Three-dimensional mine earthquake platform network efficiency evaluation method and system based on multi-dimensional index fusion

Abstract

The invention relates to the technical field of earthquake monitoring and mine safety, in particular to a three-dimensional mine earthquake platform network efficiency evaluation method and system based on multi-dimensional index fusion. The method comprises the following steps: constructing a three-dimensional detection grid of a monitoring area, and regarding each grid point as a potential seismic source; calculating a maximum station gap angle corresponding to each seismic source point, and generating a maximum station gap angle thermodynamic diagram; calculating a station convex hull volume space coverage rate; calculating a template matching omission ratio in the noise environment based on template matching; calculating an average information entropy; calculating a positioning uncertainty volume; constructing an effective monitoring area mask based on the maximum station gap angle, the information entropy and the threshold of the positioning uncertainty volume, and calculating an effective monitoring volume; the standardized scores of all the indexes are synthesized, and a comprehensive score is calculated through a geometric averaging method; and outputting a maximum station gap angle thermodynamic diagram, an effective monitoring volume and a comprehensive score. According to the invention, systematic and multi-dimensional quantitative evaluation of the mine earthquake platform network monitoring efficiency is realized, and the evaluation result is comprehensive and accurate.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of seismic monitoring and mine safety, more particularly to a three-dimensional mine seismic station network efficiency evaluation method and system based on multi-dimensional index fusion. BACKGROUND

[0002] Seismic monitoring is of great significance for preventing mine dynamic disasters, ensuring the safety of underground workers and the stable operation of mine production facilities. As a key infrastructure for monitoring mine seismic events, accurately locating the source position and evaluating the impact of mine seismic disasters, the monitoring capability of the mine seismic station network is directly related to the timeliness of mine seismic early warning, the scientificity of disaster analysis and the reliability of emergency response. Therefore, objectively and accurately evaluating the monitoring capability of the mine seismic station network is an important basis and key link for optimizing station layout, improving mine seismic monitoring level and enhancing mine safety prevention and control capability.

[0003] The reasonable layout of the mine seismic monitoring station network directly affects the detection sensitivity of mine seismic events and the accuracy of source positioning, and is the key to ensuring the effectiveness of underground safety production and disaster warning. At present, the traditional evaluation of the monitoring capability of the mine seismic station network has limitations, mostly relying on a single parameter or a few simple indicators for analysis, and usually based on empirical rules for judgment, which is difficult to comprehensively and accurately reflect the comprehensive monitoring performance of the station network in complex three-dimensional geological environment and actual working conditions. The specific manifestations are as follows: Some monitoring capability evaluations only consider the planar geometric distribution of surface stations, ignoring the actual situation that seismic events occur in three-dimensional space and some stations may be located at different depths (such as underground or inside the tunnel), resulting in a lack of stereoscopic and realistic perspective in the evaluation; Some monitoring capability evaluations do not fully consider the influence of key factors such as seismic wave propagation characteristics, signal quality (such as signal-to-noise ratio), environmental noise on monitoring results in the calculation process, resulting in a large deviation between the evaluation results and the actual monitoring performance; The evaluation indicators in the existing monitoring capability evaluation are often analyzed independently, focusing on a single performance, and lack a unified and comprehensive evaluation system, making it difficult to intuitively and systematically present the multi-dimensional and multi-factor evaluation results to managers and decision-makers, resulting in a lack of scientific basis for related departments to optimize the station network, adjust the layout or allocate resources based on the evaluation results, affecting the effectiveness, pertinence and actual disaster prevention and reduction performance of seismic monitoring work.

[0004] The existing monitoring capability evaluation often relies only on single indicators such as station layout density or surface coverage area for analysis, lacking systematic, comprehensive and quantitative evaluation of the overall monitoring performance of the mine seismic station network, which can easily lead to problems such as monitoring blind spots in the station network layout, uneven resource allocation or insufficient response to deep mine seismic events.

[0005] Especially in the complex three-dimensional environment of the mine, the prior art has not formed an effective method which can comprehensively consider the maximum station gap angle of the station spatial distribution, the monitoring information entropy (i.e. data uncertainty), the uncertainty volume of the earthquake source positioning, and the missed detection rate associated with multiple parameters and multiple dimensions, so it is difficult to comprehensively reflect the comprehensive monitoring capability of the mine earthquake network under actual working conditions. SUMMARY

[0006] Therefore, the present application provides a three-dimensional mine earthquake network efficiency evaluation method and system based on multi-dimensional index fusion, aiming to overcome the shortcomings of the existing mine earthquake network monitoring capability evaluation technology, solve the problem that it is difficult to comprehensively and accurately reflect the comprehensive monitoring performance of the network in the complex three-dimensional geological environment and actual working conditions in the traditional mine earthquake network monitoring capability evaluation which relies on single or a few simple indicators, and provide data support and decision basis for optimizing the station layout, improving the mine earthquake monitoring level, and enhancing the mine safety prevention and control capability.

