Coal mine goaf crack measuring and monitoring system and method

Abstract

The invention discloses a coal mine goaf crack measuring and monitoring system and method, and the method comprises the following steps: collecting surrounding rock energy propagation behavior data, and carrying out the preprocessing; carrying out space grid modeling, constructing a goaf three-dimensional discrete volume element set, and generating an energy density field; calculating a space gradient for the energy density field, generating an energy gradient field, and calculating an energy flow direction vector field based on the energy gradient field; calculating an energy divergence anomaly score field, executing threshold segmentation and connected domain extraction, and generating an energy dissipation anomaly region set and an energy convergence anomaly region set; streamline tracking is executed based on the energy flow direction vector field, an energy flow deflection structure set is extracted, direction consistency screening is executed, and a crack trend set is obtained; and generating a coal mine goaf crack measurement result based on the crack trend set. The method realizes goaf crack measurement based on three-dimensional energy field analysis, and has the advantages of high continuity and high precision.

Description

Technical Field

[0001] This invention relates to the field of coal mine safety monitoring and rock mass structure detection, and in particular to a coal mine goaf crack measurement and monitoring system and method. Background Technology

[0002] Under the influence of mining disturbances and stress redistribution, coal mine goaf areas are prone to fracture propagation, shear slip, and structural instability within the surrounding rock. The spatial distribution and orientation of these fractures directly affect roof stability and the evolution of disasters. Existing technologies primarily rely on borehole inspection, geological logging, or single microseismic event location results for fracture identification. These technologies typically focus on inverting the location of the event source or analyzing wave velocity anomalies, lacking sufficient ability to model the spatial continuity of surrounding rock energy propagation behavior data and thus failing to represent the three-dimensional energy distribution structure within the goaf area.

[0003] Meanwhile, existing methods lack a joint analysis mechanism for energy density field, energy gradient field and energy flow direction vector field, fail to identify energy dissipation anomaly area and energy convergence anomaly area through energy divergence anomaly score field, and have not constructed a crack orientation inversion process based on streamline tracing and energy flow deflection structure extraction. This results in strong dispersion and insufficient directional stability of crack identification results, making it difficult to achieve continuous and structured measurement and monitoring of hidden cracks in goaf areas. Summary of the Invention

[0004] One objective of this invention is to propose a coal mine goaf crack measurement and monitoring system and method. This invention achieves goaf crack measurement based on three-dimensional energy field analysis, and has the advantages of strong continuity and high accuracy.

[0005] A method for measuring and monitoring cracks in a coal mine goaf according to an embodiment of the present invention includes the following steps: Within the goaf area of ​​a coal mine, data on the energy propagation behavior of the surrounding rock are collected and preprocessed to generate a standardized energy propagation dataset. Based on a standardized energy propagation dataset, spatial gridding modeling is performed to construct a three-dimensional discrete volume element set of the goaf, calculate the energy density of each volume element, and generate an energy density field. Calculate the spatial gradient of the energy density field to generate an energy gradient field, and calculate the energy flow direction vector field based on the energy gradient field; The energy divergence anomaly score field is calculated based on the energy density field, energy gradient field, and energy flow direction vector field. Threshold segmentation and connected component extraction are then performed to generate a set of energy dissipation anomaly regions and a set of energy convergence anomaly regions. Within the sets of energy dissipation anomalies and energy convergence anomalies, streamline tracing is performed based on the energy flow direction vector field to extract the set of energy flow deflection structures, generate a candidate set of crack orientations, and perform orientation consistency screening to obtain the set of crack orientations. The measurement results of cracks in the goaf of a coal mine are generated based on the set of crack orientations.

[0006] Optionally, the surrounding rock energy propagation behavior data includes event arrival time data, waveform amplitude data, frequency band energy data, and attenuation coefficient data.

[0007] Optionally, the preprocessing includes time synchronization, outlier removal, bandpass filtering, and amplitude normalization.

[0008] Optionally, the generation of the energy density field specifically includes: Obtain a standardized energy propagation data set and a spatial coordinate matrix of energy propagation monitoring points. Establish a three-dimensional spatial coordinate framework for the goaf based on the spatial boundary of the goaf. Then, perform mesh subdivision on the three-dimensional spatial coordinate framework of the goaf to obtain a three-dimensional discrete volume element set of the goaf. For each three-dimensional discrete element in the goaf set, extract the event arrival time data, waveform amplitude data, frequency band energy data and attenuation coefficient data in the corresponding spatial neighborhood. Combine the spatial coordinate matrix of the energy propagation monitoring point to calculate the spatial distance from each microseismic sensor to the center position of the current three-dimensional discrete element, and generate the local data mapping set of each three-dimensional discrete element. Based on the local data mapping set of the voxels, energy contribution allocation is performed on each three-dimensional discrete voxel. The waveform amplitude data and frequency band energy data are mapped to the voxel input energy value, the attenuation coefficient data and spatial distance are mapped to the voxel propagation attenuation value, and the voxel energy action time window is determined by combining the event arrival data, so as to obtain the voxel energy characterization data corresponding to each three-dimensional discrete voxel. Volume normalization is performed on the volume element energy characterization data of each three-dimensional discrete volume element in the three-dimensional discrete volume element set of the goaf to obtain the energy density value corresponding to each three-dimensional discrete volume element, forming a discrete energy density distribution set. The discrete energy density distribution set is used to reconstruct the spatial continuity of the three-dimensional discrete volume element set of the goaf, generating a continuous energy density distribution result; Boundary correction is performed on the continuous energy density distribution results to obtain the energy density field.

[0009] Optionally, the energy density field is a spatial distribution representation formed by assigning a corresponding energy density value to each three-dimensional discrete element in the three-dimensional space of the coal mine goaf, used to characterize the distribution state of the surrounding rock energy within the goaf.

[0010] Optionally, the generation of the energy flow direction vector field specifically includes: Read the energy density field and the three-dimensional spatial coordinate frame of the goaf, extract the energy density value, the center coordinates of the volume element and the adjacency relationship of each three-dimensional discrete volume element in the three-dimensional discrete volume element set of the goaf, and construct the volume element neighborhood data set. For each three-dimensional discrete voxel in the three-dimensional discrete voxel set of the goaf, the energy density values ​​of adjacent three-dimensional discrete voxels are extracted along the three coordinate directions of the three-dimensional spatial coordinate frame of the goaf. The energy density change of the current three-dimensional discrete voxel in the three coordinate directions is calculated, and the corresponding voxel directional gradient components are generated. The gradient components of the volume element direction corresponding to each three-dimensional discrete volume element are combined to obtain the energy gradient vector corresponding to each three-dimensional discrete volume element. The energy gradient vector is then written into the three-dimensional spatial coordinate frame of the goaf to generate the energy gradient field. Direction normalization is performed on each energy gradient vector in the energy gradient field to extract the dominant energy change direction of each three-dimensional discrete volume element and generate the corresponding set of volume element energy flow direction vectors. The set of energy flow direction vectors of the volume element is mapped to the three-dimensional spatial coordinate frame of the goaf according to the coordinates of the volume element center, forming the energy flow direction vectors corresponding to each three-dimensional discrete volume element in the three-dimensional space of the goaf, and thus obtaining the energy flow direction vector field. The consistency between the energy gradient field and the energy flow direction vector field is verified, and the three-dimensional discrete volume elements that meet the preset direction consistency conditions are retained in the effective volume element set.

