Main earthquake group intelligent identification method considering micro-earthquake contour coefficient and spectrogram clustering

Abstract

The invention discloses a main earthquake group intelligent identification method considering a micro-earthquake contour coefficient and spectrogram clustering. The method comprises the following steps: 1, acquiring micro-earthquake monitoring data; 2, acquiring a spatial similarity matrix, an energy similarity matrix and a time similarity matrix, and weighting to form a comprehensive similarity matrix; 3, forming a feature matrix based on the comprehensive similarity matrix, and mapping data points in the feature matrix into two-dimensional space coordinates; 4, clustering data points in the two-dimensional space coordinates by adopting a DBSCAN algorithm, and obtaining a neighborhood radius and a minimum sample number optimization combination based on a micro-seismic contour coefficient; and 5, under the optimization combination of the neighborhood radius and the minimum sample number, clustering data points in the two-dimensional space coordinates by adopting a DBSCAN algorithm to obtain a plurality of micro-seismic event optimization clusters. And 6, screening out a micro-seismic event cluster with the highest comprehensive feature score as a main seismic group. The method is reasonable in design, and the main earthquake group is selected by fusing evaluation indexes of space density, energy characteristics and time characteristics.

Description

Technical Field

[0001] This invention belongs to the field of microseismic activity mainshock group identification technology, specifically involving an intelligent identification method for mainshock groups that considers microseismic profile coefficients and spectral clustering. Background Technology

[0002] In recent years, with the increase in the depth and intensity of coal mining, the dynamic adjustment and energy dissipation processes of mining-induced stress fields have exhibited highly nonlinear characteristics. Traditional analysis methods based on static monitoring or single microseismic parameters are no longer sufficient to accurately characterize stress migration paths and potential hazardous areas. Microseismic monitoring technology, due to its high sensitivity to rock mass fracturing events, has become an effective means of studying the evolution of mining-induced stress fields, especially focusing on mainshock swarms. However, existing mainshock swarm identification research still faces significant technical bottlenecks: firstly, traditional clustering methods (such as K-means) only focus on the spatial distribution of microseismic events, ignoring energy parameters and time series characteristics, leading to insufficient accuracy in mainshock swarm identification; secondly, the evaluation indicators for mainshock swarm identification are singular, using only basic parameters such as energy values.

[0003] Therefore, there is currently a lack of a well-designed intelligent identification method for mainshock groups that considers both microseismic profile coefficients and spectral clustering. In spectral clustering, a comprehensive similarity matrix is ​​formed by weighting the spatial similarity matrix, energy similarity matrix, and temporal similarity matrix. By considering the microseismic profile coefficients to obtain the neighborhood radius and the minimum number of samples, an optimized cluster of microseismic events is obtained. Furthermore, by integrating evaluation indicators such as spatial density, energy characteristics, and temporal characteristics, the mainshock group is selected, effectively extracting key microseismic event clusters that reflect stress concentration areas. Summary of the Invention

[0004] The technical problem to be solved by this invention is to address the shortcomings of the prior art by providing an intelligent identification method for mainshock groups that considers microseismic profile coefficients and spectral clustering. The method is simple in steps and reasonable in design. In spectral clustering, the spatial similarity matrix, energy similarity matrix and temporal similarity matrix are weighted to form a comprehensive similarity matrix. Considering the microseismic profile coefficients, the optimized combination of neighborhood radius and minimum sample number is obtained to obtain optimized clusters of microseismic events. The mainshock groups are selected by integrating evaluation indicators of spatial density, energy characteristics and temporal characteristics, effectively extracting key microseismic event clusters that reflect stress concentration areas.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is: an intelligent identification method for mainseismic groups considering microseismic profile coefficients and spectral clustering, characterized in that the method includes the following steps: Step 1: Obtain microseismic monitoring data: Step 2: Obtain the spatial similarity matrix, energy similarity matrix, and temporal similarity matrix based on microseismic monitoring data, and then weight them to form a comprehensive similarity matrix; Step 3: Form a feature matrix based on the comprehensive similarity matrix and map the data points in the feature matrix to two-dimensional spatial coordinates: Step 4: Use the DBSCAN algorithm to cluster data points in two-dimensional spatial coordinates, and optimize the combination of neighborhood radius and minimum number of samples based on the microseismic profile coefficient; Step 5: Under the optimized combination of neighborhood radius and minimum sample number, the DBSCAN algorithm is used to cluster the data points in the two-dimensional spatial coordinates to obtain multiple optimized clusters of microseismic events; Step 6: Extract the spatial density, energy characteristics, and temporal characteristics of each microseismic event optimized cluster, and obtain a weighted comprehensive feature score. Select the microseismic event cluster with the highest comprehensive feature score as the mainshock group.