[0007] In order to achieve the above-mentioned purpose, the present application adopts the following technical solutions: In a first aspect, the present application provides a three-dimensional mine earthquake network efficiency evaluation method based on multi-dimensional index fusion, comprising the following steps: Step 1, initialization and three-dimensional modeling: input the three-dimensional spatial coordinates of each station in the mine earthquake network, define the boundary of the mine earthquake monitoring area to be evaluated, and construct a regular three-dimensional detection grid in the monitoring area with a preset grid resolution. Each discrete grid point in the grid is regarded as a potential mine earthquake source position, i.e. a simulated source point; Step 2, key parameter calculation: calculate the maximum station gap angle corresponding to each simulated source position to obtain a maximum station gap angle heat map; calculate the convex hull volume spatial coverage rate; based on the template matching technology, calculate the template matching missed detection rate which reflects the basic detection capability in the noise environment; calculate the average information entropy; calculate the positioning uncertainty volume; Step 3, comprehensive evaluation parameter generation: based on the threshold conditions of the maximum station gap angle, the average information entropy, and the positioning uncertainty volume, construct an effective monitoring area mask and calculate the effective monitoring volume; and based on the station direction coverage uniformity reflected by the maximum station gap angle heat map, the average information entropy, the positioning uncertainty volume, and the template matching missed detection rate, generate a missed detection rate index for comprehensive evaluation; Step 4, calculation of comprehensive score and output of results: based on the standardized scores of the maximum station gap angle heat map, the average information entropy, the positioning uncertainty volume, and the missed detection rate index, calculate the comprehensive score by the geometric mean method; output the maximum station gap angle heat map, the effective monitoring volume, and the comprehensive score.

[0008] In a specific implementation, the step 2, calculating the station convex hull volume spatial coverage, specifically includes: calculating a three-dimensional convex hull volume based on the three-dimensional spatial coordinates of the stations; calculating the spatial coverage based on the three-dimensional convex hull volume and the total volume of the target monitoring area.

[0009] In a specific implementation, the step 2, calculating the maximum station gap angle corresponding to each simulated source location, specifically includes: calculating, for each simulated source location, a projection direction vector of the simulated source location to each station on a specified two-dimensional projection plane; calculating an azimuth angle corresponding to each of the projection direction vectors; calculating the maximum station gap angle based on the azimuth angles; generating a maximum station gap angle heat map based on the maximum station gap angles of all simulated source locations.

[0010] In a specific implementation, the step 2, calculating the template matching miss detection rate in a noise environment, specifically includes: constructing a template seismic signal library; generating a synthetic seismic signal for each simulated source point and superimposing noise; determining whether an event is successfully detected through cross-correlation analysis; calculating the miss detection rate based on the proportion of the number of events that are not successfully detected to the total number of test events.

[0011] In a specific implementation, the step 2, calculating the template matching miss detection rate in a noise environment, further includes: introducing a signal-to-noise ratio constraint condition, and only when the signal-to-noise ratio of the synthetic signal meets the preset condition, the signal is considered to be within the effective detection range.

[0012] In a specific implementation, the step 2, calculating the average information entropy, specifically includes: calculating the relative probability of each possible source location based on the travel time residual and the noise model; calculating the information entropy based on the probability distribution of the relative probability; normalizing the information entropy to obtain normalized information entropy.

[0013] In a specific implementation, the step 2, calculating the positioning uncertainty volume, specifically includes: constructing a covariance matrix based on the coefficient matrix of the travel time observation equation; defining the axis length of the positioning error ellipsoid through the eigenvalues of the covariance matrix; calculating the positioning uncertainty volume based on the eigenvalues.

[0014] In a specific implementable scheme, in step 3, the specific way of constructing the effective monitoring area mask is: Only when the maximum station gap angle, information entropy and positioning uncertainty volume of the potential seismic source position are all better than the preset threshold, the point is regarded as an effective monitoring point; The spatial volume formed by all the effective monitoring points is the effective monitoring volume.

[0015] In a specific implementable scheme, in step 4, the comprehensive score is calculated as follows: The maximum station gap angle heat map, the average information entropy, the positioning uncertainty volume and the normalized score of the missing detection rate index are geometrically averaged to obtain the comprehensive score.

[0016] In a second aspect, the present application provides a three-dimensional mine seismic station network efficiency evaluation system based on multi-dimensional index fusion, which is used to realize the three-dimensional mine seismic station network efficiency evaluation method based on multi-dimensional index fusion as described above, and the system comprises: A three-dimensional modeling module is used to input the three-dimensional spatial coordinates of each station in the mine seismic station network, to determine the boundary of the mine seismic monitoring area to be evaluated, and to construct a regular three-dimensional detection grid in the monitoring area with a preset grid resolution; A parameter calculation module is used to calculate the maximum station gap angle, the station convex hull volume spatial coverage, the template matching missing detection rate in a noise environment, the average information entropy and the positioning uncertainty volume corresponding to each simulated seismic source position; A comprehensive evaluation parameter generation module is used to construct an effective monitoring area mask based on the threshold conditions of the maximum station gap angle, the average information entropy and the positioning uncertainty volume, to calculate an effective monitoring volume, and to generate a missing detection rate index for comprehensive evaluation based on the maximum station gap angle heat map, the average information entropy, the positioning uncertainty volume and the template matching missing detection rate; A comprehensive evaluation and output module is used to calculate a comprehensive score by a geometric mean method based on the maximum station gap angle heat map, the average information entropy, the positioning uncertainty volume and the normalized score of the missing detection rate index, and to output the maximum station gap angle heat map, the effective monitoring volume and the comprehensive score.

[0017] Compared with the prior art, the three-dimensional mine seismic station network efficiency evaluation method and system based on multi-dimensional index fusion have the following beneficial effects: 1. By organically integrating geometric coverage, detection capability, positioning accuracy and information theory indicators in three-dimensional space, a multi-dimensional and three-dimensional comprehensive evaluation system is constructed, the evaluation perspective is comprehensive, and it is more suitable for practical application scenarios. At the same time, by introducing noise interference factors and physical constraint conditions (such as seismic wave propagation velocity model) in the real environment, the evaluation benchmark is changed from the idealized "theoretical optimum" to the "achievable performance" closer to the actual situation, thereby improving the accuracy and practicality of the evaluation results.

