An intelligent early warning method and system based on mine microseismic data analysis

Abstract

This invention discloses an intelligent early warning method and system based on mine microseismic data analysis, specifically relating to the field of mine safety monitoring technology. The method receives microseismic waveform sequences and the spatial trajectory of a coal mining machine, extracts the three-dimensional spatial coordinates, radiant energy, and corner frequencies of microseismic events through travel-time inversion and spectral analysis; reconstructs the focal mechanism solution by combining the initial amplitude ratio, screens floor tensile fracture events, and calculates dynamic stress drop; superimposes the dynamic stress drop onto the initial geostress tensor to reconstruct the mining-induced additional stress field, quantifies the intrinsic permeability increment using the dot product projection of the fracture surface normal vector and the local tensile principal stress direction vector; clusters these fractures into time-series arranged water-conducting fracture clusters, calculates the vertical approximation rate of the centroid migration three-dimensional vector projected onto the floor aquifer, and the rock mass volume expansion rate, finally outputting a water hazard early warning command with corresponding coordinates. The system described above is used to implement the above method.

Description

Technical Field

[0001] This invention relates to the field of mine safety monitoring technology, specifically to an intelligent early warning method and system based on mine microseismic data analysis. Background Technology

[0002] Water hazards in coal mines, along with gas and fire, are considered core safety threats in mining operations. As mining extends deeper and operations expand, the hidden threat posed by confined aquifers in the floor strata to the mining process is increasing. In complex geological environments, micro-fractures within the floor strata can fracture and propagate under continuous mining stress. When these fractures develop together and connect to concealed aquifers in the floor, they can easily trigger water inrush disasters. Dynamic sensing and real-time monitoring of the micro-fracture state of the floor rock mass and the evolution of water-conducting channels are crucial for ensuring the safety of underground workers and the continuity of mine production.

[0003] However, in the process of water hazard prevention, existing microseismic monitoring technologies typically use fixed empirical thresholds to determine the threat level of a single microseismic event, or rely solely on simple statistical analysis of the frequency of microseismic events within a specific time period. This analytical approach over-relies on static surface vibration data, ignoring the real-time stress disturbance differences caused by the dynamic advancement of coal mining machines on the floor strata. This often leads to misjudging non-water-conducting conventional pressure relief ruptures as precursors to water inrush, significantly increasing the false alarm rate of early warnings. Simultaneously, existing data processing models lack the ability to track the spatiotemporal evolution trajectory of microseismic clusters in three-dimensional geological space. The acquired monitoring data exhibits severe fragmentation, failing to deeply analyze the underlying physical evolution mechanisms of floor strata damage and the formation of hidden water-conducting channels. Consequently, there is a significant lag in the dynamic perception and response to water hazard risks. Therefore, it remains necessary to provide an intelligent early warning method based on mine microseismic data analysis to improve the accuracy of water hazard early warning under complex geological conditions. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, embodiments of the present invention provide an intelligent early warning method and system based on mine microseismic data analysis to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: an intelligent early warning method based on mine microseismic data analysis, comprising the following steps: S1, receiving microseismic waveform sequences and coal mining machine spatial trajectories collected by an IoT microseismic sensor network, performing phase identification and travel time inversion on the microseismic waveform sequences, extracting the three-dimensional spatial coordinates of microseismic events, and analyzing the frequency domain spectrum of the microseismic waveform sequences to extract radiation energy and corner frequencies; S2, extracting the initial amplitude ratio of P-waves and S-waves of the microseismic waveform sequences, reconstructing the focal mechanism solution by combining the three-dimensional spatial coordinates, screening the tensile fracture events of the floor according to the focal mechanism solution, and calculating the dynamic stress drop of the tensile fracture events by integrating radiation energy, corner frequencies, and three-dimensional spatial coordinates; S3, obtaining the initial ground stress tensor of the mining face, and calculating the dynamic stress drop... Superimposed on the initial geostress tensor, the mining-induced additional stress field in the space where the tensile rupture event is located is reconstructed. The direction vector of the local tensile principal stress in the mining-induced additional stress field is extracted. The dot product projection of the normal vector of the rupture surface and the direction vector of the local tensile principal stress in the focal mechanism solution is calculated. The dot product projection and the intrinsic permeability increment of dynamic stress reduction are fused. S4. Based on the intrinsic permeability increment, the tensile rupture event is subjected to three-dimensional spatial density clustering to generate water-conducting fracture clusters arranged in time sequence. The three-dimensional centroid migration vector and the rock mass volume expansion rate of the water-conducting fracture clusters in adjacent continuous time windows are calculated. The vertical approximation rate is obtained by projecting the three-dimensional centroid migration vector onto the normal of the bottom aquifer. Based on the vertical approximation rate and the rock mass volume expansion rate, the water hazard warning command corresponding to the three-dimensional spatial coordinates is output.

[0006] In a preferred embodiment, the specific process of receiving microseismic waveform sequences and coal mining machine spatial trajectories collected by an IoT microseismic sensor network, performing phase identification and travel time inversion on the microseismic waveform sequences, and extracting the three-dimensional spatial coordinates of microseismic events is as follows: obtaining the arrival phases of the P-waves and S-waves of the microseismic waveform sequences and the three-dimensional deployment coordinates of the IoT microseismic sensor network nodes, and constructing a theoretical travel time residual objective function by combining the arrival phases of the P-waves and S-waves and the three-dimensional deployment coordinates; establishing a priori spatial constraint boundary based on the coal mining machine spatial trajectory, and performing nonlinear inversion optimization on the theoretical travel time residual objective function under the priori spatial constraint boundary conditions to calculate the three-dimensional spatial coordinates of the microseismic events.

[0007] In a preferred embodiment, the specific process of extracting radiated energy and corner frequencies from the frequency domain spectrum of a microseismic waveform sequence is as follows: The displacement spectrum is extracted by performing a fast Fourier transform on the microseismic waveform sequence; the ray propagation path attenuation factor and geometric diffusion factor are calculated by combining the three-dimensional deployment coordinates and three-dimensional spatial coordinates of the IoT microseismic sensor network nodes; the intrinsic source spectrum is inverted by performing dispersion compensation on the displacement spectrum using the ray propagation path attenuation factor and geometric diffusion factor; the low-frequency limiting platform displacement parameter and the high-frequency asymptote intersection frequency parameter of the intrinsic source spectrum are extracted; the radiated energy is deduced by integrating the low-frequency limiting platform displacement parameter, and the high-frequency asymptote intersection frequency parameter is calibrated as the corner frequency.

[0008] In a preferred embodiment, the specific process of extracting the initial amplitude ratio of P-waves and S-waves from the microseismic waveform sequence and reconstructing the focal mechanism solution using three-dimensional spatial coordinates is as follows: Firstly, the initial amplitude ratio parameter is obtained by analyzing the maximum amplitudes of the first-to-arrival P-waves and the maximum amplitudes of the first-to-arrival S-waves in the microseismic waveform sequence. The source emission angle and azimuth are deduced based on three-dimensional spatial coordinates and the three-dimensional deployment coordinates of the IoT microseismic sensor network nodes. A seismic moment tensor matrix is ​​constructed by combining the initial amplitude ratio parameter, the source emission angle, and the azimuth. The principal compressive stress axis vector and the principal stress axis vector are extracted by eigenvalue decomposition of the seismic moment tensor matrix. The rupture surface normal vector and slip angle are inverted based on the principal compressive stress axis vector and the principal stress axis vector to reconstruct the focal mechanism solution.

[0009] In a preferred embodiment, the specific process of screening tensile fracture events of the base plate according to the focal mechanism and calculating the dynamic stress drop of the tensile fracture events by integrating radiation energy, corner frequency and three-dimensional spatial coordinates is as follows: Obtain the reference elevation of the base plate of the mining face, and screen candidate microseismic events of the base plate strata by comparing the three-dimensional spatial coordinates with the reference elevation of the base plate; among the candidate microseismic events of the base plate strata, screen tensile fracture events of the base plate with sliding angles in the absolute tensile fracture critical angle range, and obtain the medium shear wave velocity and rock mass density of the rock strata region where the tensile fracture events of the base plate are located; calculate the seismic moment scalar by combining radiation energy, medium shear wave velocity and rock mass density, and perform spatial stress tensor attenuation extrapolation based on the seismic moment scalar and corner frequency to quantify the dynamic stress drop of the tensile fracture events of the base plate.