[0006] The above-mentioned intelligent identification method for mainshock groups that considers microseismic profile coefficients and spectral clustering further includes, in step one, using a microseismic monitoring system to monitor the mine area to be monitored, obtaining microseismic events per unit time and the occurrence time, spatial coordinates, and energy values ​​of each microseismic event per unit time; Step two, the specific process is as follows: Step 201: Normalize the spatial coordinates of each microseismic event using the proportional normalization method to obtain the spatial normalized coordinates corresponding to each microseismic event. Step 202: Record each microseismic event in chronological order of occurrence as the first microseismic event, ..., the i-th microseismic event, ..., the I-th microseismic event; where i is a positive integer and 1 ≤ i ≤ I; Step 203: Obtain the spatial similarity matrix based on the spatially normalized coordinates corresponding to I microseismic events. ; Step 204: Obtain the energy similarity matrix based on the energy values ​​corresponding to I microseismic events. ; Step 205: Obtain the time similarity matrix based on the times corresponding to I microseismic events. ; Step 206: Based on the spatial similarity matrix Energy similarity matrix and time similarity matrix The weighted averages are used to form a comprehensive similarity matrix.

[0007] The above-mentioned intelligent mainshock group identification method considering microseismic profile coefficients and spectral clustering further includes step 203, which is specifically as follows: Step 2031: Using a computer according to The spatial distance between the i-th microseismic event and the j-th microseismic event is obtained. ;in, Represents the spatial normalized coordinates of the i-th microseismic event. Let represent the spatial normalized coordinates of the j-th microseismic event; j is a positive integer, and 1≤j≤1; Step 2032: Repeat step 2031 multiple times to obtain the spatial distance between any two microseismic events, and obtain the maximum spatial distance from all spatial distances between two microseismic events. ; Step 2033, according to The spatial similarity between the i-th microseismic event and the j-th microseismic event is obtained. ; Step 2034: Repeat step 2033 multiple times to obtain I×I spatial similarities, and use the I×I spatial similarities as matrix elements to obtain the spatial similarity matrix. .

[0008] The above-mentioned intelligent mainshock group identification method considering microseismic profile coefficients and spectral clustering further includes step 204, which is specifically as follows: Step 2041: Using a computer according to The energy difference between the i-th microseismic event and the j-th microseismic event is obtained. ;in, This represents the energy value of the i-th microseismic event. This represents the energy value of the j-th microseismic event; Step 2042: Repeat step 2041 multiple times to obtain the energy difference between any two microseismic events, and obtain the maximum energy difference from all energy differences between two microseismic events. ; Step 2043, according to The energy similarity between the i-th microseismic event and the j-th microseismic event is obtained. ; Step 2044: Repeat step 2043 multiple times to obtain I×I energy similarities, and use these I×I energy similarities as matrix elements to obtain the energy similarity matrix. .

[0009] The above-mentioned intelligent mainshock group identification method considering microseismic profile coefficients and spectral clustering further includes step 205, which is as follows: Step 2051: Using a computer according to... The time difference between the i-th microseismic event and the j-th microseismic event is obtained. ;in, This represents the time of occurrence of the i-th microseismic event. This indicates the time of occurrence of the j-th microseismic event; Step 2052: Repeat step 2051 multiple times to obtain the time difference between any two microseismic events, and obtain the maximum time difference from all the time differences between two microseismic events. ; Step 2053, according to The time similarity between the i-th microseismic event and the j-th microseismic event is obtained. ; Step 2054: Repeat step 2053 multiple times to obtain I×I time similarity scores, and use these I×I time similarity scores as matrix elements to obtain a time similarity matrix. ; Step 206, the specific process is as follows: according to The comprehensive similarity matrix is ​​obtained. ;in, Weights representing spatial similarity Weights representing energy similarity Weights representing temporal similarity. All are located between 0 and 1, and .

[0010] The above-mentioned intelligent mainshock group identification method considering microseismic profile coefficients and spectral clustering further includes the following step three: Step 301: Construct a Laplacian matrix based on the comprehensive similarity matrix, and perform eigenvalue decomposition on the Laplacian matrix to obtain eigenvalues ​​and corresponding eigenvectors; Step 302: Arrange the eigenvalues ​​in ascending order, remove eigenvalues ​​with zero values ​​and duplicate eigenvalues, and then use the eigenvectors corresponding to the remaining eigenvalues ​​as column vectors to form the original eigenma matrix; Step 303: Reduce the dimensionality of the original feature matrix using PCA to obtain the reduced matrix, and denote the reduced matrix as the feature matrix. Map the data points in the feature matrix to two-dimensional spatial coordinates.