[0011] Optionally, the generation of the set of energy dissipation anomaly regions and the set of energy accumulation anomaly regions specifically includes: Based on the energy density field, energy gradient field, and energy flow direction vector field, the energy density value, energy gradient vector, energy flow direction vector, center coordinates of the volume element and the adjacency relationship of the volume element are extracted from the three-dimensional discrete volume element set of the goaf, and an energy divergence anomaly calculation dataset is constructed. For each three-dimensional discrete volume element in the three-dimensional discrete volume element set of the goaf, the energy flow direction vector components of the adjacent three-dimensional discrete volume elements are extracted along the three coordinate directions of the three-dimensional spatial coordinate frame of the goaf. The vector change of the current three-dimensional discrete volume element in the three coordinate directions is calculated. Combined with the energy density value and energy gradient vector of the current three-dimensional discrete volume element, the corresponding volume element energy divergence value is generated. For each three-dimensional discrete volume element, the volume element energy divergence value is generated, and the energy divergence anomaly score is written into the three-dimensional spatial coordinate frame of the goaf to form an energy divergence anomaly score field. Threshold segmentation is performed on the energy divergence anomaly score field, and three-dimensional discrete volume elements with energy divergence anomaly scores greater than a preset anomaly threshold are extracted as candidate anomaly volume elements; For each candidate anomalous volume element, extract the energy flow direction vector corresponding to its adjacent three-dimensional discrete volume elements, calculate the directional convergence value of the energy flow direction vector of each adjacent three-dimensional discrete volume element pointing to the center position of the current candidate anomalous volume element, and accumulate all directional convergence values ​​to obtain the volume element direction convergence judgment value corresponding to the current candidate anomalous volume element. Candidate abnormal voxels whose voxel direction convergence determination value is greater than the preset convergence determination threshold are marked as energy convergence abnormal voxels, and candidate abnormal voxels whose voxel direction convergence determination value is less than the preset dissipation determination threshold are marked as energy dissipation abnormal voxels, thus generating a set of abnormal voxel labels. Based on the adjacency relationship of the volume elements in the three-dimensional discrete volume element set of the goaf, the connected component extraction process is performed on the volume element anomaly marker set to aggregate the three-dimensional discrete volume elements into multiple candidate anomaly connected components, thus obtaining the energy dissipation anomaly area set and the energy convergence anomaly area set.

[0012] Optionally, the generation of the crack orientation set specifically includes: Based on the set of energy dissipation anomalies, the set of energy convergence anomalies, and the energy flow direction vector field, the center coordinates of the three-dimensional discrete volume elements in each anomaly zone and the energy flow direction vector are extracted to generate a set of streamline tracing data for the anomaly zones. Using the coordinates of the center of the three-dimensional discrete volume element with the smallest spatial distance from the geometric center in each anomaly region as the starting point for streamline tracing, forward tracing is performed along the energy flow direction vector, and reverse tracing is performed along the opposite direction of the energy flow direction vector to obtain the set of forward streamlines and the set of reverse streamlines. For each streamline record in the forward and reverse streamline sets, a three-dimensional discrete element sequence is passed through. Local deflection segments are extracted based on the change in the directional angle between adjacent tracking steps to generate an energy flow deflection segment set. Based on the energy flow deflection fragment set, fragment aggregation is performed to merge local deflection fragments that are spatially continuous and have the same directional change trend into an energy flow deflection structure, thus obtaining an energy flow deflection structure set. Based on the dominant extension direction in the energy flow deflection structure set, candidate crack orientations corresponding to each energy flow deflection structure are extracted to generate a candidate crack orientation set. Perform direction consistency screening on the candidate set of crack directions to generate a set of crack directions; The set of crack orientations is mapped to the three-dimensional spatial coordinate frame of the goaf.

[0013] Optionally, the generation of the measurement results of cracks in the coal mine goaf specifically includes: The fracture orientations in the fracture orientation set are calibrated according to the three-dimensional spatial coordinate frame of the goaf area to generate a fracture orientation distribution data set. Extract the spatial location and orientation of each crack orientation from the crack orientation distribution dataset, determine the distribution location of each crack orientation in the three-dimensional space of the goaf, and generate a crack spatial location result set. Based on the set of crack spatial location results, the orientation of each crack in the three-dimensional space of the goaf is classified into regions, the distribution of crack orientation in different spatial regions is determined, and a set of crack regional distribution results is generated. The number of crack directions in the goaf is counted based on the crack direction set, and the crack directions are classified and organized according to their direction to generate a crack direction statistical result set. The results of crack spatial location, crack regional distribution, and crack direction statistics are summarized to generate the measurement results of cracks in the coal mine goaf.

[0014] A coal mine goaf crack measurement and monitoring system according to an embodiment of the present invention includes: The data acquisition module is used to collect data on the energy propagation behavior of surrounding rock within the coverage area of ​​the coal mine goaf, and to perform preprocessing to generate a standardized energy propagation data set. The energy density field construction module is used for spatial gridding modeling, constructing a three-dimensional discrete volume element set of the goaf, calculating the energy density of each volume element, and generating an energy density field. The gradient and flow direction calculation module is used to calculate the spatial gradient of the energy density field, generate the energy gradient field, and calculate the energy flow direction vector field based on the energy gradient field. The anomaly region extraction module is used to calculate the energy divergence anomaly score field based on the energy density field, energy gradient field and energy flow direction vector field, and to perform threshold segmentation and connected component extraction to generate a set of energy dissipation anomaly regions and a set of energy convergence anomaly regions. The crack orientation extraction module is used to extract the set of energy flow deflection structures by performing streamline tracing based on the energy flow direction vector field within the set of energy dissipation anomaly regions and the set of energy convergence anomaly regions, generate a candidate set of crack orientations, and perform directional consistency screening on the candidate set of crack orientations to obtain the crack orientation set. The measurement result generation module is used to generate measurement results of cracks in coal mine goaf based on the crack orientation set.

[0015] The beneficial effects of this invention are: This invention focuses on the energy propagation behavior data of surrounding rock. By constructing a three-dimensional discrete voxel set in the goaf and generating an energy density field, an energy gradient field, and an energy flow direction vector field, it achieves a continuous expression of the spatial distribution and propagation direction characteristics of energy within the surrounding rock. Compared with traditional techniques that rely solely on the location of a single microseismic event or the analysis of local anomaly parameters, this invention performs unified mapping and voxel energy characterization calculations on event arrival data, waveform amplitude data, frequency band energy data, and attenuation coefficient data. This enables the surrounding rock energy information to form a continuous energy density distribution result within a three-dimensional spatial coordinate framework. Furthermore, by calculating gradients and normalizing directions, a stable energy flow direction vector field is generated, thus providing a spatial field basis with physical constraints for subsequent crack identification and improving the spatial consistency and directional reliability of crack identification.

[0016] In terms of anomaly identification and crack orientation extraction, this invention constructs an energy divergence anomaly score field, combines threshold segmentation and connected domain extraction to distinguish between sets of energy dissipation anomaly zones and sets of energy convergence anomaly zones, giving the energy anomaly regions inside the surrounding rock clear spatial aggregation boundaries. Within the anomaly zones, streamline tracing is further performed based on the energy flow direction vector field to extract sets of energy flow deflection structures, and crack orientation sets are generated through direction consistency screening. This technical approach effectively avoids the problems of discrete crack identification results, large orientation fluctuations, and insufficient spatial continuity in traditional methods, enabling crack orientation results to have continuous extension characteristics and structured expression capabilities. It can achieve visual mapping and regional classification analysis in the three-dimensional space of the goaf, thereby improving the accuracy, stability, and engineering application value of hidden crack measurement in coal mine goafs. Attached Figure Description

[0017] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a flowchart of a method for measuring and monitoring cracks in coal mine goaf proposed in this invention; Figure 2 This is a schematic diagram illustrating the construction of an energy divergence anomaly score field for a method of measuring and monitoring cracks in coal mine goaf proposed in this invention. Detailed Implementation

[0018] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, illustrating only the basic structure of the invention, and therefore only show the components relevant to the invention.