[0010] In a preferred embodiment, the specific process of obtaining the initial geostress tensor of the mining face, superimposing the dynamic stress drop onto the initial geostress tensor, and reconstructing the mining-induced additional stress field in the space where the tensile fracture event is located is as follows: obtaining the triaxial stress test parameters and the self-weight stress parameters of the floor rock mass from the exploration borehole of the mining face, and fitting the initial geostress tensor by combining the triaxial stress test parameters and the self-weight stress parameters of the floor rock mass; mapping the dynamic stress drop into a three-dimensional spatial perturbation tensor according to the normal vector and sliding angle of the fracture surface, and performing matrix superposition and deduction of the three-dimensional spatial perturbation tensor and the initial geostress tensor to reconstruct the mining-induced additional stress field covering the space where the tensile fracture event is located.

[0011] In a preferred embodiment, the specific process of extracting the local tensile principal stress direction vector of the mining-induced additional stress field, calculating the dot product projection of the fracture surface normal vector and the local tensile principal stress direction vector in the source mechanism solution, and fusing the dot product projection with dynamic stress reduction to quantify the intrinsic permeability increment is as follows: The mining-induced additional stress field is subjected to principal stress tensor eigenvalue decomposition to extract the third principal stress component, and the spatial orientation vector of the third principal stress component is calibrated as the local tensile principal stress direction vector; a spatial inner product operation is performed on the fracture surface normal vector and the local tensile principal stress direction vector to obtain the dot product projection; based on the dot product projection and a preset intrinsic elastic modulus, the effective opening parameter of the fracture surface in the tensile fracture event is calculated; the effective opening parameter of the fracture surface, the dynamic stress reduction, and the initial permeability parameter are combined to solve the rock mechanics fluid-structure coupling equation, quantifying the intrinsic permeability increment.

[0012] In a preferred embodiment, the process of performing three-dimensional spatial density clustering on tensile fracturing events based on intrinsic permeability increments to generate time-sequentially arranged water-conducting fracture clusters, and calculating the three-dimensional vector of centroid migration and rock mass volume expansion rate of the water-conducting fracture clusters within adjacent continuous time windows is as follows: The trigger time series of microseismic waveforms recorded by the Internet of Things (IoT) microseismic sensor network is acquired and adjacent continuous time windows are divided. Within adjacent continuous time windows, the intrinsic permeability increment is converted into a spatial weighting factor. High-permeability core points are extracted by weighted spatial density clustering of the three-dimensional spatial coordinates using the spatial weighting factor. Time-sequentially adjacent high-permeability core points are connected to generate time-sequentially arranged water-conducting fracture clusters. The three-dimensional envelope convex hull volume parameters and spatial geometric distribution center of the water-conducting fracture clusters within adjacent continuous time windows are measured. The three-dimensional centroid migration vector is constructed based on the spatial displacement offset of the spatial geometric distribution center within adjacent continuous time windows. The rock mass volume expansion rate is deduced based on the temporal evolution differences of the three-dimensional envelope convex hull volume parameters.

[0013] In a preferred embodiment, the specific process of obtaining the vertical approximation rate by projecting the centroid-migrating three-dimensional vector onto the normal direction of the bottom aquifer, and outputting a water hazard warning command corresponding to the three-dimensional spatial coordinates based on the vertical approximation rate and the rock mass volume expansion rate is as follows: Obtain the three-dimensional geological spatial distribution matrix of the hidden aquifer on the bottom of the mining face; analyze the normal vector of the bottom aquifer facing the mining face in the three-dimensional geological spatial distribution matrix; perform spatial orthogonal projection of the centroid-migrating three-dimensional vector onto the normal vector of the bottom aquifer to obtain the normal extension distance parameter; combine the normal extension distance parameter with the time interval step of adjacent continuous time windows to deduce the vertical approximation rate; map the vertical approximation rate and the rock mass volume expansion rate to a preset two-dimensional safety decision matrix for risk assessment; and output a water hazard warning command corresponding to the three-dimensional spatial coordinates based on the risk trigger level.

[0014] An intelligent early warning system based on mine microseismic data analysis is used to execute the aforementioned intelligent early warning method based on mine microseismic data analysis. The system includes: a microseismic analysis module, used to receive microseismic waveform sequences and coal mining machine spatial trajectories collected by an IoT microseismic sensor network, perform phase identification and travel time inversion on the microseismic waveform sequences, extract the three-dimensional spatial coordinates of microseismic events, and analyze the frequency domain spectrum of the microseismic waveform sequences to extract radiated energy and corner frequencies; a mechanism inversion module, used to extract the initial amplitude ratio of P-waves and S-waves in the microseismic waveform sequences, reconstruct the source mechanism solution by combining the three-dimensional spatial coordinates, screen floor tensile rupture events based on the source mechanism solution, and calculate the dynamic stress drop of tensile rupture events by integrating radiated energy, corner frequencies, and three-dimensional spatial coordinates; and a permeability calculation module, used to obtain the initial ground stress tension of the mining face. The dynamic stress reduction is superimposed onto the initial geostress tensor to reconstruct the mining-induced additional stress field in the space where the tensile rupture event is located. The direction vector of the local tensile principal stress in the mining-induced additional stress field is extracted. The dot product projection of the normal vector of the rupture surface and the direction vector of the local tensile principal stress in the focal mechanism solution is calculated. The dot product projection and the intrinsic permeability increment are fused together. The evolution early warning module is used to perform three-dimensional spatial density clustering of the tensile rupture event based on the intrinsic permeability increment, generate water-conducting fracture clusters arranged in time sequence, calculate the three-dimensional centroid migration vector and rock mass volume expansion rate of the water-conducting fracture clusters in adjacent continuous time windows, project the three-dimensional centroid migration vector onto the normal of the bottom aquifer to obtain the vertical approximation rate, and output the water hazard early warning command corresponding to the three-dimensional spatial coordinates based on the vertical approximation rate and the rock mass volume expansion rate.

[0015] The technical effects and advantages of this invention are as follows:

[0016] (1) An intelligent early warning method based on mine microseismic data analysis, after receiving microseismic waveforms and coal mining machine spatial trajectories collected by the Internet of Things microseismic sensor network, performs travel-time inversion and frequency domain spectrum analysis on microseismic events, extracts three-dimensional spatial coordinates, radiation energy and corner frequencies, and reconstructs the source mechanism solution by combining the initial amplitude ratio of P-waves and S-waves, thereby accurately screening out tensile fracture events of the floor from massive monitoring data to deduce dynamic stress drop. Through the above process, interference signals from massive non-water-conducting shear fractures or mining physical noise are eliminated, and the analysis dimension is deepened from simple vibration frequency statistics to the level of physical tensile fracture mechanism inside the rock strata, improving the mechanical relevance of data analysis and the identification accuracy of effective disaster-causing events, and improving the problem of high false alarm rate caused by the over-reliance on static threshold judgment in existing methods.

[0017] (2) An intelligent early warning system based on mine microseismic data analysis reconstructs the mining-induced additional stress field by superimposing dynamic stress drop and initial geostress tensor. It quantifies the intrinsic permeability increment by combining the dot product projection of the local tensile principal stress direction and the fracture surface normal. Based on this, it performs three-dimensional spatial density clustering of tensile fracture events to generate water-conducting fracture clusters, calculates the vertical approximation rate of the three-dimensional vector of its centroid migration to the normal projection of the bottom aquifer, and the volume expansion rate of the rock mass, and finally outputs the water hazard early warning command corresponding to the three-dimensional spatial coordinates. Through the above process, the scattered microseismic point data is transformed into permeability parameters and approximation rate parameters that characterize the connectivity of hidden water-conducting channels. It realizes the dynamic tracking of the evolution trajectory of the fracture group to the aquifer in three-dimensional space, improves the problem of fragmentation and dynamic perception lag in the existing data processing mode, and improves the physical coherence of early warning of mine water hazards and the timeliness of actual disaster prevention intervention.