[0018] 2. Adopting three-dimensional visualization technology, the originally abstract and complex evaluation indicators are converted into intuitive spatial distribution atlas, which helps to quickly identify the advantage area and monitoring blind area in the layout of the seismic station network, and enhances the intuitiveness and interpretability of the analysis.

[0019] 3. Based on the evaluation results, quantitative indicators such as comprehensive score and effective monitoring volume can be output, which provides direct and clear guidance for the scientific site selection of new seismic monitoring station network, or the optimization and expansion of existing station network, and has important application value. The whole evaluation process is automatically completed by the program, which is not only efficient and reliable, but also can quickly compare and optimize different station layout schemes, greatly improving the evaluation efficiency and decision-making scientificity. BRIEF DESCRIPTION OF DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, a brief introduction will be given below to the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are only embodiments of the present application, and those skilled in the art can obtain other drawings according to the provided drawings without creating any creative labor.

[0021] Figure 1 The flowchart of the three-dimensional mine seismic station network performance evaluation method based on multi-dimensional index fusion according to the present application.

[0022] Figure 2 The three-dimensional distribution diagram of the initial station and the earthquake event.

[0023] Figure 3 The schematic diagram of the station spatial distribution and the convex hull formed thereby.

[0024] Figure 4 is a maximum station gap angle heat map on a two-dimensional plane, wherein: (a) XY plane maximum station gap angle heat map; (b) XZ plane maximum station gap angle heat map; (c) YZ plane maximum station gap angle heat map.

[0025] Figure 5 The information entropy three-dimensional distribution point cloud diagram.

[0026] Figure 6 is an isosurface map of the three-dimensional distribution of information entropy.

[0027] Figure 7 is a point cloud map of the three-dimensional distribution of positioning uncertainty volume.

[0028] Figure 8 is an isosurface map of the three-dimensional distribution of positioning uncertainty volume.

[0029] Figure 9 is a three-dimensional distribution map of the effective monitoring volume.

[0030] Figure 10 is a three-dimensional distribution map of the missed detection rate.

[0031] Figure 11 is a radar chart of the evaluation index. DETAILED DESCRIPTION

[0032] The technical solutions in the embodiments of the present application will be described clearly and completely below. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0033] The three-dimensional mine seismic station network performance evaluation method based on multi-dimensional index fusion provided by the present application is applied to the monitoring capability evaluation of a mine seismic station network. The method regards each grid point as a potential seismic source by constructing a three-dimensional detection grid in a monitoring area to simulate the occurrence of a mine seismic event. The method comprehensively fuses a spatial geometric coverage constraint index (represented by a maximum station gap angle between stations, used to measure the spatial uniformity and directional coverage capability of station distribution), an event monitoring capability index (represented by a template matching missed detection rate in a noise environment, used to reflect the capturing capability of the station network on seismic events in a noise environment), a monitoring data uncertainty index (represented by information entropy, used to quantify the constraint degree of monitoring data on seismic source parameters), and a seismic source positioning accuracy index (represented by a positioning uncertainty volume, used to represent the reliability and stability of the estimation of the seismic source position). Through multi-dimensional fusion analysis of the above four types of indexes, the present application can comprehensively depict the comprehensive monitoring performance of the mine seismic station network from the aspects of spatial geometric distribution, signal detection performance, data information constraint, and positioning accuracy. Finally, the complex multi-index evaluation results are intuitively presented through three-dimensional visualization technology.

[0034] The three-dimensional mine seismic station network performance evaluation method based on multi-dimensional index fusion provided by the present application comprises the following steps: First step: three-dimensional space discretization and station modeling Input the boundary of the monitoring area and the three-dimensional coordinates of the stations (including surface and underground layout), and discretize the monitoring area and the seismic network in three-dimensional space. Each grid point is regarded as the potential seismic source location, i.e., the simulated source point. Step 2: Assessment of the uniformity of station spatial layout (based on spatial geometric coverage) Based on three-dimensional spatial discretization and station modeling, the distribution of the direction vector from each simulated source point to all stations on a unit sphere is calculated. The maximum station gap angle (Angular Gap, AG) in the distribution of all station direction vectors on a unit sphere is analyzed and calculated. A heatmap of the maximum station gap angle is generated to quantify the uniformity of the spatial distribution of stations and the degree of enclosure of each region in three-dimensional space. Step 3: Assessment of Uncertainty in Monitoring Data (Based on Volatility Theory and Probability Statistics) Based on the geological attenuation model of seismic wave propagation, the travel time of seismic waves to each station is simulated, and noise interference that conforms to the actual environment is introduced. The probability distribution entropy values ​​of all potential earthquake source locations are calculated using probabilistic statistical methods, and then the information entropy index is obtained to characterize the uncertainty of the monitoring data. Step 4: Evaluation of event detection capability and positioning accuracy (based on signal simulation) Detection capability assessment: Construct a signal library containing multiple "template seismic events". For each simulated seismic source point, simulate the synthetic signal generated at each station and add noise. By cross-correlation detection with the template library signals, calculate the proportion of undetected points in all test points to obtain the false negative rate. The false negative rate is used to reflect the event detection capability of the network in actual noise environment. Location accuracy assessment: For each seismic source point, a system of linear equations is constructed using travel time residuals and wave direction of arrival information. By solving for the eigenvalues ​​of its covariance matrix, the volume of the location error ellipsoid is estimated to quantify the uncertainty of seismic source location.

[0035] Through the above four steps, the three-dimensional mine seismic network performance evaluation method based on multi-dimensional index fusion described in this invention comprehensively and quantitatively evaluates the monitoring performance of the mine seismic network from multiple dimensions such as spatial geometric coverage, data information entropy, event detection probability, and source location accuracy.

