Mining-induced stress evolution analysis method fusing optical flow tracking multiple indexes

Abstract

The invention discloses a mining-induced stress evolution analysis method fusing optical flow tracking multiple indexes. The mining-induced stress evolution analysis method comprises the following steps: 1, acquiring micro-seismic monitoring data; 2, generating a coordinate KDE graph based on the normalized space coordinates of the micro-seismic monitoring data by using a two-dimensional Gaussian kernel density estimation method; 3, filtering and graying the coordinate KDE image to obtain a space coordinate grey-scale image; 4, based on two adjacent space coordinate grey-scale maps, obtaining an optical flow vector of each pixel point of the space coordinate grey-scale maps by using a Farneback optical flow algorithm to obtain an optical flow vector field map; 5, performing reverse normalization on the coordinates of the optical flow vector field graph to obtain a space coordinate optical flow vector field graph; and 6, obtaining a path reflecting mining-induced stress evolution based on the space coordinate optical flow vector field graphs of different unit time. Based on a coordinate KDE graph, a Farneback optical flow algorithm is adopted to obtain an optical flow vector field graph, and the optical flow vector field graph is quantized to obtain a path reflecting mining-induced stress evolution.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of mining stress evolution, and particularly relates to a mining stress evolution analysis method fusing light flow tracking and multiple indexes. BACKGROUND

[0002] Microseismic is produced when rock mass or coal rock mass breaks, dislocates and releases energy due to stress concentration exceeding the limit of rock mass in mining activities. Microseismic monitoring technology has become an important means to study mining stress by virtue of high sensitivity to rock mass rupture events. However, the existing technical solutions usually rely on a single analysis model, and most of them ignore the dynamic evolution process of stress at different scales. Therefore, the existing methods generally lack deep understanding of stress dissipation mechanism, especially cannot quantify the complex characteristics in the stress evolution process.

[0003] On the other hand, the traditional method has shortcomings in visualization and quantitative characterization, which cannot intuitively show the stress propagation path, and lacks precise quantitative description of stress concentration, dissipation direction and intensity, resulting in insufficient disaster warning accuracy, and it is difficult to meet the actual needs of mine risk prevention and control. In addition, the existing technology also generally has the problem of dynamic analysis lag, which is difficult to capture the rapid change trend of the stress field in the mining process in real time, and cannot meet the requirements of mine dynamic monitoring and disaster warning.

[0004] Therefore, at present, there is a lack of a mining stress evolution analysis method fusing light flow tracking and multiple indexes which is reasonable in design. After generating the coordinate KDE graph based on the normalized spatial coordinates of microseismic monitoring data, the light flow vector field graph is obtained by using the Farneback light flow algorithm to quantitatively obtain the stress evolution path, realize dynamic tracking of the stress path, and intuitively show the stress dissipation path in the mining process by combining the streamline graph and the heat map, which can capture the rapid change of the stress field in real time and greatly improve the timeliness of the warning. SUMMARY

[0005] The technical problem to be solved by the present application is to provide a mining stress evolution analysis method fusing light flow tracking and multiple indexes, which is simple in method steps and reasonable in design, can quantitatively obtain the stress evolution path by generating the coordinate KDE graph based on the normalized spatial coordinates of microseismic monitoring data and using the Farneback light flow algorithm to obtain the light flow vector field graph, realize dynamic tracking of the stress path, intuitively show the stress dissipation path in the mining process by combining the streamline graph and the heat map, can capture the rapid change of the stress field in real time, and greatly improve the timeliness of the warning.

[0006] To solve the above technical problems, the technical scheme adopted by the present application is as follows: a mining stress evolution analysis method fusing light flow tracking and multiple indexes, characterized in that the method comprises the following steps: Step one, obtaining microseismic monitoring data: The microseismic monitoring system is used to monitor the area of the mine to be monitored, and the microseismic events in a unit of time and the occurrence time, spatial coordinates and energy value corresponding to each microseismic event in a unit of time are obtained; and the spatial coordinates are normalized to obtain normalized spatial coordinates; Step two, generating a coordinate KDE map based on the normalized spatial coordinates of the microseismic monitoring data using a two-dimensional Gaussian kernel density estimation method: Step three, filtering and grayscale processing of the coordinate KDE map to obtain a spatial coordinate grayscale map: Step four, based on two adjacent spatial coordinate grayscale maps, the Farneback optical flow algorithm is used to obtain the optical flow vector of each pixel point of the spatial coordinate grayscale map, and a flow vector field map is obtained; Step five, the coordinates of the optical flow vector field map are de-normalized to obtain a spatial coordinate optical flow vector field map; Step six, based on the spatial coordinate optical flow vector field maps of different units of time, a path reflecting the mining stress evolution is obtained.