[0018] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0019] Figure 1 This is a flowchart of an intelligent early warning method based on mine microseismic data analysis according to the present invention.

[0020] Figure 2 This is a schematic diagram of the three-dimensional spatial distribution of the water-conducting fracture cluster in an embodiment of the present invention;

[0021] Figure 3 This is a schematic diagram illustrating the evolution trend of the two-dimensional security decision matrix in an embodiment of the present invention;

[0022] Figure 4 This is a flowchart of an intelligent early warning system based on mine microseismic data analysis according to the present invention. Detailed Implementation

[0023] This application provides an intelligent early warning method and system based on mine microseismic data analysis. It solves the problems of existing technologies that rely solely on fixed thresholds to statistically analyze microseismic frequency or energy, making it difficult to distinguish between non-hydrothermal shear fractures and dangerous tensile fractures. Furthermore, it lacks tracking of the dynamic evolution trajectory of water-conducting fracture groups approaching aquifers in three-dimensional geological space, leading to frequent false alarms and delayed dynamic perception and response to water hazards.

[0024] The overall approach of the scheme in this application embodiment is as follows: First, it receives the microseismic waveform sequence and the spatial trajectory of the coal mining machine collected by the IoT microseismic sensor network, performs phase identification and travel time inversion on the microseismic waveform sequence, extracts the three-dimensional spatial coordinates of the microseismic event, and analyzes the frequency domain spectrum to extract the radiation energy and corner frequency; Second, it extracts the initial amplitude ratio of the P-wave and S-wave of the microseismic waveform sequence, reconstructs the focal mechanism solution by combining the three-dimensional spatial coordinates, screens out the tensile fracture event of the floor from the massive monitoring data based on the focal mechanism solution, and calculates the dynamic stress drop of the tensile fracture event by integrating the radiation energy, corner frequency, and three-dimensional spatial coordinates; Subsequently, it obtains the initial geostress tensor of the mining face, and superimposes the dynamic stress drop onto the... The initial geostress tensor is used to reconstruct the mining-induced additional stress field, and the direction vector of the local tensile principal stress is extracted. By calculating the dot product projection of the fracture surface normal vector and the direction vector of the local tensile principal stress in the focal mechanism solution, the intrinsic permeability increment of the tensile rupture event is deduced and quantified. Finally, the tensile rupture event is clustered in three-dimensional spatial density according to the intrinsic permeability increment to generate water-conducting fracture clusters arranged in time sequence. The three-dimensional centroid migration vector and rock mass volume expansion rate of the water-conducting fracture clusters in adjacent continuous time windows are calculated. The vertical approximation rate is obtained by projecting the three-dimensional centroid migration vector onto the normal of the bottom aquifer. Based on the vertical approximation rate and the rock mass volume expansion rate, the water hazard warning command corresponding to the three-dimensional spatial coordinates is output.

[0025] Example 1; please refer to Figure 1 This invention provides a technical solution: an intelligent early warning method based on mine microseismic data analysis, comprising the following steps: S1, receiving microseismic waveform sequences and coal mining machine spatial trajectories collected by an IoT microseismic sensor network, performing phase identification and travel time inversion on the microseismic waveform sequences, extracting the three-dimensional spatial coordinates of microseismic events, and analyzing the frequency domain spectrum of the microseismic waveform sequences to extract radiation energy and corner frequencies; S2, extracting the initial amplitude ratio of P-waves and S-waves of the microseismic waveform sequences, reconstructing the focal mechanism solution by combining the three-dimensional spatial coordinates, screening tensile fracture events of the floor according to the focal mechanism solution, and calculating the dynamic stress drop of tensile fracture events by fusing radiation energy, corner frequencies, and three-dimensional spatial coordinates; S3, obtaining the initial geostress tensile of the mining face, and superimposing the dynamic stress drop. Based on the initial geostress tensor, the mining-induced additional stress field in the space where the tensile rupture event is located is reconstructed. The direction vector of the local tensile principal stress in the mining-induced additional stress field is extracted. The dot product projection of the normal vector of the rupture surface and the direction vector of the local tensile principal stress in the focal mechanism solution is calculated. The dot product projection and the intrinsic permeability increment of dynamic stress reduction are fused. S4. Based on the intrinsic permeability increment, the tensile rupture event is subjected to three-dimensional spatial density clustering to generate water-conducting fracture clusters arranged in time sequence. The three-dimensional centroid migration vector and the rock mass volume expansion rate of the water-conducting fracture clusters in adjacent continuous time windows are calculated. The vertical approximation rate is obtained by projecting the three-dimensional centroid migration vector onto the normal of the bottom aquifer. Based on the vertical approximation rate and the rock mass volume expansion rate, the water hazard warning command corresponding to the three-dimensional spatial coordinates is output.

[0026] In this implementation plan, step S1 involves processing the acoustic signals actually collected downhole to determine the specific location and scale of rock strata fracturing. After acquiring the wave signals recorded by the IoT microseismic sensor network, the arrival times of P-waves and S-waves are identified and extracted from the microseismic waveform sequence for phase identification. Combined with the mine formation wave velocity model, the physical coordinates of the rock strata fracturing are inverted in three-dimensional space using the time difference between the arrival of P-waves and S-waves to achieve travel time inversion. Furthermore, the corner frequency is obtained by extracting the intersection frequency of the low-frequency horizontal segment and the high-frequency attenuation segment in the spectrum, and the actual scale of the rock mass micro-fractures is assessed by combining this with the radiant energy.

[0027] Step S2 involves filtering out fracturing events that would lead to tensile tearing of the floor from various disturbance signals generated by mining. The initial amplitude ratio parameter is extracted by obtaining the maximum amplitude ratio of the P-waves and S-waves when they first arrive at the observation point, thereby reconstructing the focal mechanism solution. Based on the reconstructed focal mechanism solution, tensile fracturing events that characterize tensile tearing of the floor strata are screened from the mining disturbance signals. Subsequently, multidimensional parameters are fused to calculate the dynamic stress drop, quantifying the shear stress difference released at the instant of tensile fracturing and the degree of degradation of the rock's mechanical properties near the fracture surface.

[0028] Step S3 involves calculating the increased permeability of the floor strata due to the opening of fractures caused by stress, based on the actual stress conditions of the coal face. The reconstructed additional stress field, obtained by superimposing the dynamic stress drop onto the initial geostress tensor, reflects the redistribution of the three-dimensional stress state within the floor strata caused by the coal machine cutting the coal seam. The local tensile principal stress direction vector is extracted from this stress field and projected onto the fracture surface normal vector using a dot product. This projection operation measures the consistency between the direction of fracture opening within the rock and the surrounding tensile stress direction from a three-dimensional geometric perspective. The quantified intrinsic permeability increment refers to the specific value of the increase in the permeability of the floor strata caused by the tensile fracturing of the rock skeleton.

[0029] Step S4 involves tracking the growth trajectory of water-conducting fractures based on changes in the permeability of the rock strata, assessing the risk of them connecting to concealed aquifers, and issuing an alarm. Three-dimensional spatial density clustering of fracture events based on intrinsic permeability increments identifies spatially dense sets of discrete fracture points with permeability characteristics as clusters of water-conducting fractures with a risk of water inrush. Calculating the three-dimensional centroid migration vector involves measuring the displacement direction and distance of the fracture cluster's geometric center within adjacent time periods, used to track the development trajectory of the fracture cluster. The rock mass volume expansion rate characterizes the rate at which the dense generation of microfractures leads to a relative increase in the volume of the underlying rock. The vertical approach rate refers to the rate at which the centroid of the fracture cluster grows vertically towards the underlying high-pressure aquifer. The system combines the vertical approach rate and the rock mass volume expansion rate to determine the critical state of water-conducting channel connection and outputs an early warning command bound to three-dimensional coordinates.

