Large dip angle stope coal rock dynamic disaster monitoring and early warning method and system based on digital twinning

Abstract

The invention provides a large-dip-angle stope coal rock dynamic disaster monitoring and early warning method and system based on digital twinning, and the method comprises the steps: obtaining sensor installation parameters and large-dip-angle stope geological environment information, building a correction model, and correcting large-dip-angle stope geological exploration data collected by a sensor; according to the corrected parameters, constructing a digital twinborn model of the large-dip-angle stope; based on historical coal and rock disaster case data and expert experience, constructing an early warning index system of the coal and rock dynamic disaster; constructing a coal rock dynamic disaster early warning model based on a machine learning algorithm; the coal rock dynamic disaster early warning model is integrated into the digital twin model, and when the digital twin model is updated, the coal rock dynamic disaster early warning model carries out real-time evaluation and early warning on the large-dip-angle stope dynamic disaster risk. According to the method, the occurrence risk of coal rock dynamic disasters is effectively reduced, the life safety of workers in a stope and the normal operation of equipment are guaranteed, and the safety and economic benefits of coal mining are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of disaster early warning for steeply inclined stope, and particularly relates to a method and system for monitoring and early warning of coal and rock dynamic disasters in a steeply inclined stope based on digital twin. Background Art

[0002] With the continuous increase of coal mining depth and intensity, the occurrence frequency and harm degree of coal and rock dynamic disasters (such as coal and gas outburst, rock burst, etc.) in steeply inclined stope are becoming increasingly serious. The traditional monitoring and early warning methods have the following defects:

[0003] 1. The monitoring data is scattered and not intuitive, making it difficult to comprehensively and accurately reflect the actual state of coal and rock mass in the stope.

[0004] 2. The accuracy of the early warning model is not high enough to meet the effective monitoring and early warning requirements of coal and rock dynamic disasters in the complex environment of steeply inclined stope.

[0005] 3. Lack of real-time interaction ability, unable to dynamically adjust the early warning results according to real-time data. Summary of the Invention

[0006] The present invention provides a method for monitoring and early warning of coal and rock dynamic disasters in a steeply inclined stope based on digital twin, realizing the monitoring and early warning of coal and rock dynamic disasters in a steeply inclined stope.

[0007] A method for monitoring and early warning of coal and rock dynamic disasters in a steeply inclined stope based on digital twin includes:

[0008] Obtaining sensor installation parameters and geological environment information of the steeply inclined stope, establishing a calibration model, and calibrating the geological exploration data of the steeply inclined stope collected by the sensor;

[0009] Constructing a digital twin model of the steeply inclined stope according to the calibrated geological exploration data of the steeply inclined stope and the stope design parameters;

[0010] Constructing an early warning index system for coal and rock dynamic disasters based on historical coal and rock disaster case data and expert experience;

[0011] Training and learning the early warning index system based on machine learning algorithms to construct a coal and rock dynamic disaster early warning model;

[0012] Integrating the coal and rock dynamic disaster early warning model into the digital twin model, and when the digital twin model is updated, the coal and rock dynamic disaster early warning model conducts real-time assessment and early warning of the dynamic disaster risk in the steeply inclined stope.

[0013] Preferably, the sensors include stress sensors, strain sensors, gas sensors, displacement sensors, and microseismic sensors;

[0014] The sensor installation parameters include the installation angle and the installation position; among them, the installation position includes the coal pillar area, the coal and rock mass in front of the shearer, and the hydraulic support area, and the sensor is arranged in association with the equipment and facilities in the stope;

[0015] The geological environment information of the steeply inclined stope includes the dip angle of the coal seam, the thickness change, and the distribution of geological structures.

[0016] Preferably, the method for establishing the calibration model of the sensor installation angle and position includes:

[0017] An adaptive array sensor group is used to dynamically adjust the installation attitude according to the dip angle of the coal seam, and the installation angle deviation is recorded in real time through laser ranging and inertial navigation unit to construct an installation angle deviation compensation model;

[0018] Establish the mapping relationship between the sensor installation parameters and the wave velocity anisotropy of the rock mass, and use the backpropagation correction matrix to compensate for signal attenuation to construct a wave velocity anisotropy compensation model;

[0019] Introduce the non-uniform medium wave field reconstruction technology, and simulate the propagation path deviation of electromagnetic waves in the coal-bearing rock fracture medium through finite difference time domain to construct a propagation path compensation model;

[0020] Based on the installation angle deviation compensation model, the wave velocity anisotropy compensation model, and the propagation path compensation model, the calibration model is obtained.

[0021] Preferably, the method for constructing the digital twin model includes:

[0022] Integrate the calibrated geological exploration data of the steeply inclined stope and the stope design parameters, construct a unified data format and coordinate system, and obtain multi-source fusion data;

[0023] Based on the multi-source fusion data, use three-dimensional modeling technology to construct a three-dimensional geometric model of the coal and rock mass;

[0024] Based on the three-dimensional geometric model, combined with the mechanical properties of coal and rock, the stress-strain law during the mining process, and the gas migration law, corresponding physical field attributes are assigned to the coal and rock mass to construct physical field models of the stress field, displacement field, and gas seepage field;

[0025] Perform three-dimensional modeling on the shearer, hydraulic support, and scraper conveyor in the stope to obtain the three-dimensional model of the equipment and facilities;

[0026] Fuse the three-dimensional geometric model, physical field model, and three-dimensional model of the equipment and facilities to obtain the digital twin model of the steeply inclined stope.

[0027] Preferably, the method for constructing the early warning index system for coal and rock dynamic disasters includes:

[0028] Extract the topological features of the historical disaster case data based on the graph convolutional network, construct an inclination similarity mapping function, perform knowledge transfer of the historical disaster case data across inclinations, and obtain three-dimensional spatio-temporal-frequency potential indicators;

[0029] Based on expert experience, screen the potential indicators to obtain candidate indicators for the early warning index system;

[0030] Use the candidate indicators as input variables, and whether a historical disaster occurs and the degree of the disaster as output variables to construct a machine learning model;

[0031] Based on the machine learning model, evaluate the contribution of the candidate indicators to the prediction of coal and rock dynamic disasters;

[0032] Based on the contribution degree, obtain the final early warning indicators and construct an early warning index system for coal and rock dynamic disasters.