[0019] refer to Figures 1-2 A method for measuring and monitoring cracks in coal mine goaf areas, comprising the following steps: Within the goaf area of ​​a coal mine, data on the energy propagation behavior of the surrounding rock are collected and preprocessed to generate a standardized energy propagation dataset. Based on a standardized energy propagation dataset, spatial gridding modeling is performed to construct a three-dimensional discrete volume element set of the goaf, calculate the energy density of each volume element, and generate an energy density field. Calculate the spatial gradient of the energy density field to generate an energy gradient field, and calculate the energy flow direction vector field based on the energy gradient field; The energy divergence anomaly score field is calculated based on the energy density field, energy gradient field, and energy flow direction vector field. Threshold segmentation and connected component extraction are then performed to generate a set of energy dissipation anomaly regions and a set of energy convergence anomaly regions. Within the sets of energy dissipation anomalies and energy convergence anomalies, streamline tracing is performed based on the energy flow direction vector field to extract the set of energy flow deflection structures, generate a candidate set of crack orientations, and perform orientation consistency screening to obtain the set of crack orientations. The measurement results of cracks in the goaf of a coal mine are generated based on the set of crack orientations.

[0020] In this embodiment, the surrounding rock energy propagation behavior data includes event arrival time data, waveform amplitude data, frequency band energy data, and attenuation coefficient data. Event arrival time data is the timestamp data recorded when the energy propagation wave first arrives at the corresponding energy propagation monitoring point when an energy release event occurs in the surrounding rock. Waveform amplitude data is the instantaneous amplitude sequence and peak amplitude data of the time-domain vibration wave signal formed by the energy propagation wave at each energy propagation monitoring point. Frequency band energy data is the energy integral value data calculated within a preset frequency range after performing frequency domain decomposition. Attenuation coefficient data is the energy propagation attenuation parameter data calculated based on the propagation distance between different energy propagation monitoring points and the corresponding waveform amplitude attenuation degree. The surrounding rock energy release event is the instantaneous rupture or slippage of the elastic strain energy accumulated in the local rock mass of the coal mine goaf under the action of mining disturbance or stress redistribution, which exceeds the strength limit, thereby releasing energy outward and generating vibration wave phenomena that can be recorded by microseismic sensors. The phenomenon is manifested as a transient energy release process formed during the expansion of internal cracks, shear slippage, or structural instability of the surrounding rock.

[0021] In this embodiment, preprocessing includes time synchronization, outlier removal, bandpass filtering, and amplitude normalization.

[0022] In this embodiment, the generation of the energy density field specifically includes: Obtain a standardized energy propagation data set and a spatial coordinate matrix of energy propagation monitoring points. Establish a three-dimensional spatial coordinate framework for the goaf based on the spatial boundary of the goaf. Then, perform mesh subdivision on the three-dimensional spatial coordinate framework of the goaf to obtain a three-dimensional discrete volume element set of the goaf. Energy propagation monitoring points are microseismic sensor installation locations within the coal mine goaf area, used to collect data on the energy propagation behavior of the surrounding rock and corresponding to specific spatial coordinate positions. The three-dimensional spatial coordinate frame of the goaf area is a three-dimensional coordinate positioning benchmark established with the spatial boundary of the coal mine goaf area as the range, used to uniformly represent the positions of each microseismic sensor, the positions of each three-dimensional discrete element, and the spatial distribution of energy density. For each three-dimensional discrete element in the goaf set, extract the event arrival time data, waveform amplitude data, frequency band energy data and attenuation coefficient data in the corresponding spatial neighborhood. Combine the spatial coordinate matrix of the energy propagation monitoring point to calculate the spatial distance from each microseismic sensor to the center position of the current three-dimensional discrete element, and generate the local data mapping set of each three-dimensional discrete element. A local data mapping set for a volume element refers to a set of data formed by associating event arrival data, waveform amplitude data, frequency band energy data, attenuation coefficient data, and the spatial distance between the corresponding microseismic sensor and the center of the three-dimensional discrete volume element within its spatial neighborhood for a certain three-dimensional discrete volume element, according to a unified volume element number. Based on the local data mapping set of the voxels, energy contribution allocation is performed on each three-dimensional discrete voxel. The waveform amplitude data and frequency band energy data are mapped to the voxel input energy value, the attenuation coefficient data and spatial distance are mapped to the voxel propagation attenuation value, and the voxel energy action time window is determined by combining the event arrival data, so as to obtain the voxel energy characterization data corresponding to each three-dimensional discrete voxel. For each 3D discrete voxel, read the waveform amplitude data, frequency band energy data, attenuation coefficient data, event arrival time data, and spatial distance corresponding to all mapping records in the voxel's local data mapping set; for each mapping record, perform a weighted sum of the waveform amplitude data and frequency band energy data to obtain the voxel input energy value corresponding to that mapping record; determine the corresponding distance attenuation factor value based on the spatial distance value, and then multiply the distance attenuation factor value by the attenuation coefficient data to obtain the voxel propagation attenuation value corresponding to that mapping record; divide the voxel input energy value by the voxel propagation attenuation value to obtain the corrected energy value corresponding to that mapping record; from the current 3D discrete voxel... The minimum arrival time value is extracted from all event arrival data corresponding to the voxel and determined as the start time value of the voxel energy action. The maximum arrival time value is then extracted and determined as the end time value of the voxel energy action. The voxel energy action time window value is obtained by subtracting the voxel energy action start time value from the end time value. Next, all corrected energy values ​​corresponding to the current 3D discrete voxel are accumulated to obtain the voxel cumulative corrected energy value. The voxel cumulative corrected energy value is divided by the voxel energy action time window value to obtain the voxel final energy value corresponding to the current 3D discrete voxel. The voxel final energy value is determined as the voxel energy representation data corresponding to the current 3D discrete voxel. Volume normalization is performed on the volume element energy characterization data of each three-dimensional discrete volume element in the three-dimensional discrete volume element set of the goaf to obtain the energy density value corresponding to each three-dimensional discrete volume element, forming a discrete energy density distribution set. The discrete energy density distribution set is used to reconstruct the spatial continuity of the three-dimensional discrete volume element set of the goaf, generating a continuous energy density distribution result; Boundary correction is performed on the continuous energy density distribution results to obtain the energy density field.

[0023] In this embodiment, the energy density field is a spatial distribution representation formed by assigning a corresponding energy density value to each three-dimensional discrete element in the three-dimensional space of the coal mine goaf, which is used to characterize the distribution state of the surrounding rock energy inside the goaf.