[0030] Specifically, the process of receiving microseismic waveform sequences and coal mining machine spatial trajectories collected by the Internet of Things (IoT) microseismic sensor network, performing phase identification and travel time inversion on the microseismic waveform sequences, and extracting the three-dimensional spatial coordinates of microseismic events is as follows: The arrival phases of the P-waves and S-waves in the microseismic waveform sequences and the three-dimensional deployment coordinates of the IoT microseismic sensor network nodes are obtained; a theoretical travel time residual objective function is constructed by combining the arrival phases of the P-waves and S-waves with the three-dimensional deployment coordinates; a priori spatial constraint boundary is established based on the coal mining machine spatial trajectory; and nonlinear inversion optimization is performed on the theoretical travel time residual objective function under the priori spatial constraint boundary conditions to calculate the three-dimensional spatial coordinates of the microseismic events.

[0031] In this implementation scheme, when extracting the three-dimensional spatial coordinates of microseismic events, in order to eliminate the interference caused by the uncertainty of the earthquake occurrence time, the actual phase difference is calculated based on the arrival phases of the P-waves and S-waves recorded by each sensor node, and then a theoretical travel time residual objective function is constructed: ;in, The residual objective function value is represented by M, which represents the total number of triggered microseismic sensing nodes; m represents the node index. This represents the spatial positioning weight coefficient of the m-th node, and its specific determination method is as follows: , The waveform attenuation constant is This represents the signal-to-noise ratio of the corresponding waveform sequence; This indicates the difference between the actual arrival times of the transverse wave and the longitudinal wave recorded at the m-th node; and X, Y, and Z represent the equivalent propagation velocities of longitudinal and transverse waves in the rock strata at the bottom of the mine, respectively; X, Y, and Z represent the three-dimensional spatial coordinate variables of the microseismic event to be solved. This represents the 3D deployment coordinates of the m-th node. Subsequently, the current real-time center coordinates of the coal mining machine are extracted. The a priori spatial constraint boundary conditions for mining-induced failure are set as follows: ,in The maximum disturbance radius for mining operations is pre-determined based on the mechanical properties of the top and bottom rock strata. Under this boundary condition constraint, a nonlinear simplex algorithm is used to perform local mesh iterative optimization of the above-mentioned theoretical travel-time residual objective function, so that the function value... The solution converges and reaches the global minimum. The coordinates of the optimal solution at this point are the three-dimensional spatial coordinates of the microseismic event output.

[0032] Specifically, the process of extracting radiated energy and corner frequencies from the frequency domain spectrum of a microseismic waveform sequence is as follows: A fast Fourier transform is performed on the microseismic waveform sequence to extract the displacement spectrum. The ray propagation path attenuation factor and geometric diffusion factor are calculated using the three-dimensional deployment coordinates and three-dimensional spatial coordinates of the IoT microseismic sensor network nodes. The intrinsic source spectrum is then inverted using the ray propagation path attenuation factor and geometric diffusion factor through dispersion compensation. The low-frequency limiting platform displacement parameter and the high-frequency asymptote intersection frequency parameter of the intrinsic source spectrum are extracted. The radiated energy is deduced based on the integral of the low-frequency limiting platform displacement parameter, and the high-frequency asymptote intersection frequency parameter is calibrated as the corner frequency.

[0033] In this implementation scheme, frequency domain spectral analysis is used to eliminate the attenuation effect of complex geological environments on signals in order to restore the physical parameters of the seismic source. After performing a fast Fourier transform on the waveform sequence to obtain the observed displacement spectrum, a dispersion compensation equation is constructed: ;in, The amplitude of the intrinsic source displacement spectrum after dispersion compensation is represented by f; f represents the frequency component. Indicates the amplitude of the observed displacement spectrum; The geometric diffusion factor corresponding to node k is obtained by calculating the Euclidean distance between the node's three-dimensional deployment coordinates and the three-dimensional spatial coordinates calculated above. The medium quality factor represents the propagation path of node k; This represents the equivalent group velocity of seismic waves in the underlying rock strata. After obtaining the intrinsic source spectrum, the spectrum curve is fitted in a double logarithmic coordinate system, and the horizontal smooth asymptote intercept in the low-frequency band is extracted as the low-frequency limiting platform displacement parameter. The frequency offset point corresponding to the intersection of the horizontal asymptote and the tangent of the high-frequency attenuation segment is extracted and directly calibrated as the corner frequency. Finally, based on the rock mass mechanics and physical fracture integral, the radiated energy is derived, and the calculation formula is as follows: ;in, Parameters representing the radiated energy of microseismic events; This indicates the density parameter of the rock mass at the bottom of the mine. The source radiation mode coefficient is determined by statistical inversion of the three-dimensional source mechanism solutions of historical mine floor fault rupture samples to obtain the mean value. This represents the extracted low-frequency limit platform displacement parameter; This represents the calibrated corner frequency parameter. The above process maps the frequency domain attenuation waveform characteristics into an absolute physical quantity reflecting the energy release from microfractures in the rock.

[0034] Specifically, the process of extracting the initial amplitude ratio of P-waves and S-waves from the microseismic waveform sequence and reconstructing the focal mechanism solution using three-dimensional spatial coordinates is as follows: First, the initial amplitude ratio parameter is obtained by analyzing the maximum amplitudes of the first-to-first P-waves and the first-to-first S-waves in the microseismic waveform sequence. The source emission angle and azimuth are deduced based on three-dimensional spatial coordinates and the three-dimensional deployment coordinates of the IoT microseismic sensor network nodes. A seismic moment tensor matrix is ​​constructed by combining the initial amplitude ratio parameter, the source emission angle, and the azimuth. The principal compressive stress axis vector and the principal stress axis vector are extracted by eigenvalue decomposition of the seismic moment tensor matrix. The rupture surface normal vector and slip angle are inverted based on the principal compressive stress axis vector and the principal stress axis vector to reconstruct the focal mechanism solution.

[0035] In this implementation scheme, after acquiring the waveform sequences recorded by each microseismic sensor node, the maximum amplitudes of the first-arrival P-wave and the first-arrival S-wave are extracted using the long-short time window averaging ratio method, and the initial amplitude ratio parameter is obtained by quotient calculation. Combining the three-dimensional spatial coordinates of the microseismic event output with the three-dimensional deployment coordinates of the IoT microseismic sensor network nodes, the spatial propagation path of the seismic wave within the rock strata is inferred using a ray tracing algorithm, thereby determining the source emission angle and azimuth. Subsequently, the initial amplitude ratio parameter, source emission angle, and azimuth are substituted into the point-source elastic dislocation model to construct a seismic moment tensor matrix reflecting the rock fracture mechanics mechanism. ;in, Represents the seismic moment tensor matrix of the optimal match; U represents the total number of effective microseismic sensor nodes participating in the mechanism inversion; u represents the effective sensor node index parameter; This represents the initial amplitude ratio parameter extracted at the u-th sensing node; This represents the source emission angle parameter from the source to the u-th sensing node; This represents the azimuth parameter from the earthquake source to the u-th sensing node; Represents the three-dimensional mechanical tensor variables to be optimized; and Let represent the directional attenuation functions of the theoretical radiation patterns of P-waves and S-waves, respectively, given tensor variables. Obtain the optimal matching seismic moment tensor matrix. Then, eigenvalue decomposition is performed on it, and the unit eigenvector corresponding to the smallest negative eigenvalue is extracted as the principal compressive stress axis vector. Extract the unit eigenvector corresponding to the largest positive eigenvalue as the assertion stress axis vector. Finally, based on the conjugate shear fracture theory, the principal stress axis vectors are orthogonally transformed to calculate the normal vector of the fracture surface. The angle between the slip vector on the fault plane and the fault strike is analyzed to invert the slip angle parameter. This completes the underlying physical reconstruction of the focal mechanism solution for microseismic events.