[0036] The process of the three-dimensional seismic network performance evaluation method based on multi-dimensional index fusion described in this invention is as follows: Figure 1As shown, the core innovation of the method is to build a multi-parameter coordinated mine seismic network efficiency evaluation process. Through the systematic integration of mine seismic network spatial layout information, source event simulation, signal detection and monitoring error analysis, the comprehensive monitoring capability of the network is quantitatively analyzed. From the spatial coverage capability (maximum station gap angle), event detection performance (template matching missed detection rate in noise environment), positioning accuracy level (positioning uncertainty volume) and monitoring uncertainty (average information entropy), the comprehensive monitoring capability of the mine seismic network in complex three-dimensional mine environment is comprehensively and quantitatively described. This method breaks through the limitations of traditional evaluation methods, which rely on single or few indicators, ignore the depth distribution of stations and the characteristics of underground sources, and do not fully consider the influence of seismic wave propagation and noise interference. It realizes scientific, systematic and multi-angle evaluation of the monitoring efficiency of mine seismic network. Specifically, the method mainly includes the following steps: 1. Initialization and 3D modeling First, input the three-dimensional spatial coordinates of each station in the mine seismic network (including surface and underground layout information at different depths), and specify the boundary of the mine seismic monitoring area to be evaluated. On this basis, a regular three-dimensional detection grid is constructed in the monitoring area with a preset grid resolution. Each discrete grid point in the grid is considered as a potential mine seismic source location, i.e., a simulated source point. As shown in Figure 2 The three-dimensional spatial distribution of initial stations and seismic event grid points in the mine seismic monitoring system is shown. The red triangles in the figure represent the monitoring station positions on the surface and underground at different depths in the mine area. The gray dot array represents a regular three-dimensional detection grid established in the monitoring area with a preset grid resolution. Each intersection point of the grid is considered as a potential mine seismic source location, i.e., a simulated source point. Through this discrete modeling method, all possible mine seismic spatial locations in the target monitoring area can be fully covered, providing a unified and high-resolution spatial reference framework for subsequent source simulation, travel time calculation, signal detection, error analysis and other steps, ensuring spatial consistency and computational repeatability in the evaluation process.

[0037] 2. Key parameter calculation: (1) Station convex hull volume spatial coverage rate calculation To quantitatively evaluate the spatial coverage capability of the mine seismic network in the target monitoring area, the method first introduces the key index of station convex hull volume spatial coverage rate, which is used to describe the area that can be "effectively surrounded" by the mine seismic station array in the three-dimensional monitoring space, so as to reflect the overall coverage degree of the network layout to the monitoring area and the spatial uniformity of the monitoring capability. Specifically, the index is obtained by calculating the three-dimensional convex hull volume composed of the three-dimensional coordinates of the mine seismic stations, and comparing it with the total volume of the preset target monitoring area, and then a standardized spatial coverage rate value is obtained; the higher the value, the larger the spatial surrounding range formed by the station in the monitoring area, the more comprehensive the monitoring coverage of the potential mine earthquake event, and the stronger the overall spatial monitoring capability of the network.

[0038] In three-dimensional space, the "convex hull" refers to the smallest convex polyhedron composed of the three-dimensional coordinates of the given mine seismic stations, which can completely contain all the station points and itself is a convex geometric body without concave parts. In the geometric sense, the convex hull defines the outermost boundary of the station group in space, which is an important geometric feature for evaluating the spatial extension and distribution form of the station array.

[0039] As for the specific calculation method of the convex hull volume, it is usually based on the triangular mesh patches composed of the convex hull: the convex hull surface is composed of multiple triangular patches, and its overall volume can be calculated by decomposing the convex hull into a series of tetrahedrons (taking a point inside the convex hull as the common vertex Each patch is a bottom surface, so any tetrahedron is composed of vertices , , , and the volume of each tetrahedron is added up. In specific implementation, the vector determinant method is often used, and the formula is as follows: Wherein, is the coordinate of the th vertex of the convex hull; is the number of convex hull faces; V is the three-dimensional convex hull volume formed by the station points; det(·) is the three-dimensional vector determinant used to calculate the volume of the tetrahedron.

[0040] To more accurately reflect the spatial efficiency of the mine seismic station in the actual monitoring process, the method introduces an "effective coverage coefficient" based on the traditional convex hull volume, thereby obtaining the "network effective coverage volume". This coefficient considers the actual monitoring capability of the station, such as sensor sensitivity, signal propagation attenuation characteristics, station layout depth (such as mixed layout of underground and surface), and spatial signal coverage radius. For example, if the effective monitoring radius of a single station is set to k, it is considered that the station not only has monitoring ability inside the convex hull, but also can effectively perceive the earthquake event within a certain range outside the convex hull; therefore, the effective coverage volume of the station network can be regarded as the expanded volume under the joint action of the convex hull volume and the effective monitoring radius, and its calculation method can be geometric inflation, density weighting or correction method based on propagation model according to actual needs. Finally, the formula for calculating the spatial coverage rate is: Among them, "the effective coverage volume of the station network" is used to reflect the spatial service ability of the station under actual monitoring conditions; "the total volume of the target monitoring area" is to evaluate the preset area.

[0041] Figure 3 The actual layout position of the mine earthquake station in the three-dimensional space, the geometric shape of the three-dimensional convex hull composed of these station points, and the spatial relationship between the convex hull and the monitoring area boundary are shown. Among them, the red triangle represents the spatial distribution position of the station in the mine earthquake network; the blue translucent polyhedron represents the three-dimensional convex hull composed of all station points; the coordinate axes correspond to the X, Y and Z directions in the mine coordinate system.

