] Example 2, as Figure 3 As shown, the coal mine pressure monitoring system based on deep learning provided by the embodiment of the present invention includes:

[0139] Establishing a coal mining geological model module 1, which is used to establish a coal mining geological model. Based on the multi-source parameter coupling method of coal mine pressure, it completes the coal mine pressure lidar magnetic field decomposition, lidar wave field progression, lidar magnetic vector solution, coal mining geological layer conductivity calculation, and coal mine pressure characteristic curve synthesis;

[0140] Model establishment and response calculation module 2 of the three upper geological zones of a typical goaf is used for model establishment and response calculation of the three upper geological zones of a typical goaf. It analyzes the characteristic curves of the original coal mine pressure in the three upper zones of the goaf in response to the mine pressure, targeting uniform strata, strata with mining holes, intangible strata, and tangible strata, and provides the original coal mine pressure characteristic curves for synthesis processing;

[0141] Curve characteristic analysis module 3 is used to analyze the characteristics of the coal mine pressure characteristic curve affected by the resistivity of the coal mine fracturing fluid and the mining hole size in the upper three-zone profile of the goaf; analyze different stratum thicknesses and different coal mine fracturing fluid resistivities, evaluate the pressure values ​​of the coal mine pressure characteristic curve in deep and shallow detection, and provide the coal mine pressure characteristic curve for stratum evaluation work;

[0142] The synthetic coal mine pressure characteristic curve acquisition module 4 is used for characteristic analysis of coal mine pressure characteristic curves affected by deformation and surrounding rock in the upper three zones of the goaf; for coal mines with mining holes and no deformation, the characteristic analysis of coal mine pressure characteristic curves in the upper three zones of the goaf affected by the surrounding rock of the mining geological model is carried out, and the synthetic coal mine pressure characteristic curve is selected to provide a basis for coal and shale identification.

[0143] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant description of other embodiments.

[0144] The deep learning-based coal mine pressure monitoring system and method provided by the embodiment of the present invention were applied in the goaf of the 5-20303 working face of a coal mine, within a range of 350m wide and 1000m long. Coal mine pressure monitoring was carried out in the collaborative dynamic coupling monitoring of the deformation process of the three zones above the goaf (curved sinking zone, fracture zone and caving zone). The spatiotemporal complementarity and process synergy of multiple means were utilized to realize dynamic monitoring of deformation under the mining hole in the goaf of coal mines, with high monitoring accuracy, providing technical support for further coal mining.

[0145] The above description is only a preferred specific implementation method of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by any technician familiar with this technical field within the technical scope disclosed by the present invention and within the spirit and principles of the present invention should be covered by the scope of protection of the present invention.

Claims

1. A coal mine pressure monitoring method based on deep learning, characterized in that: The method comprises the following steps: S1: Establish a coal mining geological model. Based on the multi-source parameter coupling method for coal mine pressure, complete the coal mine pressure lidar magnetic field decomposition, lidar wave field progression, lidar magnetic vector solution, coal mine mining geological layer conductivity calculation, and coal mine pressure characteristic curve synthesis; S2, model establishment and response calculation of the three geological zones above the typical goaf, and analysis of the original coal mine pressure characteristic curves of the three zones above the goaf in the mine pressure sensing area for uniform strata, strata with mining holes, intangible strata, and tangible strata, to provide the original coal mine pressure characteristic curves for synthesis processing; S3, in the upper three zones of the goaf, analyze the characteristics of the coal mine pressure characteristic curve affected by the resistivity of the coal mine fracturing fluid and the mining hole size; analyze different stratum thicknesses and different coal mine fracturing fluid resistivities, evaluate the pressure values ​​of the coal mine pressure characteristic curve in deep and shallow detection, and provide the coal mine pressure characteristic curve for stratum evaluation work; S4, characteristic analysis of coal mine pressure characteristic curves affected by deformation and surrounding rock in the upper three zones of goaf profile; for coal mines with mining holes and invisible deformation, characteristic analysis of coal mine pressure characteristic curves in the upper three zones of goaf affected by surrounding rock pressure induction of mining geological models, and selection of synthetic coal mine pressure characteristic curves to provide a basis for coal and shale identification work.