[0011] The above-mentioned intelligent mainshock group identification method considering microseismic profile coefficients and spectral clustering further includes step four, the specific process of which is as follows: Step 401: Set the minimum number of samples min_samples to a value of 3 to 5, and determine the neighborhood radius eps value based on the K-distance graph. Then, the neighborhood radius eps under different minimum number of samples min_samples is obtained; where the minimum number of samples min_samples is a positive integer. Step 402: Set the neighborhood radius eps to 0.5 to 1 and the minimum number of samples min_samples to 3 to 5. Increase the neighborhood radius eps by 0.1 to obtain different neighborhood radii eps and minimum number of samples min_samples. Step 403: Combine the different neighborhood radii eps and minimum sample number min_samples from steps 401 and 402 into a parameter (eps, min_samples); Step 404: Under different combinations of (eps,min_samples) parameters, the DBSCAN algorithm is used to cluster different data points in two-dimensional spatial coordinates to obtain the microseismic profile coefficients under each combination of (eps,min_samples) parameters. Step 405: Arrange the microseismic profile coefficients of each (eps,min_samples) parameter combination from smallest to largest. Then, the (eps,min_samples) parameter combination corresponding to the maximum value of the microseismic profile coefficient is used as the optimized combination of neighborhood radius and minimum number of samples.

[0012] The above-mentioned intelligent mainshock group identification method considering microseismic profile coefficients and spectral clustering further includes step 404, which is specifically processed as follows: Step 4041: Under each set of (eps, min_samples) parameter combinations, the DBSCAN algorithm is used to cluster the data points in the two-dimensional spatial coordinates to obtain multiple clusters; each cluster contains multiple microseismic event samples. Step 4042: Set the m-th cluster to contain K microseismic event samples, according to... The average distance between the k-th microseismic event sample in the m-th cluster and other microseismic event samples in the m-th cluster is obtained; where, This represents the coordinates of the data point of the k-th microseismic event sample in the m-th cluster; Let represent the coordinates of the data point of the e-th microseismic event sample in the m-th cluster; 1≤k≤K, 1≤e≤K, and k≠e; Step 4043: Obtain the distance between the kth microseismic event sample in the mth cluster and each microseismic event sample in other clusters, and record the cluster corresponding to the minimum distance as the nearest neighbor cluster of the kth microseismic event sample in the mth cluster. Step 4044: Following the method in step 4042, obtain the average distance between the k-th microseismic event sample in the m-th cluster and all microseismic event samples in its nearest neighboring cluster. ; Step 4045, according to The profile coefficients of the k-th microseismic event sample in the m-th cluster are obtained. ; Step 4046: Repeat steps 4042 and 4045 multiple times to obtain the profile coefficients of all microseismic event samples in the m-th cluster. Step 4047: Repeat steps 4042 and 4046 multiple times to obtain the profile coefficients of all microseismic event samples in all clusters, and record the average of the profile coefficients of all microseismic event samples in all clusters as the microseismic profile coefficient under each (eps,min_samples) parameter combination.

[0013] The above-mentioned intelligent mainshock group identification method considering microseismic profile coefficients and spectral clustering further includes step six, the specific process of which is as follows: Step 601: Optimize the cluster for the Cth microseismic event in Step 5, based on... The spatial density of the optimized cluster for the Cth microseismic event is obtained. Where F represents the total number of microseismic event samples in the optimized cluster of the Cth microseismic event, f is a positive integer, and 1 ≤ f ≤ F. This represents the spatial normalized coordinates of the sample of the f-th microseismic event in the optimized cluster of the C-th microseismic event. Let represent the mean of the spatially normalized coordinates of all microseismic event samples in the optimized cluster of the Cth microseismic event; Step 602, according to The energy characteristics of the optimized cluster of the Cth microseismic event are obtained. ;in, This represents the energy value corresponding to the sample of the f-th microseismic event in the optimized cluster of the C-th microseismic event; Step 603, according to The temporal characteristics of the optimized cluster of the Cth microseismic event are obtained. ;in, This represents the latest occurrence time of the microseismic event sample in the optimized cluster of the Cth microseismic event. This represents the earliest occurrence time of the microseismic event sample in the optimized cluster of the Cth microseismic event; Step 604: Repeat steps 601 to 603 multiple times to obtain the spatial density, energy characteristics, and temporal characteristics of each microseismic event optimized cluster, and obtain the maximum spatial density. Minimum spatial density Maximum energy characteristics Minimum energy characteristic Maximum time characteristics and minimum energy characteristics ; Step 605, according to The comprehensive feature score of the optimized cluster of the Cth microseismic event is obtained. ; Step 606: Repeat step 605 multiple times to obtain the comprehensive feature score of each microseismic event optimized cluster, and select the microseismic event cluster with the highest comprehensive feature score as the main seismic group.

[0014] Compared with the prior art, the present invention has the following advantages: 1. The method of the present invention has simple steps and reasonable design, and solves the problem of misjudgment of mainshock groups caused by the lack of features and single indicators in the current traditional methods.

[0015] 2. This invention obtains spatial similarity matrices, energy similarity matrices, and temporal similarity matrices based on microseismic monitoring data and weights them to form a comprehensive similarity matrix. It considers not only spatial coordinates but also energy parameters and time series characteristics. By fusing the spatial, energy, and temporal characteristics of microseismic events and standardizing the spatial density, energy, and temporal characteristics, it balances the influence weights of multi-dimensional features to suppress low-energy noise. Furthermore, by introducing weight parameters to dynamically adjust the contribution of each feature, it improves the accuracy of subsequent identification.