[0007] The above-mentioned mining stress evolution analysis method fusing optical flow tracking and multiple indexes further adopts proportional normalization to normalize the spatial coordinates of each microseismic event to obtain the normalized spatial coordinates corresponding to each microseismic event; wherein the normalized spatial coordinates include x-axis normalized coordinates, y-axis normalized coordinates and z-axis normalized coordinates; Step two, the specific process is as follows: Step 201, obtaining the maximum energy value Emax from the energy values corresponding to each microseismic event; Step 202, sequentially recording each microseismic event as the first microseismic event, the second microseismic event, the i-th microseismic event, the n-th microseismic event according to the occurrence time; wherein i is a positive integer, and 1≤i≤n; Step 203, recording the energy value of the i-th microseismic event as , and obtaining the weight Wi of the i-th microseismic event according to Wi= / Emax; Step 204, inputting the weights of n microseismic events into a computer, based on the x-axis normalized coordinates and y-axis normalized coordinates of n microseismic events, using a two-dimensional Gaussian kernel density estimation method to obtain an xy coordinate KDE map; Step 205, according to the method of step 204, inputting the weights of n microseismic events into a computer, based on the x-axis normalized coordinates and z-axis normalized coordinates of n microseismic events, using a two-dimensional Gaussian kernel density estimation method to obtain an xz coordinate KDE map; Step 206, according to the method of step 204, using a computer to input the weight of n microseismic events, based on the y-axis normalized coordinates and z-axis normalized coordinates of n microseismic events, using a two-dimensional Gaussian kernel density estimation method, to obtain a yz coordinate KDE graph.

[0008] The above-mentioned mining stress evolution analysis method of fusing light flow tracking multi-index, further, step 204, the specific process is as follows: Step 2041, based on the weight of n microseismic events, obtaining The estimated probability density function at , and ; wherein, is the bandwidth, is the kernel function, and the kernel function is a Gaussian function; represents any x-axis and y-axis coordinates, xi represents the x-axis normalized coordinates of the i-th microseismic event, and yi represents the y-axis normalized coordinates of the i-th microseismic event. Step 2042, set the value range of x-axis as 0-1, and the value range of y-axis as 0-1, and divide the entire region of x-axis coordinates and y-axis coordinates into a two-dimensional grid; Step 2043, obtain the coordinates of each grid , and input into the probability density function to obtain the probability density of each grid; Step 2044, map the probability density of each grid to a color heat map to obtain an xy coordinate KDE graph.

[0009] The above-mentioned mining stress evolution analysis method of fusing light flow tracking multi-index, further, step three, the specific process is as follows: Step 301, using a computer to perform Gaussian filtering on the xy coordinate KDE graph, the xz coordinate KDE graph and the yz coordinate KDE graph respectively, to obtain a filtered xy coordinate KDE graph, a filtered xz coordinate KDE graph and a filtered yz coordinate KDE graph; Step 302, using a computer to perform grayscale processing on the filtered xy coordinate KDE graph, the filtered xz coordinate KDE graph and the filtered yz coordinate KDE graph respectively, to obtain an xy coordinate grayscale graph, an xz coordinate grayscale graph and a yz coordinate grayscale graph.

[0010] The above-mentioned mining stress evolution analysis method of fusing light flow tracking multi-index, further, step four, the specific process is as follows: Step 401, according to the method of step two and step three, the microseismic monitoring data in the next unit time is processed, the xy coordinate gray scale map of the next unit time is obtained, and the xy coordinate gray scale map of the current unit time and the xy coordinate gray scale map of the next unit time are recorded as adjacent two xy coordinate gray scale maps; Step 402, using a computer, the Farneback optical flow algorithm is used to process the adjacent two xy coordinate gray scale maps, and the optical flow vector of each pixel point on the xy coordinate gray scale map of the current unit time is obtained; Step 403, using a computer, based on the optical flow vector of each pixel point, the optical flow vector arrow is drawn on the xy coordinate gray scale map of the current unit time, and the xy normalized coordinate optical flow vector field map is obtained; Step 404, according to the method of steps 401 to 403, the xz normalized coordinate optical flow vector field map is obtained; Step 405, according to the method of steps 401 to 403, the yz normalized coordinate optical flow vector field map is obtained.