[0033] Preferably, based on the early warning index system, use a physics-informed neural network to construct an early warning model for coal and rock dynamic disasters; wherein, a residual term of the coal and rock constitutive equation is embedded in the physics-informed neural network.

[0034] The present invention also provides a monitoring and early warning system for coal and rock dynamic disasters in a large-inclination stope based on digital twin, which is used to implement the method, and includes:

[0035] A data correction module, which is used to obtain sensor installation parameters and geological environment information of the large-inclination stope, establish a correction model, and correct the geological exploration data of the large-inclination stope collected by the sensor;

[0036] A digital twin model construction module, which is used to construct a digital twin model of the large-inclination stope according to the corrected geological exploration data of the large-inclination stope and the stope design parameters;

[0037] An early warning index system construction module, which is used to construct an early warning index system for coal and rock dynamic disasters based on historical coal and rock disaster case data and expert experience;

[0038] A disaster early warning model construction module, which is used to train and learn the early warning index system based on a machine learning algorithm to construct an early warning model for coal and rock dynamic disasters;

[0039] An early warning module, which is used to integrate the coal and rock dynamic disaster early warning model into the digital twin model. When the digital twin model is updated, the coal and rock dynamic disaster early warning model performs real-time evaluation and early warning on the dynamic disaster risk of the large-inclination stope.

[0040] Preferably, the data correction module includes:

[0041] An angle compensation unit, which is used to adopt an adaptive array sensor group to dynamically adjust the installation attitude according to the dip angle of the coal seam, and record the installation angle deviation in real time through laser ranging and an inertial navigation unit, and construct an installation angle deviation compensation model;

[0042] A wave velocity compensation unit, which is used to establish a mapping relationship between the sensor installation parameters and the anisotropy of the rock mass wave velocity, compensate for signal attenuation using a backpropagation correction matrix, and construct a wave velocity anisotropy compensation model;

[0043] A propagation path compensation unit, which is used to introduce a non-uniform medium wave field reconstruction technology, simulate the propagation path deviation of electromagnetic waves in a coal-bearing rock fracture medium through finite-difference time-domain, and construct a propagation path compensation model;

[0044] A correction model construction unit, which is used to obtain the correction model based on the installation angle deviation compensation model, the wave velocity anisotropy compensation model, and the propagation path compensation model.

[0045] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0046] Improved data accuracy: By establishing a correction model, the geological exploration data collected by the sensor can be effectively corrected, improving the accuracy and reliability of the data, and providing a higher-quality data basis for subsequent analysis and early warning.

[0047] Improved early warning accuracy: Using machine learning algorithms to train and learn the early warning index system, the constructed coal and rock dynamic disaster early warning model can more accurately evaluate and early warn the risk of coal and rock dynamic disasters, reducing false alarms and missed alarms.

[0048] Real-time monitoring and early warning: Integrating the early warning model into the digital twin model realizes the real-time assessment and early warning of the risk of coal and rock dynamic disasters in a steeply inclined stope, timely discovers potential safety hazards, and gains time for taking effective prevention and control measures.

[0049] Intelligent decision-making support: Combining the visualization and simulation analysis capabilities of the digital twin model provides intuitive stope state information and scientific decision-making basis for stope management personnel, and assists in formulating reasonable mining plans and disaster prevention and control strategies.

[0050] Significant safety benefits: Effectively reduces the occurrence risk of coal and rock dynamic disasters, guarantees the life safety of stope workers and the normal operation of equipment, and improves the safety and economic benefits of coal mining. Description of the Drawings

[0051] To more clearly illustrate the technical solution of the present invention, the following briefly introduces the drawings required in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0052] Figure 1 Flowchart of the method for monitoring and early warning of coal and rock dynamic disasters in steeply inclined stope based on digital twin in the embodiment of the present invention;

[0053] Figure 2 Schematic structural diagram of the electronic device in the embodiment of the present invention;

[0054] Figure 3 Flowchart of constructing the digital twin model in the embodiment of the present invention;

[0055] Figure 4 Flowchart of constructing the early warning index system for coal and rock dynamic disasters in the embodiment of the present invention.

[0056] Description:

[0057] 1010, processor; 1020, memory; 1030, input / output interface; 1040, communication interface; 1050, bus. Detailed implementation manners

[0058] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts belong to the protection scope of the present invention.

[0059] It should be noted that unless otherwise defined, the technical terms or scientific terms used in the embodiments of the present invention should have the ordinary meaning understood by those with ordinary skills in the field to which the present invention belongs. The "first", "second" and similar terms used in the embodiments of the present invention do not represent any order, quantity or importance, but are only used to distinguish different components. The terms such as "including" or "comprising" mean that the elements or objects appearing before this term cover the elements or objects listed after this term and their equivalents, without excluding other elements or objects. The terms such as "connected" or "coupled" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The terms such as "upper", "lower", "left", "right" are only used to represent relative positional relationships, and when the absolute position of the object being described changes, the relative positional relationship may also change accordingly.

[0060] To make the above objects, features, and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0061] Embodiment 1

[0062] As Figure 1 shown, a method for monitoring and warning coal and rock dynamic disasters in steeply inclined stope based on digital twin includes:

[0063] S1: Obtain the sensor installation parameters and the geological environment information of the steeply inclined stope, establish a calibration model, and calibrate the geological exploration data of the steeply inclined stope collected by the sensors.

[0064] A further implementation manner is that the sensors include stress sensors, strain sensors, gas sensors, displacement sensors, and microseismic sensors;

[0065] The sensor installation parameters include the installation angle and the installation position; among them, the installation position includes the coal pillar area, the coal and rock mass in front of the shearer, and the hydraulic support support area, and the sensors are associated and arranged with the equipment and facilities in the stope;

[0066] The geological environment information of the steeply inclined stope includes the dip angle of the coal seam, the thickness change, and the geological structure distribution.