[0036] Specifically, the process of screening tensile fracture events in the base plate based on the focal mechanism and calculating the dynamic stress drop of tensile fracture events by integrating radiation energy, corner frequency, and three-dimensional spatial coordinates is as follows: Obtain the baseline elevation of the base plate of the mining face; compare the three-dimensional spatial coordinates with the baseline elevation to screen candidate microseismic events in the base plate strata; among the candidate microseismic events in the base plate strata, screen tensile fracture events with a slip angle within the critical angle range of absolute tensile fracture; obtain the medium shear wave velocity and rock mass density of the stratum region where the tensile fracture event is located; calculate the seismic moment scalar by combining radiation energy, medium shear wave velocity, and rock mass density; and deduce the spatial stress tensor attenuation based on the seismic moment scalar and corner frequency to quantify the dynamic stress drop of the tensile fracture event in the base plate.

[0037] In this implementation plan, by reading the reference elevation of the floor of the mining face, the vertical spatial coordinates of microseismic events are numerically compared with the reference elevation of the floor. Unrelated vibrations occurring in the roof or inside the coal seam are eliminated, and candidate microseismic events located in the hidden space of the floor strata are screened out. Based on this, the slip angle parameter of each candidate microseismic event is extracted. The critical angle range for absolute tensile fracture is set as follows: ;in, The critical tolerance angle parameter is determined by obtaining bottom rock cores from mine exploration boreholes and linearly calibrating them proportionally based on the laboratory test ratio of the tensile strength to shear strength of the bottom rock mass. After screening out bottom tensile rupture events where the sliding angle enters this interval, the medium shear wave velocity parameter and rock mass density parameter of the specific rock strata region in the bottom plate where the rupture event is located are extracted. Combined with the radiated energy extracted in the previous steps, the seismic moment scalar is calculated by solving the rock mass elastoplastic dissipative constitutive equation. ;in, A seismic moment scalar representing a tensile rupture event in the foundation slab; This represents the rock mass density parameter indicating the region where the tensile fracture event of the floor slab occurred. This represents the corresponding medium shear wave velocity parameter; This represents the radiation energy parameter extracted in the preceding sequence; This indicates the corner frequency parameter specified in the previous calibration. This represents the radiation constant based on the anisotropy correction of the rock mass. Subsequently, based on the Madariaga disk fracture model, spatial stress tensor decay is extrapolated to quantify the dynamic stress drop during tensile fracture events in the floor slab. ;in, The parameter representing the dynamic stress drop at the fracture surface at the instant of a tensile fracture event in the base plate; This represents the disk model correction coefficient related to the propagation velocity of the seismic source rupture. The above process converts the kinematic and spectral parameters into absolute dynamic parameters that reflect the true tensile failure degree of the underlying rock strata, thus completing the quantitative solution of the mechanical source data.

[0038] Specifically, the process of obtaining the initial geostress tensor of the mining face, superimposing the dynamic stress drop onto the initial geostress tensor, and reconstructing the mining-induced additional stress field in the space where the tensile fracture event is located is as follows: The triaxial stress test parameters and the self-weight stress parameters of the floor rock mass from the exploration boreholes of the mining face are obtained; the initial geostress tensor is fitted by combining the triaxial stress test parameters and the self-weight stress parameters of the floor rock mass; the dynamic stress drop is mapped into a three-dimensional spatial perturbation tensor based on the fracture surface normal vector and the sliding angle; the three-dimensional spatial perturbation tensor and the initial geostress tensor are matrix superimposed and deduced to reconstruct the mining-induced additional stress field covering the space where the tensile fracture event is located.

[0039] In this implementation plan, the measured maximum and minimum horizontal principal stress parameters from borehole stress gauges recorded during the exploration phase of the mining face are extracted. Combined with the borehole depth and the average density of the overlying strata, the self-weight stress parameter of the floor rock mass is calculated. Based on the above triaxial stress state components, an initial geostress tensor in the form of a third-order symmetric matrix is ​​constructed in the global geographic coordinate system. The dynamic stress drop value and fracture surface normal vector calculated in the previous steps are obtained. Based on the geometric slip characteristics in the focal mechanism solution, the sliding direction vector of the fracture surface is deduced. Through vector operations, the scalar form of dynamic stress drop is mapped to a three-dimensional spatial perturbation tensor. The calculation formula is as follows: ;in, This represents the three-dimensional spatial perturbation tensor generated by the mapping; This represents the dynamic stress drop parameter extracted in the previous steps. This represents the normal vector of the rupture surface reconstructed in the preceding steps; This represents the deduced slip direction vector of the fracture surface; This represents the tensor product operator for spatial vectors. After obtaining the three-dimensional spatial perturbation tensor, it is added to the initial geostress tensor using the same coordinate system. ;in, This represents the reconstructed additional stress field tensor from mining operations; This represents the initial geostress tensor of the fit. This step, through matrix tensor superposition and derivation, directly quantifies the underlying physical stress state under the combined effects of mining disturbance and rock microfracture.

[0040] Specifically, the process of extracting the local tensile principal stress direction vector of the mining-induced additional stress field, calculating the dot product projection of the fracture surface normal vector and the local tensile principal stress direction vector in the source mechanism solution, and integrating the dot product projection with the dynamic stress reduction to quantify the intrinsic permeability increment is as follows: The third principal stress component is extracted by principal stress tensor eigenvalue decomposition of the mining-induced additional stress field, and the spatial orientation vector of the third principal stress component is calibrated as the local tensile principal stress direction vector; the dot product projection is obtained by performing a spatial inner product operation on the fracture surface normal vector and the local tensile principal stress direction vector; the effective opening parameter of the fracture surface in the tensile fracture event is calculated based on the dot product projection and the preset intrinsic elastic modulus; the rock mechanics fluid-structure interaction equation is solved by combining the effective opening parameter of the fracture surface, the dynamic stress reduction, and the initial permeability parameter to quantify the intrinsic permeability increment.

[0041] In this implementation scheme, the Jacobi iterative eigenvalue decomposition is performed on the tensor of the mining-induced additional stress field obtained above. The three real eigenvalues ​​of the tensor matrix are extracted and arranged in descending order of algebraic value. The eigenvalue with the smallest value is taken as the third principal stress component. The orthogonal eigenvector corresponding to this eigenvalue is extracted and calibrated as the local tensile principal stress direction vector. The fracture surface normal vector is extracted and its spatial inner product is performed with the local tensile principal stress direction vector. The absolute value of the cosine of the angle between the two is obtained as the dot product projection parameter. Based on the dot product projection parameter, the effective opening parameter of the fracture surface in the tensile fracture event is calculated. The calculation formula is as follows: ;in, The parameter representing the effective opening of the fracture surface in a tensile fracture event; This indicates the intrinsic elastic modulus parameter of the pre-set base rock. This represents the characteristic length parameter of the fracture calculated based on the microseismic radiation energy. This parameter represents the projection of the dot product of the fracture surface normal vector and the direction vector of the local tensile principal stress. Combined with the effective fracture opening parameter, a three-dimensional rock mechanics fluid-structure interaction equation based on the cubic law is used to quantify the intrinsic permeability increment: ;in, This represents the quantified increment of intrinsic penetration rate; This parameter represents the initial permeability of the rock strata at the bottom of the mining face. The parameter representing the initial average natural pore size of the base strata is determined by statistically analyzing the arithmetic mean of mercury intrusion porosimetry test data obtained from core samples taken during geological exploration. This represents the parameter indicating the pore compressibility coefficient of the rock mass. This represents the in-situ fluid pressure parameter of the concealed aquifer in the base plate. The calculation process, through multi-parameter nonlinear equation mapping, achieves a quantitative transformation from stress field disturbances to changes in the water-conductivity properties of the microscopic rock strata.

[0042] Specifically, the process of performing three-dimensional spatial density clustering on tensile fracturing events based on intrinsic permeability increments to generate time-series arranged water-conducting fracture clusters, and calculating the three-dimensional vector of centroid migration and rock mass volume expansion rate of the water-conducting fracture clusters within adjacent continuous time windows is as follows: The trigger time series of microseismic waveforms recorded by the Internet of Things microseismic sensor network is acquired and adjacent continuous time windows are divided. Within adjacent continuous time windows, the intrinsic permeability increment is converted into a spatial weighting factor. High-permeability core points are extracted by weighted spatial density clustering of the three-dimensional spatial coordinates using the spatial weighting factor. Temporally adjacent high-permeability core points are connected to generate time-series arranged water-conducting fracture clusters. The volume parameters of the three-dimensional envelope convex hull and the spatial geometric distribution center of the water-conducting fracture clusters within adjacent continuous time windows are measured. A three-dimensional vector of centroid migration is constructed based on the spatial displacement offset of the spatial geometric distribution center within adjacent continuous time windows. The rock mass volume expansion rate is inferred based on the temporal evolution differences of the three-dimensional envelope convex hull volume parameters.