2. The coal mine pressure monitoring method based on deep learning according to claim 1 is characterized in that: In step S1, based on the coal mine pressure multi-source parameter coupling method, the coal mine pressure lidar magnetic field decomposition, lidar wave field progression, lidar magnetic subvector solution and coal mining geological layer conductivity calculation, and coal mine pressure characteristic curve synthesis are completed, including: A coal mining geological model with M horizontal interfaces is established. Each layer is radially non-uniform and consists of mining holes, curved subsidence zones, fracture zones, caving zones, deformation zones, and original strata. (f, χ, h) are cylindrical coordinates, where f is the layer thickness, χ is the angular frequency, and h is the mining hole axis diameter. The laser radar transmission line structure is located at the ground level h s On the meridian plane, use a point (r,h s ) represents, r is the radius, h s is the distance from the ground surface; the laser radar transmission line structure generates an axisymmetric coal mine pressure laser radar magnetic field; the laser radar transmission line structure passes the alternating current A T =A0d iat , A0 is the initial value of the alternating current, d iat Because the magnetic field of the coal mine pressure lidar contains a time factor, based on the classical differential form of the equation group, we get: d 2 =a 2 β(η-iγ / a) Where, is the differential deviation, Q is the laser radar magnetic field power, a is the laser radar transmission coefficient, β is the magnetic permeability of the laser radar transmission, i is the time in seconds, d is the time factor, η is the dielectric constant, γ is the conductivity, F T is the emission current density; According to the axial symmetry, Q has nothing to do with x. For the passive region, Q in the mth layer χ =Q m , Q χ is the lidar magnetic field power at angular frequency, Q m is the magnetic field power of the mth layer lidar, then: Where a m is the transmission coefficient of the mth layer laser radar; The wave number is: Where, β m (f) is the magnetic permeability of the mth layer, η m (f) is the dielectric constant of the mth layer, γ m (f) is the electrical conductivity of the coal mining geological layer at layer m; Solving by separation of variables method, we get: Where, e m (f) is the electrical conductivity of the coal mining geological stratum separation of the mth layer, is the separation distance between the upper and lower layers of the mth layer under the diameter of the mining hole axis, γ m is an N-order diagonal matrix; After establishing the boundary conditions that the basis function should satisfy, the above formula is transposed and transformed, e m (f) Only numerical solutions are available to solve the generalized complex eigenvalues ​​and calculate the electrical conductivity of the geological layer in coal mining; The resistivity of the uniform infinite coal mining geological model is set to vary from 0.01Ω.m to 200Ω.m. The characteristics of the original coal mine pressure characteristic curves obtained from different sub-mine pressure sets are observed, and the applicable conditions of the lidar analysis equipment for the upper three zones of the mine pressure-sensing goaf are obtained.

3. The coal mine pressure monitoring method based on deep learning according to claim 1 is characterized in that: In step S2, the characteristic analysis of the original coal mine pressure characteristic curves of the upper three zones of the mine pressure sensing goaf is performed, including: (1) Establish an infinitely thick coal mining geological model considering the mining hole, and compare the original response signals of multiple sub-mine pressure sets with different frequencies under different coal mine fracturing fluid conditions; (2) Establish a multi-layer coal mining geological model that takes into account the mining hole and surrounding rock, and compare the original response signals of multiple sub-mine pressure sets with different frequencies under the conditions of different layer thicknesses of surrounding rock; (3) Establish a multi-layer coal mining geological model that takes into account the mining hole, surrounding rock, and deformation, and compare the original response signals of multiple sub-mine pressure sets of different frequencies at different radial deformation depths under high pressure conditions; (4) A multi-layer coal mining geological model is established that takes into account the mining hole, surrounding rock, and deformation. Under low pressure conditions, the original response signals of multiple sub-mine pressure sets of different frequencies at different radial deformation depths are compared.

4. The coal mine pressure monitoring method based on deep learning according to claim 3 is characterized in that: In steps (1)-(4), the coal mining geological model is used to handle the problem of immiscible multi-zones. When simulating and calculating the two-zone problem, the free surface position of the deformation zone under mine pressure and the volume of the deformation zone under mine pressure are represented by processing the deformation volume fraction equation under mine pressure; If the deformation volume fraction of a certain zone under mine pressure in each control volume is A, then A = 0 indicates that the control volume does not contain this zone, A = 1 indicates that the control volume only contains this zone, and 0 < A < 1 indicates that there is an interface of deformation between two zones in the control volume; The coal mining geological model is used to handle transient problems, and the expression of its deformation volume fraction equation under mine pressure is as follows: Where c q is the qth band density, is the volume fraction of the qth zone under pressure, l q is the qth band deformation frequency, is the momentum source term, is the rock pressure value transmitted from the pth zone to the qth zone, is the pressure value of the mine pressure transmitted from the qth zone to the pth zone, t is the time, is the differential deviation.

5. The coal mine pressure monitoring method based on deep learning according to claim 4 is characterized in that: The deformation volume fraction equation under mine pressure is: The momentum equation of the coal mining geological model: Where c is density, v is deformation angle, w is phase frequency deformation coefficient, is the phase frequency differential, for The transposed matrix, l is the phase frequency, Δp is the pressure difference, g is the gravity frequency, l T is the phase frequency transposed matrix, W is the body force; The energy equation of the coal mining geological model is: Where D is energy, p is pressure, ΔT is temperature, k is conductivity, Γ h is the energy source term.