[0067] A further implementation manner is that the method for establishing the calibration model of the sensor installation angle and position includes:

[0068] Adopt an adaptive array sensor group, dynamically adjust the installation attitude according to the dip angle of the coal seam, record the installation angle deviation in real time through laser ranging and inertial navigation unit, and construct an installation angle deviation compensation model; among them, the laser ranging array adopts an 8-point circular layout, and the inertial navigation unit adopts a sampling rate of 100Hz. Use the least squares method to fit the installation surface equation through the laser ranging data, calculate the included angle deviation between the actual normal vector and the ideal normal vector, and establish a Lie group SE(3) optimization model based on the theoretical installation point and the actual measurement point. Based on the optimization of the included angle deviation by the optimization model, implement dynamic compensation, and adjust the attitude of each sensor in the sensor group through a PID controller in combination with the real-time deviation.

[0069] It should be noted that wave velocity anisotropy refers to the characteristic that elastic waves (such as P waves, S waves) have different wave velocity characteristics in different propagation directions, which is usually caused by geological structures such as bedding, fractures, and stress fields in rock masses.

[0070] There is a close relationship between the calibration of the data collected by the sensors and wave velocity anisotropy. Especially in complex geological environments (such as steeply inclined coal seams), the anisotropic characteristics of rock masses will significantly affect the accuracy and reliability of sensor data.

[0071] In a coal seam, wave velocity anisotropy is mainly manifested as follows:

[0072] Bedding plane direction: The wave velocity is faster along the bedding plane and slower perpendicular to the bedding plane.

[0073] Fracture distribution: Fracture density and directionality can cause spatial variations in wave velocity.

[0074] Stress field: Non-uniform stress distribution can change the elastic modulus of the rock mass, thereby affecting wave velocity.

[0075] Moreover, anisotropy causes different degrees of attenuation of signals in different directions, and the signal intensity received by the sensor may be underestimated or overestimated. For example, the signal propagating perpendicular to the bedding plane attenuates more, which may result in lower displacement or stress measurement values. Wave velocity anisotropy can cause a deviation between the signal propagation time and the theoretical value, affecting the positioning accuracy of microseismic events. For example, the arrival time of the signal propagating along the bedding plane may be shorter than expected, resulting in a positioning deviation. In an anisotropic medium, the frequency components and waveforms of the signal may be distorted, affecting the analysis of acoustic emission or microseismic data. Wave velocity anisotropy has a significant impact on sensor data, so it is necessary to correct it.

[0076] Therefore, the present invention establishes a mapping relationship between the sensor installation parameters and the wave velocity anisotropy of the rock mass, uses a backpropagation correction matrix to compensate for signal attenuation, and constructs a wave velocity anisotropy compensation model. Specifically, 4 groups of P-wave transducers and 8 groups of S-wave receivers are arranged, and the three-dimensional wave velocity distribution is calculated through the time difference of flight to obtain an orthogonal elastic wave excitation-reception array. Based on the orthogonal elastic wave excitation-reception array, the sensor Euler angles, and the horizontal stress, a tensor mapping of wave velocity-installation angle is established to dynamically invert the rock mass parameters. Based on the direction-dependent attenuation coefficient and the angle between the wave propagation direction and the bedding plane, an anisotropic attenuation compensation equation is obtained. Then, the original signal received by the sensor is subjected to wavelet transform, and the direction-sensitive attenuation factor is calculated through the eigenvalues and eigenvectors of the wave velocity tensor. Based on the direction-sensitive factor and the anisotropic attenuation compensation equation, a backpropagation correction matrix is constructed, and the inverse wavelet transform is performed on the original signal after wavelet transform to obtain a reconstructed signal, completing the construction of the wave velocity anisotropy compensation model.

[0077] Introduce the non-uniform medium wave field reconstruction technology, simulate the propagation path deviation of electromagnetic waves in the coal-rock fracture medium through the finite-difference time-domain (FDTD), and construct a propagation path compensation model. Specifically, according to the physical parameters of the coal-rock medium and the fracture development situation, establish a non-uniform medium model containing coal-rock fractures. Divide the coal-rock mass into multiple small units, each unit having different electromagnetic parameters and fracture characteristics to reflect the non-uniformity and complexity of the coal-rock medium. Select a suitable FDTD simulation software, and set the initial conditions and boundary conditions of the FDTD simulation according to the coal-rock medium model and the electromagnetic wave monitoring data. This includes defining the size of the simulation area and grid division, setting the source term and receiving point position of the electromagnetic wave, specifying the time step and total duration of the simulation, etc. Run the FDTD simulation to calculate the propagation process of electromagnetic waves in the coal-rock fracture medium. Obtain information such as the electric and magnetic field distributions, propagation paths, and propagation times of electromagnetic waves in the coal-rock medium through the simulation. Analyze the reflection, refraction, scattering, etc. of electromagnetic waves at the fractures, as well as the propagation path deviation caused by the non-uniformity of the coal-rock medium. Compare the propagation path of the electromagnetic wave obtained by the FDTD simulation with the propagation path in the ideal homogeneous medium, and calculate the deviation amount between the actual propagation path and the ideal path. The deviation amount can be expressed as indicators such as the difference in path length and the offset of the propagation direction. At the same time, analyze the relationship between the propagation path deviation and the coal-rock medium parameters and fracture characteristics, and find out the main influencing factors and laws. According to the calculation results and analysis of the propagation path deviation, construct a propagation path compensation model. Among them, by collecting a large amount of propagation path deviation data and the corresponding coal-rock medium parameters, train a neural network model, use the coal-rock medium parameters as the input and the propagation path compensation amount as the output to construct the propagation path compensation model, and realize the automatic compensation prediction of the propagation path deviation.

[0078] Based on the installation angle deviation compensation model, wave velocity anisotropy compensation model, and propagation path compensation model, obtain a calibration model.

[0079] S2: According to the calibrated large dip stope geological exploration data and stope design parameters, construct a digital twin model of the large dip stope. As Figure 3 shown. Compared with nearly horizontal coal seams, the exploitation of the large dip coal seams (dip angle between 35° and 55°) studied in the present invention faces many unique challenges such as the gravity-dip angle effect, gas migration anisotropy, and insufficient adaptability of the support system. The specific solutions are reflected in using unstructured grid encryption in the areas with drastic dip angle changes (such as fault intersections) when constructing the three-dimensional geometric model, using vectorized modeling of the gravity field in the physical field calculation, and correcting the anisotropic seepage field. The adaptability of the support system is reflected in the early warning mechanism.