[0024] In this embodiment, the generation of the energy flow direction vector field specifically includes: Read the energy density field and the three-dimensional spatial coordinate frame of the goaf, extract the energy density value, the center coordinates of the volume element and the adjacency relationship of each three-dimensional discrete volume element in the three-dimensional discrete volume element set of the goaf, and construct the volume element neighborhood data set. The adjacency relationship of a volume element is the connection and correspondence between a certain three-dimensional discrete volume element and other three-dimensional discrete volume elements that are in contact or adjacent to it in the three-dimensional discrete volume element set in the goaf. The volume element neighborhood data set is a data set formed by collecting the energy density values, volume element center coordinates and volume element adjacency relationships of each three-dimensional discrete volume element adjacent to the current three-dimensional discrete volume element, with the current three-dimensional discrete volume element as the center. For each three-dimensional discrete voxel in the three-dimensional discrete voxel set of the goaf, the energy density values ​​of adjacent three-dimensional discrete voxels are extracted along the three coordinate directions of the three-dimensional spatial coordinate frame of the goaf. The energy density change of the current three-dimensional discrete voxel in the three coordinate directions is calculated, and the corresponding voxel directional gradient components are generated. For the current 3D discrete voxel in the 3D discrete voxel set of the goaf, read the voxel center coordinates and energy density value of the current 3D discrete voxel; based on the 3D spatial coordinate frame of the goaf and the voxel adjacency relationship, determine the positive and negative adjacent 3D discrete voxels of the current 3D discrete voxel in the first coordinate direction, read the energy density values ​​of the positive and negative adjacent 3D discrete voxels, and calculate the energy density change value in the first coordinate direction; using the same method, determine the positive and negative adjacent 3D discrete voxels of the current 3D discrete voxel in the second coordinate direction. For each discrete volume element, the corresponding energy density value is read, and the energy density change value in the second coordinate direction is calculated. Next, the positive and negative adjacent 3D discrete volume elements in the third coordinate direction are determined, and their corresponding energy density values ​​are read, and the energy density change value in the third coordinate direction is calculated. The energy density change values ​​in the first, second, and third coordinate directions are respectively determined as the first, second, and third direction gradient components corresponding to the current 3D discrete volume element, generating the volume element direction gradient components corresponding to the current 3D discrete volume element. The gradient components of the volume element direction corresponding to each three-dimensional discrete volume element are combined to obtain the energy gradient vector corresponding to each three-dimensional discrete volume element. The energy gradient vector is written into the three-dimensional spatial coordinate frame of the goaf according to the volume element number to generate the energy gradient field. The energy gradient field is used to characterize the changing trend of energy density at different locations. Direction normalization is performed on each energy gradient vector in the energy gradient field to extract the dominant energy change direction of each three-dimensional discrete volume element and generate the corresponding set of volume element energy flow direction vectors. For each three-dimensional discrete volume element in the energy gradient field, the corresponding energy gradient vector is read, and the first-direction gradient component, the second-direction gradient component, and the third-direction gradient component are extracted from the energy gradient vector. The gradient vector length value corresponding to the current three-dimensional discrete volume element is calculated. Based on the gradient vector length value, the first-direction gradient component, the second-direction gradient component, and the third-direction gradient component are normalized to obtain the first-direction normalized component value, the second-direction normalized component value, and the third-direction normalized component value corresponding to the current three-dimensional discrete volume element. The first-direction normalized component value, the second-direction normalized component value, and the third-direction normalized component value are combined to generate the direction normalized vector value corresponding to the current three-dimensional discrete volume element. The dominant energy change direction value of the current three-dimensional discrete volume element is determined based on the direction normalized vector value. The dominant energy change direction value is mapped to the initial vector value of the volume element energy flow direction corresponding to the current three-dimensional discrete volume element. The above process is repeated for all three-dimensional discrete volume elements in the goaf three-dimensional discrete volume element set to generate a set of volume element energy flow direction vectors. The set of energy flow direction vectors of the volume element is mapped to the three-dimensional spatial coordinate frame of the goaf according to the coordinates of the volume element center, forming the energy flow direction vectors corresponding to each three-dimensional discrete volume element in the three-dimensional space of the goaf, and thus obtaining the energy flow direction vector field. The energy flow direction vector field is the spatial distribution result of the energy flow direction vectors corresponding to each three-dimensional discrete element in the three-dimensional space of the goaf, and is used to characterize the propagation direction of energy at different locations. The consistency of the energy gradient field and the energy flow direction vector field is checked, and the three-dimensional discrete volume elements that meet the preset direction consistency conditions are retained in the effective volume element set. The preset direction consistency condition is a criterion used to determine whether the energy change direction corresponding to the same three-dimensional discrete volume element in the energy gradient field and the energy flow direction corresponding to the same energy flow direction vector field maintain a directional coordination relationship. Specifically, it is set as follows: if the directional deviation value between the energy gradient vector direction of the current three-dimensional discrete volume element and the energy flow direction correction vector direction of the volume element is less than the preset direction deviation threshold, it is determined that the current three-dimensional discrete volume element meets the preset direction consistency condition.