[0016] 3. This invention constructs a Laplace matrix based on a comprehensive similarity matrix, and through Laplace matrix eigenvalue decomposition and PCA dimensionality reduction, maps microseismic events to two-dimensional spatial coordinates that retain nonlinear correlations, which facilitates subsequent clustering.

[0017] 4. When using the DBSCAN algorithm to cluster data points in two-dimensional spatial coordinates, this invention uses the contour coefficients under different combinations of neighborhood radius and minimum number of samples to select the optimal combination of neighborhood radius and minimum number of samples corresponding to the contour coefficients. Dynamic parameter optimization is achieved by selecting the optimal contour coefficients.

[0018] 5. Under the optimized combination of neighborhood radius and minimum sample number, this invention extracts spatial density, energy characteristics and temporal characteristics of each microseismic event into an optimized cluster and calculates a weighted comprehensive score. The optimized cluster of microseismic events with the highest score is selected as the main seismic group, which significantly improves the accuracy of main seismic group identification. The high-precision identification of main seismic groups in coal mining facilitates subsequent stress evolution path tracking.

[0019] In summary, the method of this invention is simple in steps and reasonable in design. In the spectral clustering, the spatial similarity matrix, energy similarity matrix and temporal similarity matrix are weighted to form a comprehensive similarity matrix. Considering the microseismic profile coefficient to obtain the neighborhood radius and the minimum number of samples, the optimized cluster of microseismic events is obtained. Furthermore, the mainshock group is selected by integrating evaluation indicators of spatial density, energy characteristics and temporal characteristics, effectively extracting the key microseismic event clusters that reflect the stress concentration area.

[0020] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0021] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0022] like Figure 1 As shown, the present invention provides an intelligent mainshock group identification method considering microseismic profile coefficients and spectral clustering, comprising the following steps: Step 1: Obtain microseismic monitoring data: Step 2: Obtain the spatial similarity matrix, energy similarity matrix, and temporal similarity matrix based on microseismic monitoring data, and then weight them to form a comprehensive similarity matrix; Step 3: Form a feature matrix based on the comprehensive similarity matrix and map the data points in the feature matrix to two-dimensional spatial coordinates: Step 4: Use the DBSCAN algorithm to cluster data points in two-dimensional spatial coordinates, and optimize the combination of neighborhood radius and minimum number of samples based on the microseismic profile coefficient; Step 5: Under the optimized combination of neighborhood radius and minimum sample number, the DBSCAN algorithm is used to cluster the data points in the two-dimensional spatial coordinates to obtain multiple optimized clusters of microseismic events; Step 6: Extract the spatial density, energy characteristics, and temporal characteristics of each microseismic event optimized cluster, and obtain a weighted comprehensive feature score. Select the microseismic event cluster with the highest comprehensive feature score as the mainshock group.

[0023] In this embodiment, in step one, a microseismic monitoring system is used to monitor the mine area to be monitored, and to obtain microseismic events per unit time and the occurrence time, spatial coordinates and energy value of each microseismic event per unit time. Step two, the specific process is as follows: Step 201: Normalize the spatial coordinates of each microseismic event using the proportional normalization method to obtain the spatial normalized coordinates corresponding to each microseismic event. Step 202: Record each microseismic event in chronological order of occurrence as the first microseismic event, ..., the i-th microseismic event, ..., the I-th microseismic event; where i is a positive integer and 1 ≤ i ≤ I; Step 203: Obtain the spatial similarity matrix based on the spatially normalized coordinates corresponding to I microseismic events. ; Step 204: Obtain the energy similarity matrix based on the energy values ​​corresponding to I microseismic events. ; Step 205: Obtain the time similarity matrix based on the times corresponding to I microseismic events. ; Step 206: Based on the spatial similarity matrix Energy similarity matrix and time similarity matrix The weighted averages are used to form a comprehensive similarity matrix.

[0024] In this embodiment, step 203 is specifically performed as follows: Step 2031: Using a computer according to The spatial distance between the i-th microseismic event and the j-th microseismic event is obtained. ;in, Represents the spatial normalized coordinates of the i-th microseismic event. Let represent the spatial normalized coordinates of the j-th microseismic event; j is a positive integer, and 1≤j≤1; Step 2032: Repeat step 2031 multiple times to obtain the spatial distance between any two microseismic events, and obtain the maximum spatial distance from all spatial distances between two microseismic events. ; Step 2033, according to The spatial similarity between the i-th microseismic event and the j-th microseismic event is obtained. ; Step 2034: Repeat step 2033 multiple times to obtain I×I spatial similarities, and use the I×I spatial similarities as matrix elements to obtain the spatial similarity matrix. .