[0080] A further implementation manner lies in that the method for constructing the digital twin model includes:

[0081] Integrate the corrected geological exploration data of steeply inclined stope and stope design parameters, construct a unified data format and coordinate system, and obtain multi-source fusion data; specifically, collect the corrected geological exploration data of steeply inclined stope, including multi-parameter data such as stress, strain, gas concentration, displacement of coal and rock mass, and stope design parameters, such as geometric dimensions of the stope, coal seam dip angle, shearer model and parameters, hydraulic support model and parameters, etc. Systematically organize these data and establish a database. Through data consistency check, discover and correct large differences in stress data measured at the same position by different sensors; through data integrity check, supplement missing data points or perform reasonable interpolation processing.

[0082] Based on the multi-source fusion data, use 3D modeling technology to construct a 3D geometric model of coal and rock mass; specifically, during the modeling process, accurately restore the shape, size and spatial position of the coal and rock mass to ensure that the model is highly consistent with the coal and rock mass structure of the actual stope. At the same time, considering the complexity and irregularity of the coal and rock mass, dynamically adjust the grid density and direction of the 3D geometric model based on the change of the coal seam dip angle. In areas with drastic dip angle changes (geological structures, such as folds, faults, etc.), use unstructured grid encryption to ensure refined modeling of stress concentration areas and gas enrichment areas for subsequent physical field simulation and mechanical analysis. In particular, this embodiment provides a detailed process for constructing a 3D geometric model:

[0083] Construct a deep neural network (DNN), with inputs including parameters such as coal seam dip angle and thickness, stress gradient heat map, gas concentration field and microseismic event spatial clustering data, and the output is a dynamic grid division strategy. The network is trained through reinforcement learning, aiming to minimize the residual of subsequent physical field simulation. Priority is given to generating unstructured encrypted grids (side length ≤ 0.5m) in areas with sudden dip angle changes (such as > 40°) and areas with dense microseismic events, while using sparse structured grids (side length ≤ 2m) in stable areas. Through a real-time feedback mechanism, update the grid configuration every 30 minutes according to the latest sensor data to ensure that the model dynamically adapts to changes in geological conditions.

[0084] Based on the microseismic monitoring data, invert the 3D topological structure of the fracture network, use the fractal geometry algorithm to generate a statistically equivalent fracture model, and embed it into the coal and rock mass geometric model. For the bedding plane of steeply inclined coal seams, define the dip-dip tensor field, force the bedding plane to be orthogonal to the normal vector of the coal seam during modeling, and fit the geometric shape of the bedding plane through non-uniform rational B-spline (NURBS) surface. At the intersection of faults, adopt a hybrid modeling method of discrete fracture network (DFN) and continuous medium model, and connect macroscopic faults and microscopic fractures through virtual joint elements to achieve coupling of cross-scale mechanical behaviors.

[0085] Convert the mesh quality evaluation function (the condition number of the Jacobi matrix is used in this embodiment) into a Quadratic Unconstrained Binary Optimization (QUBO) problem, and use the quantum annealing algorithm to solve the optimal node positions. Aiming at the element distortion problem at the encrypted mesh boundary, a local optimization strategy based on the quantum tunneling effect is introduced, which improves the mesh smoothness by 42% while maintaining the inclination-related features. Among them, the principle of the introduced local optimization strategy based on the quantum tunneling effect is as follows: during the quantum annealing process, quantum fluctuations are introduced by applying a transverse magnetic field, enabling the system to have the ability to cross the energy barrier. When the algorithm falls into a local optimum (such as the distortion caused by the sudden change of the inclination angle of the mesh boundary nodes), the quantum tunneling effect allows the system to "cross" the classically insurmountable high-energy state in the form of probability to explore a better solution space. For example, at the junction of the encrypted mesh and the sparse mesh, the traditional gradient descent method may stagnate due to the sudden change of the element aspect ratio, while the quantum tunneling can break through such geometric constraints. The tunneling intensity is adaptively adjusted according to the local mesh features (such as the inclination gradient, stress concentration coefficient): in the high-distortion region (Jacobi condition number > 10), the tunneling probability is enhanced to accelerate jumping out of the local optimum; in the smooth region (condition number < 3), the tunneling weight is reduced, and classical optimization is preferred to improve the convergence speed. This strategy achieves a balance between global exploration and local refinement through quantum-classical hybrid optimization.

[0086] Through a Mixed Reality (MR) device, the three-dimensional geometric model is superimposed on the real stope environment, enabling engineers to adjust the mesh density and direction in real time with gesture interaction. The system automatically detects the manually modified areas and compares them with the physical field simulation results (such as the maximum principal stress error > 15%). If they are inconsistent, it triggers the adaptive mesh regeneration to form a "human-machine collaborative modeling" closed loop.

[0087] Based on the three-dimensional geometric model, combined with the mechanical properties of coal and rock, the stress-strain law during the mining process, and the gas migration law, corresponding physical field attributes are assigned to the coal and rock mass, and physical field models of the stress field, displacement field, and gas seepage field are constructed. In the physical field calculation, the gravitational acceleration is decomposed into normal and tangential components, which are respectively coupled to the stress field and displacement field solvers to accurately simulate the asymmetric load distribution caused by the dip angle. Based on the real-time dip angle data (from the inertial navigation unit), the weight of the gravity component is dynamically adjusted to achieve an adaptive characterization of the mechanical response during the stope advancement process. Specifically, through laboratory tests and in-situ tests, the mechanical property parameters of the coal and rock mass are obtained, such as elastic modulus, Poisson's ratio, compressive strength, shear strength, etc. These parameters are the basis for constructing the mechanical physical field model of the coal and rock mass and can reflect the mechanical response of the coal and rock mass during the mining process. Using numerical simulation methods (to construct the stress field and displacement field models of the coal and rock mass. Reasonable boundary conditions and load conditions are set, including the constraint conditions of the stope roof and floor, the mining pressure of the coal mining face, etc., to simulate the stress distribution and displacement changes of the coal and rock mass during the mining process. According to the gas migration law of the coal and rock mass and the gas concentration distribution information in the geological exploration data, combined with parameters such as the permeability of the coal and rock mass, a gas seepage field model is constructed. Considering the gas migration mechanisms such as diffusion and convection in the coal and rock mass, and the influence of the deformation of the coal seam during the mining process on the gas seepage path and pressure, the gas seepage distribution and dynamic changes in the coal and rock mass are simulated. Specifically, a dip angle correction coefficient is embedded in the gas seepage model to quantify the inhibitory effect of the dip angle on the permeability; the included angle between the coal seam dip direction and the main seepage direction is inversely calculated in real time through the digital twin model, and the anisotropic tensor in the seepage equation is dynamically adjusted to improve the prediction accuracy of the gas enrichment area.