[0043] In this implementation scheme, the microseismic waveform trigger time series recorded by the clock synchronization of the IoT microseismic sensor network system is first read, and multiple adjacent continuous time windows are divided according to a preset fixed step size. Within a specific time window, the intrinsic permeability increment of each tensile rupture event is extracted and converted into a spatial weighting factor. The calculation formula is as follows: ;in, Indicates the first Spatial weighting factor for individual tensile rupture events; The parameter representing the sequence number of the tensile rupture event within the current time window; This represents the nonlinear scaling factor parameter; Indicates the first Intrinsic permeability increment parameter corresponding to a single tensile rupture event; This represents the reference permeability baseline parameter for the undamaged rock strata at the base. Subsequently, weighted spatial density clustering is performed on the three-dimensional spatial coordinates based on this spatial weighting factor, defining the local weighted density for each rupture event as: ;in, Indicates the first Local weighted density parameters for individual tensile rupture events; Indicates the first Centered on one event The spatial neighborhood set parameter is the radius; The parameter represents the event sequence number within the spatial neighborhood set; Indicates the first Spatial weighting factors for each event; Indicates an event With the event The Euclidean distance parameter between them; The preset density search radius parameter is determined by taking half of the pressure step distance of the mining face floor period. Data points with local weighted density parameters greater than a set threshold are extracted as high-permeability core points, and core points within adjacent time windows are spatially connected to form water-conducting fracture clusters with an evolutionary time sequence. For water-conducting fracture clusters within a specific time window, the spatial Deloni triangulation algorithm is used to calculate the three-dimensional envelope convex hull volume parameter of its outer contour, and the weighted spatial geometric distribution center is calculated. ;in, This represents the coordinate vector of the center of the spatial geometric distribution of the water-conducting fracture cluster within the current continuous time window; Indicates the first The three-dimensional spatial coordinate vectors of each event are used. A three-dimensional centroid migration vector is constructed by calculating the difference between the coordinate vectors of the spatial geometric distribution centers of the current time window and the previous adjacent time window, while simultaneously estimating the rock mass volume expansion rate. ;in, This parameter represents the volumetric expansion rate of the rock mass. This represents the volume parameter of the 3D envelope convex hull within the current continuous time window; This represents the three-dimensional envelope convex hull volume parameter within the previous consecutive time window; This parameter represents the time interval step size between adjacent consecutive time windows. Combined with... Figure 2 As shown, after the system performs three-dimensional spatial density clustering using intrinsic permeability increment as the spatial weighting factor, the discretely distributed light-colored nodes in the figure represent precisely screened tensile fracturing events of the substrate, and their dense spatial aggregation forms a cluster of water-conducting fractures. The calculated weighted spatial geometric distribution center (i.e., Figure 2 The core centroid (high-brightness display) and its displacement trajectory within a continuous time window intuitively represent the cluster evolution characteristics of microseismic rupture points in the concealed space of the mine floor, thus providing spatial geometric support for subsequent calculation of the vertical approximation rate to the concealed aquifer.

[0044] Specifically, the process of obtaining the vertical approximation rate by projecting the centroid-migrating 3D vector onto the normal of the bottom aquifer, and outputting a water hazard warning command corresponding to the 3D spatial coordinates based on the vertical approximation rate and the rock mass volume expansion rate is as follows: Obtain the 3D geological spatial distribution matrix of the hidden aquifer on the bottom of the mining face; analyze the normal vector of the bottom aquifer facing the mining face in the 3D geological spatial distribution matrix; perform spatial orthogonal projection of the centroid-migrating 3D vector onto the normal vector of the bottom aquifer to obtain the normal extension distance parameter; combine the normal extension distance parameter with the time interval step of adjacent continuous time windows to deduce the vertical approximation rate; map the vertical approximation rate and the rock mass volume expansion rate to a preset 2D safety decision matrix for risk assessment; and output a water hazard warning command corresponding to the 3D spatial coordinates based on the risk trigger level.

[0045] In this implementation plan, the three-dimensional geological exploration data of the bottom plate of the mining face is retrieved to obtain the three-dimensional geological spatial distribution matrix of the concealed aquifer. The partial derivatives of the three-dimensional surface function of the aquifer roof represented by this matrix are solved to analytically determine the normal vector of the bottom plate aquifer strictly facing the mining face side. The calculated centroid migration three-dimensional vector is then spatially orthogonally projected onto the bottom plate aquifer normal vector to obtain the normal extension distance parameter reflecting the absolute distance of the fracture community approaching the water body. The calculation formula is as follows: ;in, This represents the normal extension distance parameter obtained by orthogonal projection into space. This represents the three-dimensional vector representing the centroid migration constructed in the preceding sequence; This represents the analytically derived normal vector of the aquifer at the bottom of the plate; The parameter representing the modulus of the normal vector of the aquifer at the bottom plate; This represents the dot product operator for three-dimensional vectors. Combined with the normal extension distance parameter, the vertical approximation rate can be derived: ;in, This represents the vertical approach rate parameter indicating the development of the water-conducting fracture cluster towards the aquifer at the bottom. Subsequently, the system calls a preset two-dimensional safety decision matrix. This matrix uses the vertical approach rate as the horizontal axis and the rock mass volume expansion rate as the vertical axis, dividing the system into four discrete risk triggering level intervals: normal disturbance, fracture development, critical water inrush, and extreme danger. The interval boundaries of this decision matrix are linearly calibrated based on the peak approach rate and expansion rate limit of historical water inrush samples from the mine. The two-dimensional feature points formed by the currently calculated vertical approach rate parameter and the rock mass volume expansion rate parameter are mapped onto this two-dimensional safety decision matrix to determine the specific interval region where the point falls, thereby extracting the corresponding risk triggering level. When the extracted risk triggering level reaches the critical water inrush or extreme danger state, the coordinate vector of the current spatial geometric distribution center of the water-conducting fracture cluster is extracted, and a water hazard warning command containing precise three-dimensional spatial coordinates and the corresponding risk triggering level is output. Combined with... Figure 3 As shown, the system constructs a two-dimensional safety decision matrix with the vertical approach rate as the horizontal axis and the rock mass volume expansion rate as the vertical axis, dividing it into a normal disturbance zone, a fracture development zone, a water inrush critical zone, and an extremely dangerous zone. The connected data points in the figure represent the dynamic evolution trajectory of water-conducting fracture clusters within adjacent consecutive time windows (such as time windows 1 to 5). As time progresses, if the centroid approaches the bottom aquifer at a faster rate and the expansion rate caused by the expansion of microfractures in the rock mass increases, the feature points will migrate across the interval boundaries to the dangerous zone in the upper right, thereby realizing intelligent decision-making from the evolution of bottom-level data to intuitive disaster prevention intervention.