[0042] (2) Maximum station gap angle analysis and heat map construction In order to further evaluate the direction coverage uniformity of the mine earthquake station network in the three-dimensional monitoring area from the spatial geometric angle, this method introduces the maximum station gap angle analysis index, and constructs the heat map accordingly. The maximum station gap angle can further quantitatively evaluate the uniformity of the station distribution in the monitoring area and its coverage ability in each direction of the space from the spatial geometric perspective.

[0043] In the three-dimensional mine earthquake station monitoring, the maximum station gap angle is a key index to measure the uniformity of the spatial distribution of the station. It represents the maximum angle blank area formed by the station direction vector in the spherical projection of the observation point in the monitoring area, that is, the maximum uncovered angle between the directions of adjacent stations. This angle reveals the "coverage gap" of the station in the spatial direction from the perspective of spherical geometry. The smaller the maximum station gap angle, the more continuous and uniform the coverage of the station in each direction of the space, and the positioning accuracy is usually higher; on the contrary, if the maximum station gap angle is large, it means that there is a obvious monitoring direction blind area, which may lead to increased error in earthquake source positioning, increased risk of missing detection in some areas, and even affect the robustness of the whole monitoring network.

[0044] In order to comprehensively evaluate the uniformity of the station coverage in three-dimensional space, this method projects the maximum station gap angle onto the XY, XZ and YZ orthogonal planes for two-dimensional analysis, and calculates the maximum station gap angle in each plane. The specific calculation process is as follows: For any earthquake source point in the monitoring area , first calculate its distance to each station the projection direction vector of each grid point: where plane represents xy, xz or yz.

[0045] Subsequently, each projection direction vector is normalized to a unit vector: In the two-dimensional projection plane, the azimuth angle (in degrees, 0° to 360°) corresponding to each unit vector is calculated: XY plane: azimuth angle XZ plane: azimuth angle YZ plane: azimuth angle All the calculated azimuth angles are sorted in ascending order, the difference between adjacent azimuth angles is calculated, and the maximum value among these differences is found, which is the maximum station gap angle of the projection plane at the source point : where represents the difference between adjacent azimuth angles after sorting; is the maximum monitoring blind angle in the given plane.

[0046] Finally, by calculating and statistics the maximum station gap angle values of all grid points (i.e. all potential source locations) in the monitoring area, a maximum station gap angle heat map can be generated, which directly displays the uniformity of station direction coverage at different positions in the monitoring area. Figure 4 shows the maximum station gap angle distribution angle heat map of the mine seismic monitoring network in three orthogonal planes, where (a) is the XY plane maximum station gap angle distribution, reflecting the uniformity of station layout in the horizontal spatial direction on the ground surface. The purple area indicates a smaller maximum station gap angle and uniform coverage, while the yellow area indicates a larger maximum station gap angle and potential blind area. (b) is the XZ plane maximum station gap angle distribution, showing the geometric coverage ability of the mine seismic monitoring system in the vertical depth direction. If the maximum station gap angle in the deep region increases, it indicates that the underground station coverage is insufficient. (c) is the YZ plane maximum station gap angle distribution, showing the coverage distribution along another vertical profile, which is complementary to the XZ plane result, and can directly reflect the observation gap in different depth directions.

[0047] The differences in the maximum station gap angle distribution across different planes in Figure 4 reveal the coverage characteristics of the monitoring network in different spatial directions. Empirically, it is generally believed that when the average maximum station gap angle is ≤90°, the spatial distribution of stations is relatively uniform, and the monitoring coverage is sufficient. However, if the maximum station gap angle in some areas is >270°, it indicates a significant lack of monitoring direction. In such cases, prioritizing the addition of new stations or adjusting the existing station layout is recommended to eliminate monitoring blind spots and improve overall monitoring efficiency and positioning reliability.

[0048] 3. Evaluate detection and localization capabilities in noisy environments. In evaluating the monitoring capabilities of mine seismic networks, besides spatial coverage and geometric uniformity, the detection sensitivity and source location accuracy of stations under actual noise interference conditions are also key aspects for measuring their overall performance. This method systematically evaluates the monitoring reliability and scientific validity of mine seismic networks under complex actual working conditions from three interrelated dimensions: event detection probability in noisy environments (i.e., missed detection rate), uncertainty of location information (i.e., information entropy), and error range of source location estimation (i.e., location uncertainty volume).

[0049] (1) Template matching false negative rate The false negative rate is a core indicator for evaluating the effectiveness of a seismic network in detecting potential seismic events under noisy conditions. It is defined as the proportion of seismic events that cannot be correctly identified or detected in a given noisy environment out of all tested events. A lower false negative rate indicates a stronger ability of the network to capture weak signals (such as small earthquakes or deep mine-related seismic events), directly reflecting the network's monitoring sensitivity and reliability. In short, the false negative rate directly reflects the network's ability to detect seismic events. It is worth noting that in three-dimensional space, due to uneven spatial distribution of stations, signal obstruction, or local differences in noise interference, the false negative rate at different locations often exhibits significant spatial heterogeneity.

[0050] This method uses template matching technology to quantitatively calculate the missed detection rate. First, high-fidelity numerical simulation results from the mining area are selected to construct a template seismic signal library. This library contains standard seismic waveforms (i.e., "template earthquakes") from multiple representative sources. Each template signal is normalized and corrected for signal-to-noise ratio, reflecting the time-frequency and waveform characteristics of typical seismic signals. Then, a group of "simulated source points" are set up in the same monitoring grid as potential events to be detected. For each simulated source point, the theoretical travel time to all stations is calculated based on the velocity model, generating a corresponding synthetic seismic signal. A noise model conforming to the seismic wave propagation law and the noise characteristics of the mining area is superimposed on this signal to simulate the waveform signal under actual observation conditions. Next, cross-correlation template matching analysis is performed between the synthetic signal and each template in the template library. If the following parallel judgment condition is met, the event is considered successfully detected: the cross-correlation coefficient is high at at least M stations. all exceed the threshold ; signal-to-noise ratio of each station all are not less than a preset minimum signal-to-noise ratio ; the event equivalent energy E reaches a minimum detection energy threshold If any condition is not met, it is considered as a missed detection.