[0088] Particularly, in this embodiment, aiming at the asymmetric stress characteristics of the roof and floor of the steeply inclined stope, a boundary constraint equation related to the dip angle is introduced into the physical field model to simulate the interaction between the coal seam slip trend and the support structure.

[0089] Three-dimensional modeling is carried out on the shearer, hydraulic support, and scraper conveyor in the stope to obtain the three-dimensional model of the equipment and facilities; specifically, according to the actual size, shape, and performance parameters of the equipment, a precise model of the equipment is created using three-dimensional modeling software and placed at the corresponding position in the three-dimensional geometric model of the coal and rock mass to ensure the accurate spatial relationship between the equipment and the coal and rock mass. At the same time, corresponding kinematic and dynamic attributes are added to the equipment model, such as the coal cutting action of the shearer, the lifting action of the hydraulic support, etc., so that the equipment model can dynamically simulate the operating state during the actual mining process.

[0090] Integrate the three-dimensional geometric model, physical field model, and three-dimensional model of equipment and facilities to obtain a digital twin model of the steeply inclined stope. Specifically, during the integration process, ensure that the data interfaces and communication protocols between each sub-model match to achieve seamless data transmission and sharing. Combine the calculation results of the physical field model with the three-dimensional geometric model of coal and rock masses, so that the geometric shape of the coal and rock masses can be updated in real time according to the changes in the physical field; couple the operating state data of the equipment and facilities with the physical field model of the coal and rock masses to simulate the impact of equipment operation on the coal and rock masses and the feedback effect of the state change of the coal and rock masses on equipment operation. Combine the microseismic monitoring data to update the fracture topology structure in the digital twin model in real time, and simulate the correlation between fracture propagation and coal and rock mass instability through the discrete element method (DEM).

[0091] In this embodiment, the on-site real-time data collected by the sensors is interacted with the constructed virtual models in the following ways:

[0092] Sensors such as stress, strain, and gas are networked in a hybrid of 5G / industrial Ethernet, and the Precision Time Protocol (PTP) is used to ensure the consistency of data time series. The sampling frequency is 100Hz - 1kHz. Deploy an embedded edge server in the stope to run a lightweight calibration model (such as an installation angle deviation compensation model) to filter, denoise, and normalize the coordinates of the original data, and compress the data volume to 30% of the original size.

[0093] Push the preprocessed data to the cloud through the MQTT protocol. Enable the UDP acceleration channel for key parameters (such as microseismic event waveforms and sudden changes in gas concentration), and the end-to-end delay < 50ms.

[0094] Adopt a "hot and cold stratification" storage strategy:

[0095] Hot storage: Real-time data is stored in a time series database (such as InfluxDB), and the data of the most recent 72 hours is retained to support millisecond-level queries;

[0096] Cold storage: Historical data is transferred to a distributed file system (such as HDFS) for classification and archiving to facilitate model training and retrospective analysis.

[0097] Regarding the injection of real-time data, access the sensor data stream to the digital twin platform through the RESTful API, and use the Kafka message queue to achieve high-concurrency data buffering to ensure stable throughput of tens of thousands of data points per second.

[0098] For physical quantities such as the stress field and displacement field, perform a local recalculation of the finite element model every 10 seconds, and only update the area grid with a change rate > 5%; when the microseismic energy > 10 3When the mutation of J or gas concentration gradient > 20%, the full model iteration is immediately triggered to ensure the accurate mapping of the precursor of the disaster. Compare the stress distribution predicted by the digital twin model with the measured data. If the mean square error (MSE) > 15%, the sensor calibration or the recalibration of model parameters is automatically triggered; the support adjustment instructions output by the early warning model are sent to the hydraulic support controller via the OPC UA protocol. At the same time, the execution results are recorded and fed back to the digital twin model to form a closed-loop verification. Through the Hololens2 device, the real-time sensor data (such as displacement vector, gas concentration cloud map) is superimposed on the stope real scene in the form of AR, supporting gesture interaction to retrieve the historical data trend at any position. Based on the WebGL technology, the dynamic three-dimensional rendering of the coal and rock mass fracture network and the equipment operation status is realized in the digital twin platform, and the data update delay < 200ms to ensure the frame synchronization between the virtual model and the physical world.

[0099] S3: Based on the historical coal and rock disaster case data and expert experience, construct an early warning index system for coal and rock dynamic disasters. As Figure 4 shown.

[0100] A further implementation manner lies in that the method for constructing the early warning index system for coal and rock dynamic disasters includes:

[0101] Extract the topological features of the historical disaster case data based on the graph convolutional network, construct an inclination similarity mapping function, perform knowledge transfer of the historical disaster case data across inclinations, and obtain three-dimensional potential indicators in time, space, and frequency domains; among them, the time-domain indicators include displacement acceleration rate and stress change rate; the space-domain indicators include stress gradient and bedding plane angle; the frequency-domain indicators include the main frequency band energy of acoustic emission and the b value of microseismic events. In this embodiment, auxiliary indicators are also included: coal seam inclination, burial depth, fracture density, groundwater pressure, and ground temperature.

[0102] In particular, this embodiment also introduces the following potential indicators based on the collected historical disaster case data for the unique disaster mode of the large-inclination stope:

[0103] Inclination-stress coupling factor: By calculating the cosine value of the angle between the coal seam inclination and the principal stress direction, the contribution degree of the inclination to the stress redistribution is quantified, and the threshold is set as where n is the normal vector of the bedding plane;

[0104] Tangential displacement mutation rate: Monitor the change of the displacement gradient along the coal seam dip direction, use wavelet transform to extract the singular points of the displacement signal, and identify the precursor of slip instability;

[0105] Gas migration inclination effect coefficient: Based on the modified formula of Darcy's law, introduce an inclination correction term to characterize the attenuation law of gas permeability caused by the increase of inclination;

[0106] Support structure dip angle adaptability: Evaluate the adaptability of the support system to steeply inclined coal seams through the real-time deviation between the direction of the support force of the hydraulic support and the normal direction of the coal seam. An early warning is triggered when the deviation exceeds 5°. The implementation process is as follows: The digital twin model calculates the ideal support force direction of the hydraulic support based on the real-time changes of the normal vector of the coal seam (measured by the micro-inertial navigation system), generates control instructions and sends them to the electro-hydraulic system of the support; A support force-displacement coupling feedback mechanism is embedded in the model. If the deviation between the actual support direction of the support and the normal direction of the coal seam is detected to be >5°, an early warning is immediately triggered and a rectification program is started.