[0025] In this embodiment, step 204 is specifically performed as follows: Step 2041: Using a computer according to... The energy difference between the i-th microseismic event and the j-th microseismic event is obtained. ;in, This represents the energy value of the i-th microseismic event. This represents the energy value of the j-th microseismic event; Step 2042: Repeat step 2041 multiple times to obtain the energy difference between any two microseismic events, and obtain the maximum energy difference from all energy differences between two microseismic events. ; Step 2043, according to The energy similarity between the i-th microseismic event and the j-th microseismic event is obtained. ; Step 2044: Repeat step 2043 multiple times to obtain I×I energy similarities, and use the I×I energy similarities as matrix elements to obtain the energy similarity matrix. .

[0026] In this embodiment, step 205 is specifically performed as follows: Step 2051: Using a computer according to... The time difference between the i-th microseismic event and the j-th microseismic event is obtained. ;in, This represents the time of occurrence of the i-th microseismic event. This indicates the time of occurrence of the j-th microseismic event; Step 2052: Repeat step 2051 multiple times to obtain the time difference between any two microseismic events, and obtain the maximum time difference from all the time differences between two microseismic events. ; Step 2053, according to The time similarity between the i-th microseismic event and the j-th microseismic event is obtained. ; Step 2054: Repeat step 2053 multiple times to obtain I×I time similarity scores, and use these I×I time similarity scores as matrix elements to obtain a time similarity matrix. ; Step 206, the specific process is as follows: according to The comprehensive similarity matrix is ​​obtained. ;in, Weights representing spatial similarity Weights representing energy similarity Weights representing temporal similarity. All are located between 0 and 1, and .

[0027] In this embodiment, step three is as follows: Step 301: Construct a Laplacian matrix based on the comprehensive similarity matrix, and perform eigenvalue decomposition on the Laplacian matrix to obtain eigenvalues ​​and corresponding eigenvectors; Step 302: Arrange the eigenvalues ​​in ascending order, remove eigenvalues ​​with zero values ​​and duplicate eigenvalues, and then use the eigenvectors corresponding to the remaining eigenvalues ​​as column vectors to form the original eigenma matrix; Step 303: Reduce the dimensionality of the original feature matrix using PCA to obtain the reduced matrix, and denote the reduced matrix as the feature matrix. Map the data points in the feature matrix to two-dimensional spatial coordinates.

[0028] In this embodiment, step four is as follows: Step 401: Set the minimum number of samples min_samples to a value of 3 to 5, and determine the neighborhood radius eps value based on the K-distance graph. Then, the neighborhood radius eps under different minimum number of samples min_samples is obtained; where the minimum number of samples min_samples is a positive integer. Step 402: Set the neighborhood radius eps to 0.5 to 1 and the minimum number of samples min_samples to 3 to 5. Increase the neighborhood radius eps by 0.1 to obtain different neighborhood radii eps and minimum number of samples min_samples. Step 403: Combine the different neighborhood radii eps and minimum sample number min_samples from steps 401 and 402 into a parameter (eps, min_samples); Step 404: Under different combinations of (eps,min_samples) parameters, the DBSCAN algorithm is used to cluster different data points in two-dimensional spatial coordinates to obtain the microseismic profile coefficients under each combination of (eps,min_samples) parameters. Step 405: Arrange the microseismic profile coefficients of each (eps,min_samples) parameter combination from smallest to largest. Then, the (eps,min_samples) parameter combination corresponding to the maximum value of the microseismic profile coefficient is used as the optimized combination of neighborhood radius and minimum number of samples.

[0029] In this embodiment, step 404 is specifically performed as follows: Step 4041: Under each set of (eps, min_samples) parameter combinations, the DBSCAN algorithm is used to cluster the data points in the two-dimensional spatial coordinates to obtain multiple clusters; each cluster contains multiple microseismic event samples. Step 4042: Set the m-th cluster to contain K microseismic event samples, according to... The average distance between the k-th microseismic event sample in the m-th cluster and other microseismic event samples in the m-th cluster is obtained; where, This represents the coordinates of the data point of the k-th microseismic event sample in the m-th cluster; Let represent the coordinates of the data point of the e-th microseismic event sample in the m-th cluster; 1≤k≤K, 1≤e≤K, and k≠e; Step 4043: Obtain the distance between the kth microseismic event sample in the mth cluster and each microseismic event sample in other clusters, and record the cluster corresponding to the minimum distance as the nearest neighbor cluster of the kth microseismic event sample in the mth cluster. Step 4044: Following the method in step 4042, obtain the average distance between the k-th microseismic event sample in the m-th cluster and all microseismic event samples in its nearest neighboring cluster. ; Step 4045, according to The profile coefficients of the k-th microseismic event sample in the m-th cluster are obtained. ; Step 4046: Repeat steps 4042 and 4045 multiple times to obtain the profile coefficients of all microseismic event samples in the m-th cluster. Step 4047: Repeat steps 4042 and 4046 multiple times to obtain the profile coefficients of all microseismic event samples in all clusters, and record the average of the profile coefficients of all microseismic event samples in all clusters as the microseismic profile coefficient under each (eps,min_samples) parameter combination.