[0051] Unlike traditional template matching detection, the template matching process of the present method is not used to directly identify the actual event in the continuous waveform, but to simulate the "detectability" of the network under the noise condition, that is, the probability of successful template matching represents the probability that the event at the position can be effectively identified by the network. In order to obtain the monitoring probability in a continuous sense, K times of template matching simulation are repeatedly performed for each simulated seismic source point under different noise random fields. After K times of repeated simulation, the detection probability of the test point can be calculated: Wherein, the symbol "∧" represents the "and (AND)" relationship, and 1(·) is an indicator function; missed detection rate The calculation formula is: After obtaining the detection probability of each grid point in the whole domain, After obtaining the detection probability of each grid point in the whole domain, Wherein, represents the total number of grid points in the monitoring area.

[0052] In addition, in order to ensure the effectiveness of the detection result, the signal-to-noise ratio (SNR) constraint condition is introduced in the present method, that is, only when the signal-to-noise ratio of the synthesized signal is greater than or equal to the preset minimum signal-to-noise ratio (SNR min ), it is considered that the signal is within the effective detection range, which is defined as: .

[0053] (2) Calculate the average information entropy In the evaluation of the monitoring capability of mine seismic network, the average information entropy is a key index based on the Shannon information entropy theory, which is used to quantitatively measure the dispersion degree of the posterior probability distribution of the seismic event source position under the given station layout and noise environment, that is, the uncertainty of the positioning result. The higher the entropy value, the weaker the constraint ability of the observation data on the source position, and the more unreliable the positioning result; on the contrary, the lower the entropy value, the more concentrated the probability distribution of the source position, and the higher the positioning accuracy.

[0054] Based on the Bayesian inference principle, for each test point Firstly, the travel-time residuals of each station are calculated: where, is the travel-time residual of the ith station, is the travel-time matrix pre-computed based on the velocity model and the grid, is the simulated observed travel-time.

[0055] Subsequently, to simulate the noise disturbance in real environment, a distance-dependent noise term is introduced for each station: , where, is the distance from the station to the test point, is the base noise level, where the coefficients 0.5206 and 0.0481 are empirical parameters fitted by the measured noise data, is the empirical noise model.

[0056] Based on the Gaussian likelihood function, the relative probability of each possible source location is further calculated: where N is the total number of stations involved in the calculation, is the posterior probability of the grid point corresponding to the test point.

[0057] After the probability distribution is normalized, the information entropy H is defined as: The entropy value is normalized to the interval [0, 1]: where, is the number of grid points. The lower the normalized information entropy, the smaller the uncertainty of the source location in that area, and the stronger the positioning ability of the network for the seismic event.

[0058] The isosurface map of the three-dimensional distribution of information entropy ( Figure 6 ) shows the spatial isosurface morphology of the normalized information entropy in the monitoring volume. The high-entropy area corresponds to the spatial range with weak network positioning constraints and large errors. The point cloud map of the three-dimensional distribution of information entropy ( Figure 5 ) displays the same result in the form of discrete grid points. The color of the points represents the local information entropy. Both maps are based on Figure 2The monitoring grid and station three-dimensional coordinates are shown, and the raw data includes the spatial coordinates of each station, the grid node position and the noise model. The calculation process of information entropy is completed by using the above Bayesian likelihood estimation method, and the results obtained are plotted into point cloud and isosurface form by three-dimensional visualization module, so as to intuitively reveal the spatial distribution characteristics of the positioning uncertainty in the mine earthquake monitoring space. The high-entropy area indicates the area with weak positioning ability and dispersed source probability distribution, which provides a basis for subsequent network optimization and key monitoring area identification.

[0059] (3) Positioning uncertainty volume The positioning uncertainty volume is the core index for evaluating the estimation accuracy of the mine earthquake network to the source position. Its essence is to describe the possible distribution range of the source parameters in the three-dimensional space through statistical methods, reflecting the comprehensive influence of the station geometric layout, travel time observation error and noise environment on the source positioning result. The larger the volume is, the wider the possible spatial range of the source is, and the lower the reliability of the positioning result is; on the contrary, the positioning accuracy is higher, and the source position is closer to the true value.

[0060] Based on the linearization theory, the method constructs the coefficient matrix of the travel time observation equation, calculates the covariance matrix, and then extracts the eigenvalues to define the axis length of the positioning error ellipsoid, and finally quantitatively describes the positioning uncertainty through the ellipsoid volume.

[0061] Assuming that the positioning problem is linearized, the positioning error covariance matrix C is: Where A is the direction cosine matrix (each row corresponds to the direction vector of a station [cosα, cosβ, cosγ]) reflecting the influence of station geometric layout on positioning accuracy; , is the standard deviation of travel time error, representing the level of single station travel time error; d is the distance from the source to the station.

[0062] The eigenvalues of the covariance matrix C , , represent the three-axis length of the error ellipse, and their square roots represent the positioning standard deviation in X, Y and Z directions respectively. Therefore, the volume of the error ellipsoid is: If the covariance matrix cannot be directly calculated, the average distance between the station and the source can be used as an approximation: This index provides a direct geometric reference for network optimization design, helps to identify areas with weak positioning ability, and guides station supplement and layout adjustment.