[0107] Based on expert experience, screen potential indicators to obtain candidate indicators for the early warning indicator system;

[0108] Take the candidate indicators as input variables, and whether historical disasters occur and the degree of disasters as output variables to build a machine learning model;

[0109] Based on the machine learning model, evaluate the contribution of the candidate indicators to the prediction of coal and rock dynamic disasters;

[0110] Based on the contribution degree, obtain the final early warning indicators and build an early warning indicator system for coal and rock dynamic disasters.

[0111] S4: Train and learn the early warning indicator system based on machine learning algorithms to build a coal and rock dynamic disaster early warning model; A further implementation method is to build a coal and rock dynamic disaster early warning model based on the early warning indicator system using a physics-informed neural network; Among them, in the physics-informed neural network, the residual term of the coal and rock constitutive equation is embedded. Specifically, design a deep neural network structure. The input layer receives early warning indicator data (such as stress, strain, gas concentration, etc.), the hidden layer uses an appropriate activation function (such as ReLU), and the output layer outputs the occurrence probability or risk level of coal and rock dynamic disasters. The coal and rock constitutive equation describes the relationship between coal and rock stress and strain, and its form is:

[0112] σ = f(∈, coal and rock parameters),

[0113] where σ is stress, ∈ is strain, and coal and rock parameters include elastic modulus, Poisson's ratio, etc. In the neural network, in addition to the data-driven loss term, the residual term of the constitutive equation is introduced as a physical constraint. The residual term is defined as the difference between the stress predicted by the network and the stress calculated by the constitutive equation:

[0114] R = σ 预测 -f(∈ 预测 , coal and rock parameters),

[0115] Take the residual term R as part of the network optimization to ensure that the prediction results of the network conform to the physical properties of coal and rock.

[0116] S5: Integrate the coal and rock dynamic disaster early warning model into the digital twin model. When the digital twin model is updated, the coal and rock dynamic disaster early warning model conducts real-time assessment and early warning of the dynamic disaster risks in the steeply inclined stope. Specifically, the early warning mechanism constructed by the present invention achieves precise response in the steeply inclined stope through the following innovations:

[0117] 1. Based on the real-time data of the digital twin model, use the spatio-temporal graph neural network (STGNN) to model the dip-sensitive disaster chain of "microseismic events → crack propagation → stress transfer → macroscopic instability", and predict the disaster propagation path;

[0118] 2. Use the fuzzy logic algorithm to fuse expert experience and real-time data, and adaptively correct the early warning threshold, thereby reducing the false alarm rate;

[0119] 3. Through the mixed reality (MR) device, superimpose the high-risk areas in the digital twin model onto the real stope view, mark the slip risk level with a dip gradient color band, and dynamically display the disaster avoidance route;

[0120] 4. After the early warning signal is triggered, automatically adjust the cutting trajectory of the shearer and the support parameters of the hydraulic support to form a "monitoring - early warning - control" closed loop to suppress the dip-induced dynamic disaster.

[0121] Specifically, regarding the dynamic update and real-time feedback mechanism of the digital twin model with the stope state:

[0122] Through the process of interaction between the on-site real-time data collected by the aforementioned sensors in this embodiment and the constructed virtual models, model-driven closed-loop control is realized. Then, in the early warning part, this embodiment also involves feedback delay compensation: use the Kalman filter to fuse the sensor data stream and the model prediction results, compensate for the network transmission and calculation delay (typical value 50 ms), and ensure that the virtual model is strictly synchronized with the physical world state. Based on the advanced simulation ability of the digital twin model, simulate the effects of different control strategies (such as a 10% increase or decrease in the support force of the support) 1 minute in advance, select the optimal strategy to execute, and avoid the control lag caused by feedback delay.

[0123] After each control instruction is executed, collect the actual effect data (such as the displacement change rate after the support force is adjusted), update the early warning model parameters through reinforcement learning (RL), design a loss function to comprehensively consider safety (such as suppressing stress mutation), efficiency (such as the advance amount of the shearer), and energy consumption (such as the power of the hydraulic system), and use the NSGA-II algorithm to achieve Pareto optimal control.

[0124] To ensure the accuracy of disaster early warning, this embodiment adopts human-machine collaborative intervention:

[0125] Dynamic permission allocation: In low - risk scenarios (disaster probability < 5%), the system runs fully automatically; in medium - high - risk scenarios (5% - 30%), warning information is pushed to the manual monitoring interface, supporting engineers to modify control instructions through AR gesture overlay.

[0126] Traceability audit: All automatically generated instructions and manual intervention records are stored on the blockchain (based on Hyperledger Fabric) to ensure traceability of operations and clear responsibilities.

[0127] It should be noted that the method of the embodiment of the present invention can be executed by a single device, such as a computer or a server. The method of this embodiment can also be applied to a distributed scenario, and completed by multiple devices cooperating with each other. In this case of a distributed scenario, one of these multiple devices can only execute one or more steps of the method of the embodiment of the present invention, and these multiple devices will interact with each other to complete the described method.