[0030] In this embodiment, step six is ​​as follows: Step 601: Optimize the cluster for the Cth microseismic event in Step 5, based on... The spatial density of the optimized cluster for the Cth microseismic event is obtained. Where F represents the total number of microseismic event samples in the optimized cluster of the Cth microseismic event, f is a positive integer, and 1 ≤ f ≤ F. This represents the spatial normalized coordinates of the sample of the f-th microseismic event in the optimized cluster of the C-th microseismic event. Let represent the mean of the spatially normalized coordinates of all microseismic event samples in the optimized cluster of the Cth microseismic event; Step 602, according to The energy characteristics of the optimized cluster of the Cth microseismic event are obtained. ;in, This represents the energy value corresponding to the sample of the f-th microseismic event in the optimized cluster of the C-th microseismic event; Step 603, according to The temporal characteristics of the optimized cluster of the Cth microseismic event are obtained. ;in, This represents the latest occurrence time of the microseismic event sample in the optimized cluster of the Cth microseismic event. This represents the earliest occurrence time of the microseismic event sample in the optimized cluster of the Cth microseismic event; Step 604: Repeat steps 601 to 603 multiple times to obtain the spatial density, energy characteristics, and temporal characteristics of each microseismic event optimized cluster, and obtain the maximum spatial density. Minimum spatial density Maximum energy characteristics Minimum energy characteristic Maximum time characteristics and minimum energy characteristics ; Step 605, according to The comprehensive feature score of the optimized cluster of the Cth microseismic event is obtained. ; Step 606: Repeat step 605 multiple times to obtain the comprehensive feature score of each microseismic event optimized cluster, and select the microseismic event cluster with the highest comprehensive feature score as the main seismic group.

[0031] In this embodiment, the mine area to be monitored includes the working face and roadways; in actual use, the coordinate system of the spatial coordinates is not limited, as long as it meets the requirements of actual monitoring. The spatial coordinates are three-dimensional spatial coordinates.

[0032] In this embodiment, the unit of time is a day, which can be adjusted according to actual requirements.

[0033] In this embodiment, the weight of spatial similarity is specifically implemented. The weight of energy similarity is 0.3. The weight of time similarity is 0.4. The value is 0.3. It should be noted that the weights can be adjusted according to actual requirements to meet the analysis needs.

[0034] In this embodiment, the combined similarity matrix and Laplacian matrix are N×N in specific implementation, and the feature matrix in step 303 is N×2 in order to reduce the dimension to 2-dimensional, so as to facilitate the mapping of data points in the feature matrix to two-dimensional spatial coordinates.

[0035] In this embodiment, the DBSCAN algorithm (Density-Based Spatial Clustering of Applications with Noise) is used in actual applications. It is a density-based clustering algorithm.

[0036] In this embodiment, in actual use, when using the proportional normalization method in step 201, the maximum values ​​of the X-axis coordinate, Y-axis coordinate, and Z-axis coordinate are obtained respectively, and the maximum values ​​of the X-axis coordinate / X-axis coordinate, Y-axis coordinate / Y-axis coordinate, and Z-axis coordinate / Z-axis coordinate of the spatial coordinates of each microseismic event are normalized.

[0037] In this embodiment, in actual use, step 401 obtains the neighborhood radius eps with min_samples values ​​of 3, 4 and 5; step 402 obtains the neighborhood radius eps with values ​​of 0.5, 0.6, 0.7, 0.8, 0.9 and 1, and the combination of the minimum number of samples min_samples with values ​​of 3, 4 and 5.

[0038] In summary, the method of this invention is simple in steps and reasonable in design. In the spectral clustering, the spatial similarity matrix, energy similarity matrix and temporal similarity matrix are weighted to form a comprehensive similarity matrix. Considering the microseismic profile coefficient to obtain the neighborhood radius and the minimum number of samples, the optimized cluster of microseismic events is obtained. Furthermore, the mainshock group is selected by integrating evaluation indicators of spatial density, energy characteristics and temporal characteristics, effectively extracting the key microseismic event clusters that reflect the stress concentration area.

[0039] The above description is merely a preferred embodiment of the present invention and does not constitute any limitation on the present invention. Any simple modifications, alterations, or equivalent structural changes made to the above embodiments based on the technical essence of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A method for intelligent identification of mainshock groups considering microseismic profile coefficients and spectral clustering, characterized in that, The method includes the following steps: Step 1: Obtain microseismic monitoring data: Step 2: Obtain the spatial similarity matrix, energy similarity matrix, and temporal similarity matrix based on microseismic monitoring data, and then weight them to form a comprehensive similarity matrix; Step 3: Form a feature matrix based on the comprehensive similarity matrix and map the data points in the feature matrix to two-dimensional spatial coordinates: Step 4: Use the DBSCAN algorithm to cluster data points in two-dimensional spatial coordinates, and optimize the combination of neighborhood radius and minimum number of samples based on the microseismic profile coefficient; Step 5: Under the optimized combination of neighborhood radius and minimum sample number, the DBSCAN algorithm is used to cluster the data points in the two-dimensional spatial coordinates to obtain multiple optimized clusters of microseismic events; Step 6: Extract the spatial density, energy characteristics, and temporal characteristics of each microseismic event optimized cluster, and obtain a weighted comprehensive feature score. Select the microseismic event cluster with the highest comprehensive feature score as the mainshock group.