6. The coal mine pressure monitoring method based on deep learning according to claim 5 is characterized in that: In the coal mining geological model, the surface repulsive force model is a continuous surface force model. When the surface repulsive force is activated, a source term is added to the momentum equation; when the surface repulsive force is a constant and only the normal force of the interface is considered, the expression of the pressure difference on both sides of the surface is: In the formula, p2 - p1 is the pressure difference on both sides of the surface, ξ is the surface repulsive force coefficient; E1 and E2 are the curvature radii of the zone interface.

7. The coal mine pressure monitoring method based on deep learning according to claim 5 is characterized in that: In the coal mining geological model, the deformation angle model between the zone surface and the deformation zone under mine pressure is used in combination with the surface repulsive force model. The normal direction of the zone interface near the zone surface is adjusted by the deformation angle formed by the deformation zone under mine pressure and the zone surface, which results in a change in the curvature of the zone interface near the zone surface; Define v w is the deformation angle formed by the deformation band phase and the band surface under the mine pressure, and the normal expression of the unit band interface of adjacent band surfaces is: X=X w cosv w +t w sinv w Where X is the normal angle of the unit strip interface between adjacent strip faces, v w is the deformation angle formed by the deformation zone phase and the zone surface under the mine pressure, X w is the unit vector of the surface normal, t w is the unit vector tangent to the strip surface.

8. The coal mine pressure monitoring method based on deep learning according to claim 1 is characterized in that: In step S3, the analysis of different formation thicknesses and different resistivities of coal mine fracturing fluids is carried out to evaluate the mine pressure values of the mine pressure characteristic curves for deep and shallow detection of coal mines, providing the mine pressure characteristic curves for formation evaluation work, including: (1) Establish a multi-layer coal mining geological model, set the radius of the mining hole, the type of coal mine fracturing fluid changes from brine coal mine fracturing fluid to fresh water coal mine fracturing fluid, and the formation thickness changes from thin to thick, and determine the characteristics of the synthetic coal mine pressure characteristic curves of mine pressure induction affected by the coal mine fracturing fluid in the mining hole; (2) Establish a uniformly infinite large coal mining geological model, and determine the characteristics of the coal mine pressure characteristic curves of different detection depths with the same resolution when the mining hole changes under different coal mine fracturing fluid conditions.

9. The coal mine pressure monitoring method based on deep learning according to claim 1, characterized in that: In step S4, the synthetic coal mine pressure characteristic curves providing a basis for coal and shale identification work are selected, including: (1) Establish a three-layer coal mining geological model considering deformation under high-pressure surrounding rock conditions under the condition of saltwater mud, set the radius of the mining hole, and determine the characteristics of the synthetic coal mine pressure characteristic curves of mine pressure induction affected by the coal mine fracturing fluid in the mining hole; (2) Establish a three-layer coal mining geological model considering deformation under low-pressure surrounding rock conditions under the condition of fresh water mud, set the radius of the mining hole, and determine the characteristics of the synthetic coal mine pressure characteristic curves of mine pressure induction affected by the coal mine fracturing fluid in the mining hole.

10. A coal mine pressure monitoring system based on deep learning, characterized in that: This system implements the deep learning-based coal mine pressure monitoring method described in any one of claims 1-9. This system includes: A module (1) for establishing a coal mining geological model, which is used to establish a coal mining geological model, and based on the multi-source parameter coupling method of coal mine pressure, complete the decomposition of the lidar magnetic field of coal mine pressure, the progression of the lidar wave field, the solution of the lidar magnetic vector, and the calculation of the conductivity of the coal mining geological layer and the synthesis of the coal mine pressure characteristic curves; The model establishment and response calculation module (2) of the three geological layers above the typical goaf is used for the model establishment and response calculation of the three geological layers above the typical goaf. The characteristic analysis of the original coal mine pressure characteristic curve of the three layers above the goaf in the mine pressure induction area is carried out for uniform strata, strata with mining holes, intangible strata, and tangible strata, providing the original coal mine pressure characteristic curve for synthesis processing. The curve characteristic analysis module (3) is used to analyze the characteristics of the coal mine pressure characteristic curve affected by the resistivity of the coal mine fracturing fluid and the mining hole size in the upper three-zone profile of the goaf; analyze different stratum thicknesses and different coal mine fracturing fluid resistivities, evaluate the pressure values ​​of the coal mine pressure characteristic curve in deep and shallow detection, and provide the coal mine pressure characteristic curve for stratum evaluation work; The synthetic coal mine pressure characteristic curve acquisition module (4) is used for characteristic analysis of coal mine pressure characteristic curves affected by deformation and surrounding rock in the upper three zones of the goaf; for coal mines with mining holes and no deformation, characteristic analysis of coal mine pressure characteristic curves affected by surrounding rock in the mining geological model is performed, and a synthetic coal mine pressure characteristic curve is selected to provide a basis for coal and shale identification.

Images