[0046] Example 2; please refer to Figure 4An intelligent early warning system based on mine microseismic data analysis is used to execute an intelligent early warning method based on mine microseismic data analysis described in the embodiments. The system includes: a microseismic analysis module, used to receive microseismic waveform sequences and coal mining machine spatial trajectories collected by an IoT microseismic sensor network, perform phase identification and travel time inversion on the microseismic waveform sequences, extract the three-dimensional spatial coordinates of microseismic events, and analyze the frequency domain spectrum of the microseismic waveform sequences to extract radiation energy and corner frequencies; a mechanism inversion module, used to extract the initial amplitude ratio of P-waves and S-waves of the microseismic waveform sequences, reconstruct the source mechanism solution by combining the three-dimensional spatial coordinates, screen the tensile fracture events of the floor according to the source mechanism solution, and calculate the dynamic stress drop of the tensile fracture events by integrating radiation energy, corner frequencies, and three-dimensional spatial coordinates; and a permeability calculation module, used to obtain the initial ground surface of the mining face. The stress tensor is used to superimpose dynamic stress drop onto the initial geostress tensor to reconstruct the mining-induced additional stress field in the space where the tensile rupture event is located. The direction vector of the local tensile principal stress in the mining-induced additional stress field is extracted. The dot product projection of the normal vector of the rupture surface and the direction vector of the local tensile principal stress in the focal mechanism solution is calculated. The dot product projection and the intrinsic permeability increment of dynamic stress drop are fused together. The evolution early warning module is used to perform three-dimensional spatial density clustering of tensile rupture events based on the intrinsic permeability increment, generate water-conducting fracture clusters arranged in time sequence, calculate the three-dimensional centroid migration vector and rock mass volume expansion rate of the water-conducting fracture clusters in adjacent continuous time windows, project the three-dimensional centroid migration vector onto the normal of the bottom aquifer to obtain the vertical approximation rate, and output the water hazard early warning command corresponding to the three-dimensional spatial coordinates based on the vertical approximation rate and the rock mass volume expansion rate.

[0047] In this implementation scheme, the microseismic analysis module, as the front-end sensing data processing unit of the system, directly connects to the underground IoT microseismic sensor network to acquire basic waveform sequences. After receiving continuous waveform data and the spatial trajectory of the coal mining machine, the module performs phase identification and travel time inversion calculations through its internally configured positioning algorithm program to accurately calculate the three-dimensional spatial coordinates of the microseismic event. Simultaneously, the module calls the Fast Fourier Transform component to perform frequency domain transformation on the time-domain waveform, fitting and extracting the radiation energy and corner frequency parameters from the spectrum, thereby completing the automatic conversion of the original physical signal into basic spatial and energy dimension data.

[0048] The mechanism inversion module receives the spatiotemporal and energy parameters output from the front end and is responsible for calculating the physical mechanism of the underlying rupture. After extracting the initial amplitude ratio of the waveform, this module constructs a seismic moment tensor matrix within the system using three-dimensional spatial coordinates and performs eigenvalue decomposition to reconstruct the focal mechanism solution. Subsequently, based on the geometric slip characteristics in the mechanism solution, the module runs a screening program to isolate tensile rupture events with tensile-tear properties from a massive dataset. It then calls the corresponding dynamic derivation equations to quantitatively calculate the dynamic stress drop value at the occurrence of this specific tensile rupture event.

[0049] The permeability calculation module is used to perform coupled calculations of the disturbance of the underlying stress field and the changes in the water conductivity of the rock. This module reads the initial geostress tensor of the mining face pre-stored in the system database, performs tensor mapping on the dynamic stress drop output by the previous module, and performs matrix superposition operations to reconstruct the real-time mining-induced additional stress field in the system. Next, the module performs principal stress eigenvalue decomposition on the additional stress field to obtain the local tensile principal stress direction vector. The control processor performs a spatial dot product operation between this vector and the fracture surface normal vector, and finally, combined with the set rock mechanics equations, quantitatively outputs the intrinsic permeability increment of the tensile fracture event.

[0050] The evolutionary early warning module, serving as the system's terminal decision-making and disaster prevention command issuance unit, is responsible for transforming discrete mechanical parameters into macroscopic physical disaster prevention information. This module uses the intrinsic permeability increment as a weighting parameter to perform a three-dimensional spatial density clustering algorithm on tensional fracturing events, generating a time-series spatial model of water-conducting fracture clusters in computational memory. Subsequently, the module calculates the three-dimensional vector of the centroid migration and the rock mass volume expansion rate of the fracture clusters within adjacent time windows, retrieves the spatial distribution matrix of the hidden aquifer in the base plate, calculates the vertical approximation rate of the centroid towards the aquifer boundary through spatial orthogonal projection, and finally combines the numerical values ​​with a preset safety decision matrix to generate and output a water hazard early warning command bound to specific three-dimensional spatial coordinates to external devices.

[0051] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0052] This invention is described with reference to flowchart illustrations and / or block diagrams of systems, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0053] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0054] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0055] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0056] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. An intelligent early warning method based on mine microseismic data analysis, characterized in that, Includes the following steps: S1. Receive the micro-seismic waveform sequence and the spatial trajectory of the coal mining machine collected by the Internet of Things micro-seismic sensor network, perform phase identification and travel time inversion on the micro-seismic waveform sequence, extract the three-dimensional spatial coordinates of the micro-seismic event, and analyze the frequency domain spectrum of the micro-seismic waveform sequence to extract the radiation energy and corner frequency. S2. Extract the initial amplitude ratio of P-waves and S-waves from the microseismic waveform sequence, reconstruct the focal mechanism solution by combining it with three-dimensional spatial coordinates, screen the tensile rupture events of the base plate according to the focal mechanism solution, and calculate the dynamic stress drop of the tensile rupture events by integrating radiation energy, corner frequency and three-dimensional spatial coordinates. S3. Obtain the initial geostress tensor of the mining face, superimpose the dynamic stress drop onto the initial geostress tensor, reconstruct the mining-induced additional stress field in the space where the tensile rupture event is located, extract the local tensile principal stress direction vector of the mining-induced additional stress field, calculate the dot product projection of the rupture surface normal vector and the local tensile principal stress direction vector in the source mechanism solution, and fuse the dot product projection with the dynamic stress drop to quantify the intrinsic permeability increment. S4. Based on the intrinsic permeability increment, perform three-dimensional spatial density clustering of tensile fracture events to generate water-conducting fracture clusters arranged in time sequence. Calculate the three-dimensional centroid migration vector and rock mass volume expansion rate of the water-conducting fracture clusters in adjacent continuous time windows. Project the three-dimensional centroid migration vector onto the normal direction of the bottom aquifer to obtain the vertical approximation rate. Based on the vertical approximation rate and rock mass volume expansion rate, output the corresponding three-dimensional spatial coordinate water hazard warning command.

2. The intelligent early warning method based on mine microseismic data analysis according to claim 1, characterized in that: The specific process of receiving microseismic waveform sequences and coal mining machine spatial trajectories collected by the Internet of Things (IoT) microseismic sensor network, performing phase identification and travel time inversion on the microseismic waveform sequences, and extracting the three-dimensional spatial coordinates of microseismic events is as follows: Obtain the arrival time phases of the P-waves and S-waves of the microseismic waveform sequence and the three-dimensional deployment coordinates of the IoT microseismic sensor network nodes. Combine the arrival time phases of the P-waves and S-waves with the three-dimensional deployment coordinates to construct a theoretical travel time residual objective function. Based on the spatial trajectory of the coal mining machine, a priori spatial constraint boundary is established. Under the condition of the priori spatial constraint boundary, the objective function of the theoretical travel time residual is optimized by nonlinear inversion to solve the three-dimensional spatial coordinates of the microseismic event.

3. The intelligent early warning method based on mine microseismic data analysis according to claim 1 or 2, characterized in that: The specific process of extracting radiated energy and corner frequencies from the frequency domain spectrum of a microseismic waveform sequence is as follows: The displacement spectrum is extracted by performing a fast Fourier transform on the microseismic waveform sequence, and the ray propagation path attenuation factor and geometric diffusion factor are calculated by combining the three-dimensional deployment coordinates and three-dimensional spatial coordinates of the IoT microseismic sensor network nodes. The intrinsic source spectrum is inverted by using the ray propagation path attenuation factor and geometric diffusion factor to perform dispersion compensation on the displacement spectrum, and the low-frequency limit plateau displacement parameter and the high-frequency asymptote intersection frequency parameter of the intrinsic source spectrum are extracted. The radiated energy is derived by integrating the displacement parameters of the low-frequency limit platform, and the frequency parameters of the intersection points of the high-frequency asymptotes are calibrated as the corner frequencies.