[0128] It should be noted that some embodiments of the present invention have been described above. Other embodiments are within the scope of the appended claims. In some cases, it should be understood that the order of the numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiment of the present invention. The actions or steps recorded in the claims can be executed in a different order from that in the above embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0129] Embodiment 2

[0130] Based on the same inventive concept, corresponding to the method of any of the above embodiments, the present invention further provides a digital - twin - based monitoring and early - warning system for coal - rock dynamic disasters in steeply - inclined stope, used to implement the method, including:

[0131] A data correction module, used to obtain sensor installation parameters and geological environment information of the steeply - inclined stope, establish a correction model, and correct the geological exploration data of the steeply - inclined stope collected by the sensor;

[0132] A digital - twin model construction module, used to construct a digital - twin model of the steeply - inclined stope according to the corrected geological exploration data of the steeply - inclined stope and stope design parameters;

[0133] An early - warning index system construction module, used to construct an early - warning index system for coal - rock dynamic disasters based on historical coal - rock disaster case data and expert experience;

[0134] A disaster warning model construction module, which is used to train and learn the warning index system based on machine learning algorithms to construct a coal and rock dynamic disaster warning model;

[0135] A warning module, which is used to integrate the coal and rock dynamic disaster warning model into the digital twin model. When the digital twin model is updated, the coal and rock dynamic disaster warning model conducts real-time assessment and warning of the dynamic disaster risk in the steeply inclined stope.

[0136] A further implementation manner is that the data correction module includes:

[0137] An angle compensation unit, which is used to adopt an adaptive array sensor group to dynamically adjust the installation attitude according to the dip angle of the coal seam, record the installation angle deviation in real time through laser ranging and inertial navigation units, and construct an installation angle deviation compensation model;

[0138] A wave velocity compensation unit, which is used to establish a mapping relationship between the sensor installation parameters and the anisotropy of the rock mass wave velocity, compensate the signal attenuation by using a backpropagation correction matrix, and construct an anisotropy compensation model for wave velocity;

[0139] A propagation path compensation unit, which is used to introduce a non-uniform medium wave field reconstruction technology, simulate the propagation path deviation of electromagnetic waves in the coal and rock fractured medium through finite difference time domain, and construct a propagation path compensation model;

[0140] A correction model construction unit, which is used to obtain a correction model based on the installation angle deviation compensation model, the anisotropy compensation model for wave velocity, and the propagation path compensation model.

[0141] The system of the above embodiment is used to implement the corresponding method in any of the foregoing embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be elaborated here.

[0142] It should be noted that the above system is embodied in the form of functional units. The term "module" here can be implemented in software and / or hardware forms, and no specific limitation is made thereto.

[0143] For example, the "module" can be a software program, a hardware circuit, or a combination of the two to implement the above functions. The hardware circuit may include an application specific integrated circuit (ASIC), an electronic circuit, a processor (such as a shared processor, a proprietary processor, or a group of processors, etc.) for executing one or more software or firmware programs, a memory, a combined logic circuit, and / or other suitable components to support the described functions.

[0144] Embodiment III

[0145] Based on the same inventive concept, corresponding to the method of any of the above embodiments, the present invention further provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the method of any of the above embodiments is implemented.

[0146] Figure 2 FIG. shows a more specific schematic diagram of the hardware structure of the electronic device provided in this embodiment. The device may include: a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. Among them, the processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040 are communicatively connected to each other inside the device through the bus 1050.

[0147] The processor 1010 may be implemented in a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, etc., and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.

[0148] The memory 1020 may be implemented in the form of a ROM (Read Only Memory), a RAM (Random Access Memory), a static storage device, a dynamic storage device, etc. The memory 1020 may store an operating system and other application programs. When implementing the technical solutions provided in the embodiments of this specification through software or firmware, the relevant program codes are stored in the memory 1020 and are called and executed by the processor 1010.

[0149] The input / output interface 1030 is used to connect to the input / output module to implement information input and output. The input / output module may be configured as a component in the device (not shown in the figure) or externally connected to the device to provide corresponding functions. Among them, the input device may include a keyboard, a mouse, a touch screen, a microphone, various sensors, etc., and the output device may include a display, a speaker, a vibrator, an indicator light, etc.

[0150] The communication interface 1040 is used to connect to a communication module (not shown in the figure) to enable communication and interaction between this device and other devices. The communication module can achieve communication through wired means (such as USB (Universal Serial Bus), network cable, etc.) or wireless means (such as mobile network, WIFI (Wireless Fidelity), Bluetooth, etc.).

[0151] The bus 1050 includes a path for transmitting information between various components of the device (such as the processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040).

[0152] It should be noted that although the above device only shows the processor 1010, the memory 1020, the input / output interface 1030, the communication interface 1040, and the bus 1050, in the specific implementation process, the device may also include other components necessary for normal operation. In addition, those skilled in the art can understand that the above device may also only include the components necessary to implement the solution of the embodiments of this specification, and does not necessarily include all the components shown in the figure.

[0153] The system of the above embodiment is used to implement the corresponding method in any of the foregoing embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be elaborated here.

[0154] Embodiment 4

[0155] Based on the same inventive concept, corresponding to the method of any of the above embodiments, the present invention also provides a non-transitory computer-readable storage medium storing computer instructions for causing the computer to execute the method described in any of the above embodiments.

[0156] The computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory, or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD), or other optical storage, magnetic cassette tapes, magnetic disk storage, or other magnetic storage devices, or any other non-transmission medium that can be used to store information accessible by a computing device.

[0157] The computer instructions stored in the storage medium of the above embodiments are used to cause the computer to execute the method described in any one of the above embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be elaborated herein.

[0158] Those of ordinary skill in the art should understand that: The discussion of any of the above embodiments is only exemplary, and is not intended to imply that the scope of the present invention (including the claims) is limited to these examples; Under the concept of the present invention, the technical features in the above embodiments or different embodiments can also be combined, and the steps can be implemented in any order, and there are many other variations in different aspects of the embodiments of the present invention as described above, and they are not provided in detail for the sake of brevity.

[0159] In addition, for simplicity of explanation and discussion, and in order not to make the embodiments of the present invention difficult to understand, the well-known power / ground connections to integrated circuit (IC) chips and other components may or may not be shown in the provided drawings. In addition, the device may be shown in block diagram form in order to avoid making the embodiments of the present invention difficult to understand, and this also takes into account the fact that the details of the implementation of these block diagram devices are highly dependent on the platform on which the embodiments of the present invention are to be implemented (i.e., these details should be fully within the understanding of those skilled in the art). In the case where specific details (such as circuits) are set forth to describe exemplary embodiments of the present invention, it will be apparent to those skilled in the art that the embodiments of the present invention can be implemented without these specific details or with variations of these specific details. Therefore, these descriptions should be considered illustrative rather than restrictive.