[0063] ​The three-dimensional distribution isosurface map of positioning uncertainty volume Figure 8 shows the isosurface morphology of positioning uncertainty volume at different depths and spatial positions, and the outer packaging area of the isosurface corresponds to the spatial range with lower positioning accuracy; the three-dimensional distribution point cloud map of positioning uncertainty volume Figure 7 shows the discrete visualization results of the same data, each point represents a grid node, and the color corresponds to the positioning uncertainty value. Figure 7 With Figure 8 , the spatial differences of positioning accuracy in the mine earthquake monitoring area are intuitively reflected, the high uncertainty volume area represents the area with insufficient station geometric constraints and large travel time error, and the low uncertainty area represents the spatial range with strong station network positioning ability and stable and reliable results, which provides quantitative basis for subsequent network optimization and regional monitoring capacity improvement.

[0064] 4. Calculate the effective monitoring area and comprehensive score After completing the quantitative analysis and spatial distribution characterization of the key performance indicators of the mine earthquake network in three-dimensional space (including the maximum station gap angle, information entropy, positioning uncertainty, and missed detection rate), the method further constructs an effective monitoring area mask to identify the three-dimensional spatial range that the network can reliably and effectively capture earthquake events in the actual monitoring process from the perspective of multi-index coordination, and calculates the effective monitoring volume accordingly, providing spatial and quantitative support for the overall monitoring efficiency of the network.

[0065] Specifically, the effective monitoring area is defined as: in the three-dimensional monitoring grid, only when the maximum station gap angle (AG), information entropy (H), and positioning uncertainty volume (V) of the potential source location are better than the pre-set reasonable threshold, the point is considered as an “effective monitoring point”; the spatial volume composed of the collection of all such effective monitoring points is the effective monitoring volume (V Figure 9 ). This volume directly reflects the spatial range of the mine earthquake network that truly has reliable monitoring and precise positioning ability in the monitoring area, and is a key indicator for evaluating the actual “availability” and “service coverage” of the network.

[0066] Figure 10The effective monitoring area distribution of the network in three-dimensional space is intuitively displayed in the form of a three-dimensional point cloud, and the spatial range in which the network can reliably work is clearly indicated, thereby providing direct visual basis for network layout optimization and identification of key monitoring areas. The index reflects the "perception ability" and "response ability" of the network to seismic events from a macro perspective. The lower the missed detection rate, the higher the monitoring sensitivity of the network to various earthquakes (especially small earthquakes and weak signals), and the stronger the event capture ability. It should be noted that, in order to ensure the objectivity and scientificity of the missed detection rate calculation, the "total number of actual earthquakes" should be determined based on independent and reliable reference standards, such as the earthquake records of adjacent high-sensitivity networks, authoritative historical earthquake catalogs or regional seismic observation network data, to avoid relying solely on the limited data of the network to be evaluated, thereby effectively avoiding the self-reference bias and underestimation of false negatives.

[0067] To further comprehensively reflect the overall performance of the mine seismic network in terms of spatial coverage, geometric uniformity, signal detection, positioning accuracy and uncertainty control, the geometric mean method is used to calculate the scores of the aforementioned four core evaluation indexes (i.e. maximum station gap angle, average information entropy, positioning uncertainty volume and missed detection rate) to obtain a composite score (Composite Score), which can be mathematically expressed as: Composite Score The composite score quantifies the overall monitoring performance of the mine seismic network in a complex three-dimensional geological and actual noise environment. The higher the value, the more balanced the performance of the network in multiple key dimensions and the stronger the overall monitoring capability. Conversely, it indicates that the network has obvious short boards in some aspects and needs to be optimized.

[0068] Figure 11 The evaluation index radar chart is an integrated and visualized display of the final evaluation results of the method. By presenting the scores of the network in the maximum station gap angle, information entropy, missed detection rate and positioning uncertainty in the form of a radar chart, the performance of the mine seismic network in different monitoring capability dimensions is intuitively reflected. This chart not only provides a "one-stop" evaluation view of the overall performance of the network, but also provides intuitive clues for identifying specific short boards, thereby providing scientific basis and clear direction for subsequent network optimization decisions (such as adding stations, adjusting station depth, optimizing instrument parameters, strengthening noise control, etc.).

[0069] It is worth mentioning that the method is not only suitable for the evaluation of single mine seismic network, but also can be applied to the comparison and analysis of network capacity under different mine areas, different layout schemes and different monitoring targets. Through the horizontal comparison of the evaluation results of multiple networks or different design schemes, the advantages and disadvantages of various networks in monitoring capacity can be effectively identified, which provides strong theoretical support and technical guidance for the scientific planning, reasonable layout, dynamic optimization and performance improvement of mine seismic monitoring network.

[0070] The various embodiments in the specification are described in a progressive manner, and each embodiment focuses on the difference from other embodiments. The same or similar parts between the various embodiments can be referred to each other. The above description of the disclosed embodiments enables a person skilled in the art to implement or use the present application. Various modifications to the embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A three-dimensional seismic network performance evaluation method based on multi-dimensional index fusion, characterized in that, Includes the following steps: Step 1, Initialization and 3D Modeling: Input the 3D spatial coordinates of each station in the seismic network, clarify the boundary of the seismic monitoring area to be evaluated, and construct a regular 3D detection grid in the monitoring area with a preset grid resolution. Each discrete grid point in the grid is regarded as a potential seismic source location, i.e., a simulated seismic source point. Step 2, Calculation of key parameters: Calculate the maximum station gap angle corresponding to each simulated seismic source location to obtain a heat map of the maximum station gap angle; Calculate the spatial coverage of the convex hull of the station; based on template matching technology, calculate the template matching false negative rate, which reflects the basic detection capability under noisy conditions; Calculate the average information entropy; Calculate the volume of positioning uncertainty; Step 3, Generation of comprehensive evaluation parameters: Based on the threshold conditions of the maximum station gap angle, the average information entropy, and the positioning uncertainty volume, construct an effective monitoring area mask and calculate the effective monitoring volume; and based on the station direction coverage uniformity reflected by the maximum station gap angle heatmap, the average information entropy, the positioning uncertainty volume, and the template matching false negative rate, generate a false negative rate index for comprehensive evaluation. Step 4, Calculate the comprehensive score and output the results: Based on the standardized scores of the maximum station gap angle heatmap, the average information entropy, the positioning uncertainty volume, and the missed detection rate index, the comprehensive score is calculated using the geometric mean method; Output the maximum station gap angle heatmap, the effective monitoring volume, and the comprehensive score.