[0025] In this embodiment, the generation of the set of energy dissipation anomaly regions and the set of energy accumulation anomaly regions specifically includes: Based on the energy density field, energy gradient field, and energy flow direction vector field, the energy density value, energy gradient vector, energy flow direction vector, center coordinates of the volume element and the adjacency relationship of the volume element are extracted from the three-dimensional discrete volume element set of the goaf, and an energy divergence anomaly calculation dataset is constructed. For each three-dimensional discrete volume element in the three-dimensional discrete volume element set of the goaf, the energy flow direction vector components of the adjacent three-dimensional discrete volume elements are extracted along the three coordinate directions of the three-dimensional spatial coordinate frame of the goaf. The vector change of the current three-dimensional discrete volume element in the three coordinate directions is calculated. Combined with the energy density value and energy gradient vector of the current three-dimensional discrete volume element, the corresponding volume element energy divergence value is generated. For each three-dimensional discrete element in the three-dimensional discrete element set of the goaf, first read the center coordinates of the current three-dimensional discrete element to obtain the current element coordinate value; then, based on the adjacency relationship of the elements, locate the forward adjacent three-dimensional discrete elements and the backward adjacent three-dimensional discrete elements of the current three-dimensional discrete element in the first coordinate direction, read the directional component value of the energy flow direction vector of the forward adjacent three-dimensional discrete element in the first coordinate direction and the directional component value of the energy flow direction vector of the backward adjacent three-dimensional discrete element in the first coordinate direction, and perform difference processing on the two to obtain the first directional vector change value; locate the forward adjacent three-dimensional discrete elements and the backward adjacent three-dimensional discrete elements of the current three-dimensional discrete element in the second coordinate direction in the same way, read the corresponding directional component values ​​and perform difference processing to obtain the second directional vector change value; Next, locate the forward and backward adjacent 3D discrete voxels of the current 3D discrete voxel in the third coordinate direction, read the corresponding direction component values ​​and perform interpolation to obtain the change value of the third direction vector; take the absolute values ​​of the change values ​​of the first direction vector, the second direction vector, and the third direction vector and perform accumulation processing to obtain the cumulative value of the voxel direction change; read the energy density value corresponding to the current 3D discrete voxel to obtain the current voxel energy density value; then read the energy gradient vector corresponding to the current 3D discrete voxel and perform magnitude synthesis processing on each direction component of the energy gradient vector to obtain the current voxel gradient intensity value; normalize the cumulative value of the voxel direction change, the current voxel energy density value, and the current voxel gradient intensity value and perform weighted summation to generate the voxel energy divergence value corresponding to the current 3D discrete voxel; For each three-dimensional discrete volume element, the volume element energy divergence value is generated, and the energy divergence anomaly score is written into the three-dimensional spatial coordinate frame of the goaf to form an energy divergence anomaly score field. For each three-dimensional discrete voxel in the three-dimensional discrete voxel set of the goaf, firstly, the voxel energy divergence value corresponding to the current three-dimensional discrete voxel is read to obtain the current divergence value; then, based on the adjacency relationship of the voxels, adjacent three-dimensional discrete voxels within the neighborhood range centered on the current three-dimensional discrete voxel are extracted to form the local neighborhood set of the current voxel; the voxel energy divergence values ​​corresponding to all three-dimensional discrete voxels in the local neighborhood set of the current voxel are read, and the mean value of all the read voxel energy divergence values ​​is calculated to obtain the local neighborhood divergence mean; then, discrete averaging is performed on the voxel energy divergence values ​​corresponding to all three-dimensional discrete voxels in the local neighborhood set of the current voxel. The calculation process yields the local neighborhood divergence standard deviation; the difference between the current divergence value and the local neighborhood divergence mean is processed to obtain the divergence deviation value; the absolute value of the divergence deviation value is then taken to obtain the degree of deviation value; the ratio between the degree of deviation value and the local neighborhood divergence standard deviation is processed to obtain the degree of dispersion value; the degree of dispersion value is normalized to obtain the standardized outlier value corresponding to the current three-dimensional discrete volume element; this serves as the energy divergence anomaly score corresponding to the current three-dimensional discrete volume element; the energy divergence anomaly scores corresponding to each three-dimensional discrete volume element are written into the three-dimensional spatial coordinate frame of the goaf according to the volume element number and the volume element center coordinates to form an energy divergence anomaly score field; Threshold segmentation is performed on the energy divergence anomaly score field, and three-dimensional discrete volume elements with energy divergence anomaly scores greater than a preset anomaly threshold are extracted as candidate anomaly volume elements; The preset anomaly threshold is set based on the statistical distribution results of all energy divergence anomaly scores in the three-dimensional discrete element set of the goaf. For each candidate anomalous volume element, extract the energy flow direction vector corresponding to its adjacent three-dimensional discrete volume elements, calculate the directional convergence value of the energy flow direction vector of each adjacent three-dimensional discrete volume element pointing to the center position of the current candidate anomalous volume element, and accumulate all directional convergence values ​​to obtain the volume element direction convergence judgment value corresponding to the current candidate anomalous volume element. For each candidate anomalous voxel, the voxel center coordinates of the current candidate anomalous voxel are read, and the voxel center coordinates of each adjacent 3D discrete voxel are also read. Based on the voxel center coordinates of each adjacent 3D discrete voxel and the voxel center coordinates of the current candidate anomalous voxel, a neighborhood pointing vector is determined. The energy flow direction vectors corresponding to each adjacent 3D discrete voxel are read, and the angle between the energy flow direction vector and the corresponding neighborhood pointing vector is calculated to obtain the direction angle value. The direction angle value is mapped to generate the direction approximation value corresponding to each adjacent 3D discrete voxel, where the smaller the direction angle value, the larger the corresponding direction approximation value. Candidate abnormal voxels whose voxel direction convergence judgment value is greater than the preset convergence judgment threshold are marked as energy convergence abnormal voxels, and candidate abnormal voxels whose voxel direction convergence judgment value is less than the preset dissipation judgment threshold are marked as energy dissipation abnormal voxels, generating a set of abnormal voxel labels. The remaining candidate abnormal voxels do not participate in the subsequent abnormal connected component extraction. The preset convergence judgment threshold and the preset dissipation judgment threshold are set according to the statistical distribution results of the convergence judgment values ​​of the volume elements corresponding to the candidate abnormal volume elements, and the preset convergence judgment threshold is greater than the preset dissipation judgment threshold. Based on the adjacency relationship of the volume elements in the three-dimensional discrete volume element set of the goaf, the connected component extraction process is performed on the volume element anomaly marker set to aggregate the three-dimensional discrete volume elements into multiple candidate anomaly connected components, thus obtaining the energy dissipation anomaly area set and the energy convergence anomaly area set. When performing connected component extraction on the set of anomaly markers, the process first reads the voxel numbers, center coordinates, and anomaly marker values ​​of all marked 3D discrete voxels in the set to generate an anomaly voxel index set. Then, an unaggregated marked 3D discrete voxel is selected from the index set as the current seed voxel, and its anomaly marker value is read to obtain the current anomaly type value. Based on the voxel adjacency relationship, spatially adjacent 3D discrete voxels are extracted, and adjacent 3D discrete voxels with the same anomaly marker value and current anomaly type value are selected to generate the first-level connected component set. Subsequently, each 3D discrete voxel in the first-level connected component set is used as an extended voxel, and adjacent 3D discrete voxels corresponding to each extended voxel are extracted. Three-dimensional discrete voxels with the same anomaly marker value and current anomaly type value that have not been aggregated are selected to obtain a new connected component set. The new connected component set is then processed... Merge the current connected component set and repeat the adjacent extraction, similar filtering, and set merging processes until the number of voxels in the newly added connected component set is zero, generating the current candidate abnormal connected domain. Perform voxel count and spatial range extraction on all 3D discrete voxels in the current candidate abnormal connected domain to obtain the connected domain voxel count value and connected domain spatial boundary value. Mark all 3D discrete voxels in the current candidate abnormal connected domain as aggregated voxels, and continue to select the next unaggregated marked 3D discrete voxel from the abnormal voxel index set and repeat the above process until all marked 3D discrete voxels in the abnormal voxel index set have completed connected domain aggregation. All candidate abnormal connected domains with anomaly type values ​​corresponding to energy dissipation abnormal voxels are assigned to the energy dissipation abnormal region set, and all candidate abnormal connected domains with anomaly type values ​​corresponding to energy convergence abnormal voxels are assigned to the energy convergence abnormal region set.