2. The intelligent mainshock group identification method considering microseismic profile coefficients and spectral clustering according to claim 1, characterized in that: In step one, a microseismic monitoring system is used to monitor the mine area to be monitored, and to obtain the microseismic events per unit time and the occurrence time, spatial coordinates and energy value of each microseismic event per unit time. Step two, the specific process is as follows: Step 201: Normalize the spatial coordinates of each microseismic event using the proportional normalization method to obtain the spatial normalized coordinates corresponding to each microseismic event. Step 202: Record each microseismic event in chronological order of occurrence as the first microseismic event, ..., the i-th microseismic event, ..., the I-th microseismic event; where i is a positive integer and 1 ≤ i ≤ I; Step 203: Obtain the spatial similarity matrix based on the spatially normalized coordinates corresponding to I microseismic events. ; Step 204: Obtain the energy similarity matrix based on the energy values ​​corresponding to I microseismic events. ; Step 205: Obtain the time similarity matrix based on the times corresponding to I microseismic events. ; Step 206: Based on the spatial similarity matrix Energy similarity matrix and time similarity matrix The weighted averages are used to form a comprehensive similarity matrix.

3. The intelligent mainshock group identification method considering microseismic profile coefficients and spectral clustering according to claim 2, characterized in that: Step 203, the specific process is as follows: Step 2031: Using a computer according to The spatial distance between the i-th microseismic event and the j-th microseismic event is obtained. ;in, Represents the spatial normalized coordinates of the i-th microseismic event. Let represent the spatial normalized coordinates of the j-th microseismic event; j is a positive integer, and 1≤j≤1; Step 2032: Repeat step 2031 multiple times to obtain the spatial distance between any two microseismic events, and obtain the maximum spatial distance from all spatial distances between two microseismic events. ; Step 2033, according to The spatial similarity between the i-th microseismic event and the j-th microseismic event is obtained. ; Step 2034: Repeat step 2033 multiple times to obtain I×I spatial similarities, and use the I×I spatial similarities as matrix elements to obtain the spatial similarity matrix. .

4. The intelligent mainshock group identification method considering microseismic profile coefficients and spectral clustering according to claim 2, characterized in that: Step 204, the specific process is as follows: Step 2041: Using a computer according to... The energy difference between the i-th microseismic event and the j-th microseismic event is obtained. ;in, This represents the energy value of the i-th microseismic event. This represents the energy value of the j-th microseismic event; Step 2042: Repeat step 2041 multiple times to obtain the energy difference between any two microseismic events, and obtain the maximum energy difference from all energy differences between two microseismic events. ; Step 2043, according to The energy similarity between the i-th microseismic event and the j-th microseismic event is obtained. ; Step 2044: Repeat step 2043 multiple times to obtain I×I energy similarities, and use these I×I energy similarities as matrix elements to obtain the energy similarity matrix. .

5. The intelligent mainshock group identification method considering microseismic profile coefficients and spectral clustering according to claim 2, characterized in that: Step 205, the specific process is as follows: Step 2051: Using a computer according to... The time difference between the i-th microseismic event and the j-th microseismic event is obtained. ;in, This represents the time of occurrence of the i-th microseismic event. This indicates the time of occurrence of the j-th microseismic event; Step 2052: Repeat step 2051 multiple times to obtain the time difference between any two microseismic events, and obtain the maximum time difference from all the time differences between two microseismic events. ; Step 2053, according to The time similarity between the i-th microseismic event and the j-th microseismic event is obtained. ; Step 2054: Repeat step 2053 multiple times to obtain I×I time similarity scores, and use these I×I time similarity scores as matrix elements to obtain a time similarity matrix. ; Step 206, the specific process is as follows: according to The comprehensive similarity matrix is ​​obtained. ;in, Weights representing spatial similarity Weights representing energy similarity Weights representing temporal similarity. All are located between 0 and 1, and .

6. The intelligent mainshock group identification method considering microseismic profile coefficients and spectral clustering according to claim 1, characterized in that: Step three, the specific process is as follows: Step 301: Construct a Laplacian matrix based on the comprehensive similarity matrix, and perform eigenvalue decomposition on the Laplacian matrix to obtain eigenvalues ​​and corresponding eigenvectors; Step 302: Arrange the eigenvalues ​​in ascending order, remove eigenvalues ​​with zero values ​​and duplicate eigenvalues, and then use the eigenvectors corresponding to the remaining eigenvalues ​​as column vectors to form the original eigenma matrix; Step 303: Reduce the dimensionality of the original feature matrix using PCA to obtain the reduced matrix, and denote the reduced matrix as the feature matrix. Map the data points in the feature matrix to two-dimensional spatial coordinates.