2. The method for evaluating the effectiveness of a three-dimensional mine seismic network based on multi-dimensional index fusion as described in claim 1, characterized in that, In step 2, the spatial coverage of the station's convex hull volume is calculated as follows: Calculate the three-dimensional convex hull volume based on the three-dimensional spatial coordinates of the station; The spatial coverage rate is calculated based on the three-dimensional convex hull volume and the total volume of the target monitoring area.

3. The method for evaluating the effectiveness of a three-dimensional mine seismic network based on multi-dimensional index fusion as described in claim 1, characterized in that, In step 2, the maximum station gap angle corresponding to each simulated seismic source location is calculated, specifically including: For each simulated seismic source location, calculate its projection direction vector to each station on the specified two-dimensional projection plane; Calculate the azimuth angle corresponding to each of the projection direction vectors; Calculate the maximum station gap angle based on the azimuth angle; A heatmap of the maximum station gap angle is generated based on the maximum station gap angle at all simulated seismic source locations.

4. The method for evaluating the effectiveness of a three-dimensional mine seismic network based on multi-dimensional index fusion as described in claim 1, characterized in that, In step 2, the template matching false negative rate under noisy conditions is calculated as follows: Construct a template seismic signal library; For each simulated seismic source point, a synthetic seismic signal is generated and noise is superimposed; Cross-correlation analysis is used to determine whether an event has been successfully detected. The false negative rate is calculated based on the ratio of the number of undetected events to the total number of test events.

5. The method for evaluating the effectiveness of a three-dimensional mine seismic network based on multi-dimensional index fusion as described in claim 4, characterized in that, The calculation of template matching false negative rate under noisy conditions also includes: A signal-to-noise ratio (SNR) constraint is introduced, and the signal is considered to be within the effective detection range only when the SNR of the synthesized signal meets the preset condition.

6. The method for evaluating the effectiveness of a three-dimensional seismic network based on multi-dimensional index fusion as described in claim 1, characterized in that, In step 2, the average information entropy is calculated as follows: The relative probability of each possible source location is calculated based on the travel time residual and noise model; The information entropy is calculated based on the probability distribution of the relative probabilities; The information entropy is normalized to obtain the normalized information entropy.

7. The method for evaluating the effectiveness of a three-dimensional seismic network based on multi-dimensional index fusion as described in claim 1, characterized in that, In step 2, the positioning uncertainty volume is calculated, specifically as follows: Construct the covariance matrix based on the coefficient matrix of the travel-time observation equation; The axial length of the positioning error ellipsoid is defined by the eigenvalues ​​of the covariance matrix. The location uncertainty volume is calculated based on the aforementioned eigenvalues.

8. The method for evaluating the effectiveness of a three-dimensional seismic network based on multi-dimensional index fusion as described in claim 1, characterized in that, In step 3, the specific method for constructing the effective monitoring area mask is as follows: A point is considered a valid monitoring point only when the maximum station gap angle, information entropy, and location uncertainty volume of a potential seismic source location are all better than a preset threshold. The volume of space formed by all effective monitoring points is the effective monitoring volume.

9. The method for evaluating the effectiveness of a three-dimensional mine seismic network based on multi-dimensional index fusion as described in claim 1, characterized in that, In step 4, the comprehensive score is calculated as follows: The comprehensive score is obtained by geometrically averaging the standardized scores of the maximum station gap angle heatmap, the average information entropy, the positioning uncertainty volume, and the missed detection rate index.

10. A three-dimensional mine seismic network performance evaluation system based on multi-dimensional index fusion, characterized in that, The system for implementing the three-dimensional seismic network performance evaluation method based on multi-dimensional index fusion as described in any one of claims 1 to 9, comprises: The 3D modeling module is used to input the 3D spatial coordinates of each station in the seismic network, clarify the boundary of the seismic monitoring area to be evaluated, and construct a regular 3D detection grid within the monitoring area with a preset grid resolution. The parameter calculation module is used to calculate the maximum station gap angle, station convex hull volume spatial coverage, template matching missed detection rate, average information entropy, and positioning uncertainty volume corresponding to each simulated seismic source location. The comprehensive evaluation parameter generation module is used to construct an effective monitoring area mask based on the threshold conditions of the maximum station gap angle, the average information entropy, and the positioning uncertainty volume, calculate the effective monitoring volume, and generate a false negative rate index for comprehensive evaluation based on the maximum station gap angle heatmap, the average information entropy, the positioning uncertainty volume, and the template matching false negative rate. The comprehensive evaluation and output module is used to calculate a comprehensive score based on the standardized scores of the maximum station gap angle heatmap, the average information entropy, the positioning uncertainty volume, and the missed detection rate index using the geometric mean method, and output the maximum station gap angle heatmap, the effective monitoring volume, and the comprehensive score.

Images