[0026] In this embodiment, the generation of the crack orientation set specifically includes: Based on the set of energy dissipation anomalies, the set of energy convergence anomalies, and the energy flow direction vector field, the center coordinates of the three-dimensional discrete volume elements in each anomaly zone and the energy flow direction vector are extracted to generate a set of streamline tracing data for the anomaly zones. Using the coordinates of the center of the three-dimensional discrete volume element with the smallest spatial distance from the geometric center in each anomaly region as the starting point for streamline tracing, forward tracing is performed along the energy flow direction vector, and reverse tracing is performed along the opposite direction of the energy flow direction vector to obtain the set of forward streamlines and the set of reverse streamlines. For each anomalous region in the set of energy dissipation anomalies and the set of energy convergence anomalies, calculate the average coordinates of the centers of all three-dimensional discrete elements within the current anomalous region to generate a geometric center position value. Calculate the spatial distance between the center coordinates of each three-dimensional discrete element and the geometric center position value within the current anomalous region, and select the three-dimensional discrete element with the smallest spatial distance to generate an initial element number value. Determine the center coordinates of the element corresponding to the initial element number value as the streamline tracing starting point, and read the energy flow direction vector of the three-dimensional discrete element corresponding to the initial element number value to generate an initial direction vector value. Using the streamline tracing starting point as the current tracing position value and the initial direction vector value as the current tracing direction value, calculate the direction angle between the energy flow direction vector of each adjacent three-dimensional discrete element and the current tracing direction value, and select the adjacent three-dimensional discrete element with the smallest direction angle to generate the next tracing element number value. The center coordinates of the voxel corresponding to the voxel number value are determined as the next tracking position value. The energy flow direction vector of the three-dimensional discrete voxel corresponding to the next tracking voxel number value is read to generate the next tracking direction value. The current tracking position value and the next tracking position value are sequentially written into the forward tracking coordinate sequence to generate the result value of one forward tracking step. The next tracking voxel number value generation, next tracking position value update and forward tracking coordinate sequence writing processes are repeated until the next tracking position value exceeds the current anomaly zone boundary or there are no adjacent three-dimensional discrete voxels that meet the tracking conditions, generating the forward streamline value. The starting direction vector value is inverted to generate the reverse starting direction value. The streamline tracking starting point is used as the reverse current tracking position value, and the reverse starting direction value is used as the reverse current tracking direction value. Adjacent three-dimensional discrete voxels are selected according to the principle of minimum direction angle, and the reverse tracking position update process is repeated to generate the reverse streamline value. The above processes are performed on all anomaly zones to generate the forward streamline set and the reverse streamline set. For each streamline record in the forward and reverse streamline sets, a three-dimensional discrete element sequence is passed through. Local deflection segments are extracted based on the change in the directional angle between adjacent tracking steps to generate an energy flow deflection segment set. For each streamline in the forward and reverse streamline sets, the corresponding 3D discrete volume element number is read according to the tracking order, generating a 3D discrete volume element sequence value; based on the volume element center coordinates of two adjacent 3D discrete volume elements in the 3D discrete volume element sequence value, the displacement direction value corresponding to each tracking step is calculated, generating a displacement direction sequence value; the directional angle value between two adjacent displacement direction values ​​is calculated according to the tracking order, generating a directional angle sequence value; each directional angle value is progressively determined to be greater than a preset deflection angle threshold. If it is greater than the preset deflection angle threshold, the corresponding tracking step is marked as a deflection step, generating a deflection step mark sequence value; the deflection step mark sequence is then processed. Tracking steps consecutively marked as deflection steps in the column values ​​are sequentially aggregated, and the corresponding three-dimensional discrete element number intervals are extracted to generate local deflection interval values. Based on each local deflection interval value, the corresponding three-dimensional discrete element subsequences are extracted, and the center coordinates of the first and last elements and the overall extension direction value of each three-dimensional discrete element subsequence are extracted to generate energy flow deflection segment values. The above processing is repeated for all streamlines to obtain a set of energy flow deflection segments. The dominant extension direction value is determined by the center coordinates of the first and last elements in the local deflection segment, which represents the direction of the overall spatial extension of the local deflection segment. Based on the energy flow deflection fragment set, fragment aggregation is performed to merge local deflection fragments that are spatially continuous and have the same directional change trend into an energy flow deflection structure, thus obtaining an energy flow deflection structure set. For each energy flow deflection segment in the energy flow deflection segment set, extract the coordinates of the center of the first voxel, the coordinates of the center of the last voxel, and the dominant extension direction value to generate segment feature values; calculate the difference between the endpoint distance and the dominant extension direction between any two energy flow deflection segments to generate segment association values; identify energy flow deflection segments whose endpoint distance is lower than a preset distance threshold and whose dominant extension direction difference is lower than a preset direction threshold as segments in the same group to generate segment grouping result values; merge each group of segments according to the spatial connection order, extract the merged three-dimensional discrete voxel subsequence, the coordinates of the center of the first voxel, the coordinates of the center of the last voxel, and the overall dominant extension direction value to generate energy flow deflection structure values; summarize all energy flow deflection structure values ​​to obtain the energy flow deflection structure set; Based on the dominant extension direction in the energy flow deflection structure set, candidate crack orientations corresponding to each energy flow deflection structure are extracted to generate a candidate crack orientation set. For each energy flow deflection structure in the energy flow deflection structure set, the coordinates of the center of the first and last volume elements and the overall dominant extension direction are read to generate structural direction feature values. Using the overall dominant extension direction value as the direction reference value for the current energy flow deflection structure, the coordinates of the center of the first volume element are projected onto the three-dimensional spatial coordinate frame of the goaf along the direction reference value to generate the initial projection position value. Similarly, the coordinates of the center of the last volume element are projected onto the three-dimensional spatial coordinate frame of the goaf along the direction reference value to generate the final projection position value. Based on the initial and final projection position values, the spatial extension axis of the current energy flow deflection structure is determined, generating candidate orientation axis values. The direction value corresponding to the candidate orientation axis value is determined as the fracture orientation candidate value, and the fracture orientation candidate value is associated with the coordinates of the center of the first and last volume elements of the current energy flow deflection structure to generate a single fracture orientation candidate record value. The above process is repeated for all energy flow deflection structures to obtain a fracture orientation candidate set. Perform direction consistency screening on the crack orientation candidate set, calculate the direction difference and overlap ratio between adjacent crack orientation candidates, retain crack orientation candidates whose direction difference is lower than the preset direction difference threshold and whose overlap ratio is higher than the preset overlap threshold, and generate a crack orientation set. The crack orientation set is a data set composed of multiple identified crack orientation results, and each crack orientation result includes the spatial location and orientation information of the corresponding crack. The set of crack orientations is mapped to the three-dimensional spatial coordinate frame of the goaf.

[0027] In this embodiment, the generation of coal mine goaf crack measurement results specifically includes: The fracture orientations in the fracture orientation set are calibrated according to the three-dimensional spatial coordinate frame of the goaf area to generate a fracture orientation distribution data set. Extract the spatial location and orientation of each crack orientation from the crack orientation distribution dataset, determine the distribution location of each crack orientation in the three-dimensional space of the goaf, and generate a crack spatial location result set. Based on the set of crack spatial location results, the orientation of each crack in the three-dimensional space of the goaf is classified into regions, the distribution of crack orientation in different spatial regions is determined, and a set of crack regional distribution results is generated. The number of crack directions in the goaf is counted based on the crack direction set, and the crack directions are classified and organized according to their direction to generate a crack direction statistical result set. The results of crack spatial location, crack regional distribution, and crack direction statistics are summarized to generate the measurement results of cracks in the coal mine goaf.

[0028] A coal mine goaf crack measurement and monitoring system, comprising: The data acquisition module is used to collect data on the energy propagation behavior of surrounding rock within the coverage area of ​​the coal mine goaf, and to perform preprocessing to generate a standardized energy propagation data set. The energy density field construction module is used for spatial gridding modeling, constructing a three-dimensional discrete volume element set of the goaf, calculating the energy density of each volume element, and generating an energy density field. The gradient and flow direction calculation module is used to calculate the spatial gradient of the energy density field, generate the energy gradient field, and calculate the energy flow direction vector field based on the energy gradient field. The anomaly region extraction module is used to calculate the energy divergence anomaly score field based on the energy density field, energy gradient field and energy flow direction vector field, and to perform threshold segmentation and connected component extraction to generate a set of energy dissipation anomaly regions and a set of energy convergence anomaly regions. The crack orientation extraction module is used to extract the set of energy flow deflection structures by performing streamline tracing based on the energy flow direction vector field within the set of energy dissipation anomaly regions and the set of energy convergence anomaly regions, generate a candidate set of crack orientations, and perform directional consistency screening on the candidate set of crack orientations to obtain the crack orientation set. The measurement result generation module is used to generate measurement results of cracks in coal mine goaf based on the crack orientation set.

[0029] Example 1: To verify the feasibility of the present invention in practice, it was applied to the scenario of monitoring cracks in the goaf of a certain underground coal mine. Under the long-term mining disturbance and stress redistribution, there is a risk of hidden crack propagation in the roof and surrounding rock of the goaf. Traditional methods relying on borehole inspection and manual geological logging are difficult to grasp the spatial direction and extension trend of cracks in a timely manner. Although the microseismic monitoring system can record energy release events in the surrounding rock, it is difficult to reflect the continuous spatial morphology of the crack structure based solely on the source location results. The crack direction judgment has discretization and directional fluctuation problems, which cannot provide a stable basis for subsequent support optimization and safety early warning.

[0030] A microseismic sensor array was deployed within the goaf area of ​​the mine to continuously collect data on the energy propagation behavior of the surrounding rock, including event arrival time data, waveform amplitude data, frequency band energy data, and attenuation coefficient data. Time synchronization, anomaly segment removal, bandpass filtering, and amplitude normalization were performed at the ground monitoring center to form a standardized energy propagation data set. Subsequently, a three-dimensional discrete voxel set was established within the three-dimensional spatial coordinate framework of the goaf area. Energy contribution allocation was calculated for each voxel to generate a continuous energy density field. Based on the energy density field, the energy gradient field and energy flow direction vector field were further calculated. The characteristics of energy propagation direction change in space were obtained through voxel adjacency analysis. Combining the energy density value, gradient intensity, and cumulative value of direction change, an energy divergence anomaly score field was constructed. Through threshold segmentation and connected component extraction, sets of energy dissipation anomaly areas and sets of energy convergence anomaly areas were formed. Streamline tracing was performed along the energy flow direction vector within the anomaly areas to extract the energy flow deflection structure set, thereby generating a set of crack orientations. This set was then mapped onto the three-dimensional spatial model of the goaf area to achieve visualization of the spatial location and orientation of the cracks.