[0160] Although the present invention has been described in connection with specific embodiments of the present invention, many alternatives, modifications, and variations of these embodiments will be apparent to those of ordinary skill in the art based on the foregoing description. For example, other memory architectures (such as dynamic RAM (DRAM)) may be used with the embodiments discussed.

[0161] Therefore, the units of the examples described in the embodiments of the present application can be implemented in electronic hardware, or in a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present application.

[0162] Embodiments of the present invention are intended to cover all such substitutions, modifications, and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the embodiments of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for monitoring and early warning of coal and rock dynamic disasters in steeply inclined stope based on digital twin, characterized in that, Including: Obtain sensor installation parameters and geological environment information of a steeply inclined stope, establish a calibration model, and calibrate the geological exploration data of the steeply inclined stope collected by the sensor; Construct a digital twin model of the steeply inclined stope according to the calibrated geological exploration data of the steeply inclined stope and stope design parameters; Construct an early warning index system for coal and rock dynamic disasters based on historical coal and rock disaster case data and expert experience; Train and learn the early warning index system based on a machine learning algorithm to construct a coal and rock dynamic disaster early warning model; Integrate the coal and rock dynamic disaster early warning model into the digital twin model. When the digital twin model is updated, the coal and rock dynamic disaster early warning model conducts real-time assessment and early warning of the dynamic disaster risk in the steeply inclined stope.

2. The method according to claim 1, wherein The sensors include stress sensors, strain sensors, gas sensors, displacement sensors, and microseismic sensors; The sensor installation parameters include the installation angle and installation position; wherein, the installation position includes the coal pillar area, the coal and rock mass in front of the shearer, and the hydraulic support support area, and the sensors are associated and arranged with the equipment and facilities in the stope; The geological environment information of the steeply inclined stope includes the dip angle of the coal seam, thickness change, and geological structure distribution.

3. The method according to claim 2, wherein The method for establishing a calibration model for the sensor installation angle and position includes: Adopt an adaptive array sensor group, dynamically adjust the installation attitude according to the dip angle of the coal seam, and record the installation angle deviation in real time through laser ranging and inertial navigation units to construct an installation angle deviation compensation model; Establish a mapping relationship between the sensor installation parameters and the wave velocity anisotropy of the rock mass, and use a backpropagation correction matrix to compensate for signal attenuation to construct a wave velocity anisotropy compensation model; Introduce a non-uniform medium wave field reconstruction technology, and simulate the propagation path deviation of electromagnetic waves in the coal and rock fracture medium through finite difference time domain to construct a propagation path compensation model; Based on the installation angle deviation compensation model, the wave velocity anisotropy compensation model, and the propagation path compensation model, obtain the calibration model.

4. The method according to claim 1, characterized in that The method for constructing the digital twin model includes: Integrate the calibrated geological exploration data of the steeply inclined stope and stope design parameters, construct a unified data format and coordinate system, and obtain multi-source fusion data; Based on the multi-source fusion data, use 3D modeling technology to construct a 3D geometric model of the coal and rock mass; Based on the 3D geometric model, combine the mechanical properties of coal and rock, the stress and strain laws during the mining process, and the gas migration laws to endow the coal and rock mass with corresponding physical field attributes, and construct physical field models of the stress field, displacement field, and gas seepage field; Conduct 3D modeling on the shearer, hydraulic support, and scraper conveyor in the stope to obtain 3D models of equipment and facilities; Fuse the 3D geometric model, physical field model, and 3D models of equipment and facilities to obtain the digital twin model of the steeply inclined stope.

5. The method according to claim 1, wherein The method for constructing an early warning index system for coal and rock dynamic disasters includes: Extract the topological features of historical disaster case data based on graph convolutional networks, construct an inclination similarity mapping function, perform knowledge transfer of historical disaster case data across inclinations, and obtain three-dimensional potential indicators of time, space, and frequency; Based on expert experience, screen the potential indicators to obtain candidate indicators for the early warning index system; Use the candidate indicators as input variables and whether a historical disaster occurred and the degree of the disaster as output variables to construct a machine learning model; Based on the machine learning model, evaluate the contribution of the candidate indicators to the prediction of coal and rock dynamic disasters; Based on the contribution degree, obtain the final early warning indicators and construct an early warning index system for coal and rock dynamic disasters.

6. The method according to claim 1, characterized in that, Based on the early warning index system, use a physics-informed neural network to construct an early warning model for coal and rock dynamic disasters; wherein, in the physics-informed neural network, a residual term of the coal and rock constitutive equation is embedded.

7. A monitoring and early warning system for coal and rock dynamic disasters in steeply inclined stope based on digital twin, which is used to implement the method described in any one of claims 1-6, and is characterized in that Including: A data correction module for obtaining sensor installation parameters and large-inclination stope geological environment information, establishing a correction model, and correcting the large-inclination stope geological exploration data collected by the sensors; A digital twin model construction module for constructing a digital twin model of the large-inclination stope according to the corrected large-inclination stope geological exploration data and stope design parameters; An early warning index system construction module for constructing an early warning index system for coal and rock dynamic disasters based on historical coal and rock disaster case data and expert experience; A disaster early warning model construction module for training and learning the early warning index system based on machine learning algorithms to construct an early warning model for coal and rock dynamic disasters; An early warning module for integrating the coal and rock dynamic disaster early warning model into the digital twin model, and when the digital twin model is updated, the coal and rock dynamic disaster early warning model performs real-time assessment and early warning of the dynamic disaster risk of the large-inclination stope.

8. The system according to claim 7, wherein The data correction module includes: An angle compensation unit for using an adaptive array sensor group to dynamically adjust the installation attitude according to the inclination of the coal seam, and constructing an installation angle deviation compensation model by recording the installation angle deviation in real time through laser ranging and an inertial navigation unit; A wave velocity compensation unit for establishing a mapping relationship between sensor installation parameters and the anisotropy of rock mass wave velocity, and constructing a wave velocity anisotropy compensation model by compensating signal attenuation using a backpropagation correction matrix; A propagation path compensation unit for introducing a non-uniform medium wave field reconstruction technology, simulating the propagation path deviation of electromagnetic waves in a coal and rock fracture medium by finite difference time domain, and constructing a propagation path compensation model; A correction model construction unit for obtaining the correction model based on the installation angle deviation compensation model, the wave velocity anisotropy compensation model, and the propagation path compensation model.

Images