7. The intelligent mainshock group identification method considering microseismic profile coefficients and spectral clustering according to claim 1, characterized in that: Step four, the specific process is as follows: Step 401: Set the minimum number of samples min_samples to a value of 3 to 5, and determine the neighborhood radius eps value based on the K-distance graph. Then, the neighborhood radius eps under different minimum number of samples min_samples is obtained; where the minimum number of samples min_samples is a positive integer. Step 402: Set the neighborhood radius eps to 0.5 to 1 and the minimum number of samples min_samples to 3 to 5. Increase the neighborhood radius eps by 0.1 to obtain different neighborhood radii eps and minimum number of samples min_samples. Step 403: Combine the different neighborhood radii eps and minimum sample number min_samples from steps 401 and 402 into a parameter (eps, min_samples); Step 404: Under different combinations of (eps,min_samples) parameters, the DBSCAN algorithm is used to cluster different data points in two-dimensional spatial coordinates to obtain the microseismic profile coefficients under each combination of (eps,min_samples) parameters. Step 405: Arrange the microseismic profile coefficients of each (eps,min_samples) parameter combination from smallest to largest. Then, the (eps,min_samples) parameter combination corresponding to the maximum value of the microseismic profile coefficient is used as the optimized combination of neighborhood radius and minimum number of samples.

8. The intelligent mainshock group identification method considering microseismic profile coefficients and spectral clustering according to claim 7, characterized in that: Step 404, the specific process is as follows: Step 4041: Under each set of (eps, min_samples) parameter combinations, the DBSCAN algorithm is used to cluster the data points in the two-dimensional spatial coordinates to obtain multiple clusters; each cluster contains multiple microseismic event samples. Step 4042: Set the m-th cluster to contain K microseismic event samples, according to... The average distance between the k-th microseismic event sample in the m-th cluster and other microseismic event samples in the m-th cluster is obtained; where, This represents the coordinates of the data point of the k-th microseismic event sample in the m-th cluster; Let represent the coordinates of the data point of the e-th microseismic event sample in the m-th cluster; 1≤k≤K, 1≤e≤K, and k≠e; Step 4043: Obtain the distance between the kth microseismic event sample in the mth cluster and each microseismic event sample in other clusters, and record the cluster corresponding to the minimum distance as the nearest neighbor cluster of the kth microseismic event sample in the mth cluster. Step 4044: Following the method in step 4042, obtain the average distance between the k-th microseismic event sample in the m-th cluster and all microseismic event samples in its nearest neighboring cluster. ; Step 4045, according to The profile coefficients of the k-th microseismic event sample in the m-th cluster are obtained. ; Step 4046: Repeat steps 4042 and 4045 multiple times to obtain the profile coefficients of all microseismic event samples in the m-th cluster. Step 4047: Repeat steps 4042 and 4046 multiple times to obtain the profile coefficients of all microseismic event samples in all clusters, and record the average of the profile coefficients of all microseismic event samples in all clusters as the microseismic profile coefficient under each (eps,min_samples) parameter combination.

9. The intelligent mainshock group identification method considering microseismic profile coefficients and spectral clustering according to claim 1, characterized in that: Step six, the specific process is as follows: Step 601: Optimize the cluster for the Cth microseismic event in Step 5, based on... The spatial density of the optimized cluster for the Cth microseismic event is obtained. Where F represents the total number of microseismic event samples in the optimized cluster of the Cth microseismic event, f is a positive integer, and 1 ≤ f ≤ F. This represents the spatial normalized coordinates of the sample of the f-th microseismic event in the optimized cluster of the C-th microseismic event. Let represent the mean of the spatially normalized coordinates of all microseismic event samples in the optimized cluster of the Cth microseismic event; Step 602, according to The energy characteristics of the optimized cluster of the Cth microseismic event are obtained. ;in, This represents the energy value corresponding to the sample of the f-th microseismic event in the optimized cluster of the C-th microseismic event; Step 603, according to The temporal characteristics of the optimized cluster of the Cth microseismic event are obtained. ;in, This represents the latest occurrence time of the microseismic event sample in the optimized cluster of the Cth microseismic event. This represents the earliest occurrence time of the microseismic event sample in the optimized cluster of the Cth microseismic event; Step 604: Repeat steps 601 to 603 multiple times to obtain the spatial density, energy characteristics, and temporal characteristics of each microseismic event optimized cluster, and obtain the maximum spatial density. Minimum spatial density Maximum energy characteristics Minimum energy characteristic Maximum time characteristics and minimum energy characteristics ; Step 605, according to The comprehensive feature score of the optimized cluster of the Cth microseismic event is obtained. ; Step 606: Repeat step 605 multiple times to obtain the comprehensive feature score of each microseismic event optimized cluster, and select the microseismic event cluster with the highest comprehensive feature score as the main seismic group.

Images