[0031] During the continuous monitoring period, the system operates stably in the eastern wing working face and return airway of the goaf. By comparing and analyzing the energy divergence anomaly subfields over multiple time periods, the evolution trend of the crack extension direction can be clearly reflected. Compared with the original method based on the distribution statistics of seismic source points, the crack orientation set formed by this invention is more stable in terms of spatial continuity and directional consistency, and the crack area distribution boundary is clearer. It can form a coherent crack extension axis structure in the three-dimensional space of the goaf.

[0032] To verify the performance of the present invention, it was compared with the traditional method. The comparison results are shown in Table 1.

[0033] Table 1. Comparison of Overall Performance of Crack Identification Methods in Goaf Areas As shown in Table 1, the traditional microseismic source clustering method achieves an accuracy of 78.6% in crack orientation identification, while the method of this invention reaches 92.4%, an improvement of 13.8 percentage points. Traditional methods rely solely on statistical clustering based on the spatial distribution of seismic sources, lacking constraints related to the continuity of the energy field. Therefore, directional overlap easily occurs in areas where multiple cracks intersect. This invention constructs an energy density field, an energy gradient field, and an energy flow direction vector field, allowing the crack orientation to be derived from the energy flow deflection structure. This provides a physical continuity basis for orientation determination, thus improving identification accuracy.

[0034] In terms of spatial continuity scoring, the traditional method scores 0.62, while the present invention achieves 0.88. This score is based on the calculation of the spatial continuity of the crack axis. The crack results generated by the traditional method are mostly discrete point groups, which are difficult to form a coherent extended structure. However, the present invention performs streamline tracing in the abnormal zone and performs structural aggregation on the deflection segments, so that the crack direction result is a continuous axis, thus significantly enhancing the spatial coherence.

[0035] The stability fluctuation value of the crack extension direction decreased from 14.3° to 6.1°, and the directional fluctuation was reduced by more than half. This shows that the present invention effectively filters out the directional jump problem caused by local noise interference under the action of the directional consistency screening mechanism. The combined use of the energy divergence anomaly score field and the directional convergence judgment value makes the internal structure of the anomaly area have a consistent directional trend, thereby reducing the degree of directional dispersion.

[0036] The data processing cycle was shortened from 4.5 hours to 2.8 hours, mainly because this invention adopts a structured calculation method based on volume element adjacency relationships. After the energy density field and gradient field are generated, each step can be executed in parallel, reducing the computational burden of multiple iterative localizations in traditional source inversion. Structured spatial calculation improves overall efficiency.

[0037] In terms of safety management, the early warning response time of this invention reaches 14 hours, which is 8 hours earlier than the traditional method. Since this invention can identify the potential crack evolution trend based on the formation stage of the energy divergence anomaly zone, while the traditional method often needs to wait for the source to concentrate and enhance before forming obvious clustering characteristics, it has a higher identification capability in the time dimension.

[0038] After the support adjustment, the energy anomaly convergence rate increased from 41.5% to 68.9%, indicating that after targeted support optimization based on the set of fracture spatial location results generated by this invention, the energy release behavior of the surrounding rock tends to be stable and the area of ​​the abnormal region is significantly reduced. The fundamental reason is that the spatial positioning of the fracture is more accurate, and the support measures can be more concentrated on the actual fracture extension zone.

[0039] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for measuring and monitoring cracks in coal mine goaf areas, characterized in that, Includes the following steps: Within the goaf area of ​​a coal mine, data on the energy propagation behavior of the surrounding rock are collected and preprocessed to generate a standardized energy propagation dataset. Based on a standardized energy propagation dataset, spatial gridding modeling is performed to construct a three-dimensional discrete volume element set of the goaf, calculate the energy density of each volume element, and generate an energy density field. Calculate the spatial gradient of the energy density field to generate an energy gradient field, and calculate the energy flow direction vector field based on the energy gradient field; The energy divergence anomaly score field is calculated based on the energy density field, energy gradient field, and energy flow direction vector field. Threshold segmentation and connected component extraction are then performed to generate a set of energy dissipation anomaly regions and a set of energy convergence anomaly regions. Within the sets of energy dissipation anomalies and energy convergence anomalies, streamline tracing is performed based on the energy flow direction vector field to extract the set of energy flow deflection structures, generate a candidate set of crack orientations, and perform orientation consistency screening to obtain the set of crack orientations. The measurement results of cracks in the goaf of a coal mine are generated based on the set of crack orientations.

2. The method for measuring and monitoring cracks in coal mine goafs according to claim 1, characterized in that, The surrounding rock energy propagation behavior data includes event arrival time data, waveform amplitude data, frequency band energy data, and attenuation coefficient data.

3. The method for measuring and monitoring cracks in coal mine goafs according to claim 1, characterized in that, The preprocessing includes time synchronization, outlier removal, bandpass filtering, and amplitude normalization.

4. The method for measuring and monitoring cracks in coal mine goafs according to claim 1, characterized in that, The generation of the energy density field specifically includes: Obtain a standardized energy propagation data set and a spatial coordinate matrix of energy propagation monitoring points. Establish a three-dimensional spatial coordinate framework for the goaf based on the spatial boundary of the goaf. Then, perform mesh subdivision on the three-dimensional spatial coordinate framework of the goaf to obtain a three-dimensional discrete volume element set of the goaf. For each three-dimensional discrete element in the goaf set, extract the event arrival time data, waveform amplitude data, frequency band energy data and attenuation coefficient data in the corresponding spatial neighborhood. Combine the spatial coordinate matrix of the energy propagation monitoring point to calculate the spatial distance from each microseismic sensor to the center position of the current three-dimensional discrete element, and generate the local data mapping set of each three-dimensional discrete element. Based on the local data mapping set of the voxels, energy contribution allocation is performed on each three-dimensional discrete voxel. The waveform amplitude data and frequency band energy data are mapped to the voxel input energy value, the attenuation coefficient data and spatial distance are mapped to the voxel propagation attenuation value, and the voxel energy action time window is determined by combining the event arrival data, so as to obtain the voxel energy characterization data corresponding to each three-dimensional discrete voxel. Volume normalization is performed on the volume element energy characterization data of each three-dimensional discrete volume element in the three-dimensional discrete volume element set of the goaf to obtain the energy density value corresponding to each three-dimensional discrete volume element, forming a discrete energy density distribution set. The discrete energy density distribution set is used to reconstruct the spatial continuity of the three-dimensional discrete volume element set of the goaf, generating a continuous energy density distribution result; Boundary correction is performed on the continuous energy density distribution results to obtain the energy density field.

5. The method for measuring and monitoring cracks in coal mine goafs according to claim 4, characterized in that, The energy density field is a spatial distribution representation formed by assigning a corresponding energy density value to each three-dimensional discrete element in the three-dimensional space of the coal mine goaf, and is used to characterize the distribution state of the surrounding rock energy within the goaf.