4. The intelligent early warning method based on mine microseismic data analysis according to claim 1, characterized in that: The specific process of extracting the P-wave and S-wave initial amplitude ratios from the microseismic waveform sequence and reconstructing the focal mechanism solution using three-dimensional spatial coordinates is as follows: The initial amplitude ratio parameter is obtained by analyzing the maximum amplitude of the first arrival P-wave and the maximum amplitude of the first arrival S-wave in the microseismic waveform sequence. The source emission angle and azimuth are deduced based on the three-dimensional spatial coordinates and the three-dimensional deployment coordinates of the IoT microseismic sensor network nodes. A seismic moment tensor matrix is ​​constructed by combining the initial amplitude ratio parameter, source emission angle and azimuth angle. The principal compressive stress axis vector and principal stress axis vector are extracted by eigenvalue decomposition of the seismic moment tensor matrix. The source mechanism solution is reconstructed by inverting the normal vector and slip angle of the rupture surface based on the principal compressive stress axis vector and the principal stress axis vector.

5. The intelligent early warning method based on mine microseismic data analysis according to claim 1 or 4, characterized in that: The specific process for calculating the dynamic stress drop of tensile rupture events based on focal mechanism analysis and by integrating radiated energy, corner frequency, and three-dimensional spatial coordinates is as follows: Obtain the reference elevation of the bottom plate of the mining face, and compare the three-dimensional spatial coordinates with the reference elevation of the bottom plate to screen candidate microseismic events in the bottom plate rock strata; Among the candidate microseismic events in the base strata, screening for base tensile fracture events with slip angles within the critical range of absolute tensile fracture angles is conducted, and the medium shear wave velocity and rock mass density of the strata region where the base tensile fracture events are located are obtained. By combining radiant energy, medium shear wave velocity, and rock mass density, the seismic moment scalar is calculated. Based on the seismic moment scalar and corner frequency, the spatial stress tensor attenuation is extrapolated, and the dynamic stress drop of the tensile rupture event of the base plate is quantified.

6. The intelligent early warning method based on mine microseismic data analysis according to claim 1, characterized in that: The specific process of obtaining the initial geostress tensor of the mining face, superimposing the dynamic stress drop onto the initial geostress tensor, and reconstructing the mining-induced additional stress field in the space where the tensile fracture event occurs is as follows: The triaxial stress test parameters and the self-weight stress parameters of the bottom rock mass of the mining face exploration borehole are obtained, and the initial geostress tensor is fitted by combining the triaxial stress test parameters and the self-weight stress parameters of the bottom rock mass. Based on the normal vector of the rupture surface and the slip angle, the dynamic stress drop is mapped into a three-dimensional spatial perturbation tensor. The three-dimensional spatial perturbation tensor is then superimposed with the initial geostress tensor to reconstruct the mining-induced additional stress field in the space where the tensile rupture event is located.

7. The intelligent early warning method based on mine microseismic data analysis according to claim 1, characterized in that: The specific process of extracting the local tensile principal stress direction vector of the mining-induced additional stress field, calculating the dot product projection of the rupture surface normal vector and the local tensile principal stress direction vector in the focal mechanism solution, and fusing the dot product projection with the intrinsic permeability increment quantified by dynamic stress reduction is as follows: The third principal stress component is extracted by principal stress tensor eigenvalue decomposition of the additional stress field caused by mining, and the spatial orientation vector of the third principal stress component is calibrated as the direction vector of the local tensile principal stress. The dot product projection is obtained by performing a spatial inner product operation on the fracture surface normal vector and the local tensile principal stress direction vector. Based on the dot product projection and the preset intrinsic elastic modulus, the effective opening parameter of the fracture surface in the tensile fracture event is calculated. The rock mechanics fluid-structure interaction equations were solved by combining the effective opening parameter of the fracture surface, the dynamic stress drop and the initial permeability parameter to quantify the intrinsic permeability increment.

8. The intelligent early warning method based on mine microseismic data analysis according to claim 1, characterized in that: Based on the intrinsic permeability increment, three-dimensional spatial density clustering of tensile fracturing events is performed to generate time-series arranged water-conducting fracture clusters. The specific process of calculating the three-dimensional vector of centroid migration of the water-conducting fracture clusters and the volumetric expansion rate of the rock mass within adjacent consecutive time windows is as follows: The trigger time series of microseismic waveforms recorded by the Internet of Things microseismic sensor network are obtained and divided into adjacent continuous time windows. The intrinsic permeability increment is converted into a spatial weighting factor within the adjacent continuous time windows. High-permeability core points are extracted by weighted spatial density clustering of three-dimensional spatial coordinates using spatial weighting factors, and temporally adjacent high-permeability core points are connected to generate water-conducting fracture clusters arranged in time sequence. Calculate the volume parameters of the three-dimensional envelope convex hull and the spatial geometric distribution center of the water-conducting fracture cluster within adjacent consecutive time windows; A three-dimensional centroid migration vector is constructed based on the spatial displacement offset of the spatial geometric distribution center within adjacent continuous time windows, and the rock mass volume expansion rate is inferred based on the temporal evolution differences of the three-dimensional envelope convex hull volume parameters.

9. The intelligent early warning method based on mine microseismic data analysis according to claim 1, characterized in that: The specific process of obtaining the vertical approximation rate by projecting the centroid migration three-dimensional vector onto the normal direction of the bottom aquifer, and outputting the corresponding three-dimensional spatial coordinates of the water hazard early warning command based on the vertical approximation rate and the rock mass volume expansion rate is as follows: Obtain the three-dimensional geological spatial distribution matrix of the hidden aquifer on the bottom plate of the mining face, and analyze the normal vector of the bottom aquifer facing the mining face in the three-dimensional geological spatial distribution matrix; The normal extension distance parameter is obtained by spatial orthogonally projecting the three-dimensional vector of the centroid migration onto the normal vector of the bottom aquifer. The vertical approximation rate is then deduced by combining the normal extension distance parameter with the time interval step of adjacent continuous time windows. The vertical approach rate and the volume expansion rate of the rock mass are mapped to a preset two-dimensional safety decision matrix for risk assessment, and the water hazard warning command corresponding to the three-dimensional spatial coordinates is output in combination with the risk trigger level.

10. An intelligent early warning system based on mine microseismic data analysis. An intelligent early warning method based on mine microseismic data analysis as described in any one of claims 1-9, characterized in that, include: The microseismic analysis module is used to receive the microseismic waveform sequence and the spatial trajectory of the coal mining machine collected by the Internet of Things microseismic sensor network, perform phase identification and travel time inversion on the microseismic waveform sequence, extract the three-dimensional spatial coordinates of the microseismic event, and analyze the frequency domain spectrum of the microseismic waveform sequence to extract the radiation energy and corner frequency. The mechanism inversion module is used to extract the initial amplitude ratio of P-waves and S-waves in the microseismic waveform sequence, reconstruct the source mechanism solution by combining it with three-dimensional spatial coordinates, screen the tensile rupture events of the base plate according to the source mechanism solution, and calculate the dynamic stress drop of the tensile rupture event by integrating radiation energy, corner frequency and three-dimensional spatial coordinates. The permeability calculation module is used to obtain the initial geostress tensor of the mining face, superimpose the dynamic stress drop onto the initial geostress tensor, reconstruct the mining-induced additional stress field in the space where the tensile rupture event is located, extract the local tensile principal stress direction vector of the mining-induced additional stress field, calculate the dot product projection of the rupture surface normal vector and the local tensile principal stress direction vector in the focal mechanism solution, and fuse the dot product projection with the dynamic stress drop to quantify the intrinsic permeability increment. The evolution early warning module is used to perform three-dimensional spatial density clustering of tensile fracturing events based on intrinsic permeability increments, generate water-conducting fracture clusters arranged in time sequence, calculate the three-dimensional vector of centroid migration and rock mass volume expansion rate of the water-conducting fracture clusters in adjacent continuous time windows, project the three-dimensional vector of centroid migration onto the normal direction of the bottom aquifer to obtain the vertical approximation rate, and output water hazard early warning commands with corresponding three-dimensional spatial coordinates based on the vertical approximation rate and rock mass volume expansion rate.

Images