6. The method for measuring and monitoring cracks in coal mine goafs according to claim 1, characterized in that, The generation of the energy flow direction vector field specifically includes: Read the energy density field and the three-dimensional spatial coordinate frame of the goaf, extract the energy density value, the center coordinates of the volume element and the adjacency relationship of each three-dimensional discrete volume element in the three-dimensional discrete volume element set of the goaf, and construct the volume element neighborhood data set. For each three-dimensional discrete voxel in the three-dimensional discrete voxel set of the goaf, the energy density values ​​of adjacent three-dimensional discrete voxels are extracted along the three coordinate directions of the three-dimensional spatial coordinate frame of the goaf. The energy density change of the current three-dimensional discrete voxel in the three coordinate directions is calculated, and the corresponding voxel directional gradient components are generated. The gradient components of the volume element direction corresponding to each three-dimensional discrete volume element are combined to obtain the energy gradient vector corresponding to each three-dimensional discrete volume element. The energy gradient vector is then written into the three-dimensional spatial coordinate frame of the goaf to generate the energy gradient field. Direction normalization is performed on each energy gradient vector in the energy gradient field to extract the dominant energy change direction of each three-dimensional discrete volume element and generate the corresponding set of volume element energy flow direction vectors. The set of energy flow direction vectors of the volume element is mapped to the three-dimensional spatial coordinate frame of the goaf according to the coordinates of the volume element center, forming the energy flow direction vectors corresponding to each three-dimensional discrete volume element in the three-dimensional space of the goaf, and thus obtaining the energy flow direction vector field. The consistency between the energy gradient field and the energy flow direction vector field is verified, and the three-dimensional discrete volume elements that meet the preset direction consistency conditions are retained in the effective volume element set.

7. The method for measuring and monitoring cracks in coal mine goafs according to claim 1, characterized in that, The generation of the set of energy dissipation anomaly regions and the set of energy accumulation anomaly regions specifically includes: Based on the energy density field, energy gradient field, and energy flow direction vector field, the energy density value, energy gradient vector, energy flow direction vector, center coordinates of the volume element and the adjacency relationship of the volume element are extracted from the three-dimensional discrete volume element set of the goaf, and an energy divergence anomaly calculation dataset is constructed. For each three-dimensional discrete volume element in the three-dimensional discrete volume element set of the goaf, the energy flow direction vector components of the adjacent three-dimensional discrete volume elements are extracted along the three coordinate directions of the three-dimensional spatial coordinate frame of the goaf. The vector change of the current three-dimensional discrete volume element in the three coordinate directions is calculated. Combined with the energy density value and energy gradient vector of the current three-dimensional discrete volume element, the corresponding volume element energy divergence value is generated. For each three-dimensional discrete volume element, the volume element energy divergence value is generated, and the energy divergence anomaly score is written into the three-dimensional spatial coordinate frame of the goaf to form an energy divergence anomaly score field. Threshold segmentation is performed on the energy divergence anomaly score field, and three-dimensional discrete volume elements with energy divergence anomaly scores greater than a preset anomaly threshold are extracted as candidate anomaly volume elements; For each candidate anomalous volume element, extract the energy flow direction vector corresponding to its adjacent three-dimensional discrete volume elements, calculate the directional convergence value of the energy flow direction vector of each adjacent three-dimensional discrete volume element pointing to the center position of the current candidate anomalous volume element, and accumulate all directional convergence values ​​to obtain the volume element direction convergence judgment value corresponding to the current candidate anomalous volume element. Candidate abnormal voxels whose voxel direction convergence determination value is greater than the preset convergence determination threshold are marked as energy convergence abnormal voxels, and candidate abnormal voxels whose voxel direction convergence determination value is less than the preset dissipation determination threshold are marked as energy dissipation abnormal voxels, thus generating a set of abnormal voxel labels. Based on the adjacency relationship of the volume elements in the three-dimensional discrete volume element set of the goaf, the connected component extraction process is performed on the volume element anomaly marker set to aggregate the three-dimensional discrete volume elements into multiple candidate anomaly connected components, thus obtaining the energy dissipation anomaly area set and the energy convergence anomaly area set.

8. The method for measuring and monitoring cracks in coal mine goafs according to claim 1, characterized in that, The generation of the crack orientation set specifically includes: Based on the set of energy dissipation anomalies, the set of energy convergence anomalies, and the energy flow direction vector field, the center coordinates of the three-dimensional discrete volume elements in each anomaly zone and the energy flow direction vector are extracted to generate a set of streamline tracing data for the anomaly zones. Using the coordinates of the center of the three-dimensional discrete volume element with the smallest spatial distance from the geometric center in each anomaly region as the starting point for streamline tracing, forward tracing is performed along the energy flow direction vector, and reverse tracing is performed along the opposite direction of the energy flow direction vector to obtain the set of forward streamlines and the set of reverse streamlines. For each streamline record in the forward and reverse streamline sets, a three-dimensional discrete element sequence is passed through. Local deflection segments are extracted based on the change in the directional angle between adjacent tracking steps to generate an energy flow deflection segment set. Based on the energy flow deflection fragment set, fragment aggregation is performed to merge local deflection fragments that are spatially continuous and have the same directional change trend into an energy flow deflection structure, thus obtaining an energy flow deflection structure set. Based on the dominant extension direction in the energy flow deflection structure set, candidate crack orientations corresponding to each energy flow deflection structure are extracted to generate a candidate crack orientation set. Perform direction consistency screening on the candidate set of crack directions to generate a set of crack directions; The set of crack orientations is mapped to the three-dimensional spatial coordinate frame of the goaf.

9. The method for measuring and monitoring cracks in a coal mine goaf according to claim 1, characterized in that, The generation of the measurement results of cracks in the coal mine goaf specifically includes: The fracture orientations in the fracture orientation set are calibrated according to the three-dimensional spatial coordinate frame of the goaf area to generate a fracture orientation distribution data set. Extract the spatial location and orientation of each crack orientation from the crack orientation distribution dataset, determine the distribution location of each crack orientation in the three-dimensional space of the goaf, and generate a crack spatial location result set. Based on the set of crack spatial location results, the orientation of each crack in the three-dimensional space of the goaf is classified into regions, the distribution of crack orientation in different spatial regions is determined, and a set of crack regional distribution results is generated. The number of crack directions in the goaf is counted based on the crack direction set, and the crack directions are classified and organized according to their direction to generate a crack direction statistical result set. The results of crack spatial location, crack regional distribution, and crack direction statistics are summarized to generate the measurement results of cracks in the coal mine goaf.

10. A coal mine goaf crack measurement and monitoring system, comprising the coal mine goaf crack measurement and monitoring method according to any one of claims 1 to 9, characterized in that, include: The data acquisition module is used to collect data on the energy propagation behavior of surrounding rock within the coverage area of ​​the coal mine goaf, and to perform preprocessing to generate a standardized energy propagation data set. The energy density field construction module is used for spatial gridding modeling, constructing a three-dimensional discrete volume element set of the goaf, calculating the energy density of each volume element, and generating an energy density field. The gradient and flow direction calculation module is used to calculate the spatial gradient of the energy density field, generate the energy gradient field, and calculate the energy flow direction vector field based on the energy gradient field. The anomaly region extraction module is used to calculate the energy divergence anomaly score field based on the energy density field, energy gradient field and energy flow direction vector field, and to perform threshold segmentation and connected component extraction to generate a set of energy dissipation anomaly regions and a set of energy convergence anomaly regions. The crack orientation extraction module is used to extract the set of energy flow deflection structures by performing streamline tracing based on the energy flow direction vector field within the set of energy dissipation anomaly regions and the set of energy convergence anomaly regions, generate a candidate set of crack orientations, and perform directional consistency screening on the candidate set of crack orientations to obtain the crack orientation set. The measurement result generation module is used to generate measurement results of cracks in coal mine goaf based on the crack orientation set.

Images