[0011] The above-mentioned fusion optical flow tracking multi-index mining stress evolution analysis method, further, step five, the specific process is as follows: Step 501, the horizontal coordinate and the vertical coordinate in the xy normalized coordinate optical flow vector field map are de-normalized, and the xy space coordinate optical flow vector field map is obtained; Step 502, the horizontal coordinate and the vertical coordinate in the xz normalized coordinate optical flow vector field map are de-normalized, and the xz space coordinate optical flow vector field map is obtained; Step 503, the horizontal coordinate and the vertical coordinate in the yz normalized coordinate optical flow vector field map are de-normalized, and the yz space coordinate optical flow vector field map is obtained.

[0012] The above-mentioned fusion optical flow tracking multi-index mining stress evolution analysis method, further, step six, the specific process is as follows: Step 601, the optical flow vector of the jth pixel point on the xy space coordinate optical flow vector field map of the current unit time is recorded as (Vx(j), Vy(j)); wherein, Vx(j) represents the velocity component of the jth pixel point in the x-axis direction, and Vy(j) represents the velocity component of the jth pixel point in the y-axis direction; Step 602, the optical flow vectors of the jth pixel point on the xy space coordinate optical flow vector field maps of different unit times are accumulated, and the stress evolution path of the jth pixel point on the xy space coordinate optical flow vector field map is obtained; Step 603, steps 601 to 602 are repeated multiple times, and the stress evolution paths of each pixel point on the xy space coordinate optical flow vector field map are obtained; Step 604, according to , the flow velocity amplitude of the jth pixel point is obtained ; according to , the optical flow intensity corresponding to the xy spatial coordinate optical flow vector field diagram of the current unit time is obtained ; wherein, 1≤j≤J, J represents the total number of pixel points; The median probability density is selected from the probability density corresponding to the xy spatial coordinate optical flow vector field diagram of the current unit time; The pixel points greater than the optical flow intensity and the probability density greater than the median probability density on the xy spatial coordinate optical flow vector field diagram of the current unit time are recorded as the stress concentration area of the current unit time; Step 605, the stress evolution path of the stress concentration area is filtered out from the stress evolution paths of each pixel point on the xy spatial coordinate optical flow vector field diagrams of different unit times, and is recorded as the stress evolution path of the xy spatial stress concentration area; Step 606, the stress evolution path of the xz spatial stress concentration area is obtained according to the method of steps 601 to 605; Step 607, the stress evolution path of the yz spatial stress concentration area is obtained according to the method of steps 601 to 605.

[0013] The above-mentioned mining stress evolution analysis method fusing optical flow tracking and multiple indexes further includes the following steps after step six: Step 701, according to , the vector angle of the jth pixel point is obtained; and according to , the optical flow vector concentration is obtained; Step 702, according to , the total energy data of the microseismic event corresponding to the xy spatial coordinate optical flow vector field diagram of the current unit time is obtained ; according to , the energy mean value corresponding to the xy spatial coordinate optical flow vector field diagram of the current unit time is obtained ; Step 703, according to , the energy dissipation rate corresponding to the adjacent two xy spatial coordinate optical flow vector field diagrams is obtained ; wherein, represents the total energy data of the microseismic event corresponding to the xy spatial coordinate optical flow vector field diagram of the next unit time, represents the xy spatial coordinate optical flow vector field diagram of the current unit time, represents the unit time; Step 704, according to , the information entropy corresponding to the xy spatial coordinate optical flow vector field diagram of the current unit time is obtained ; wherein, represents the probability density of the jth pixel point coordinate in the xy coordinate KDE corresponding to the current unit time; Step 705, the minimum total cost of the probability distribution of each pixel point of the xy space coordinate optical flow vector field graph corresponding to the current unit time to the probability distribution of the xy space coordinate optical flow vector field graph corresponding to the next unit time is taken as the first order Wasserstein distance; Step 706, repeat steps 701 to step 705 multiple times to obtain the related parameters of the xz space coordinate optical flow vector field graph; Step 707, repeat steps 701 to step 705 multiple times to obtain the related parameters of the yz space coordinate optical flow vector field graph.

[0014] Compared with the prior art, the present application has the following advantages: 1. The method of the present application has simple steps and reasonable design, and solves the problem of dynamic tracking and comprehensive quantification of stress dissipation process based on mining stress evolution analysis.

[0015] 2. The present application generates coordinate KDE graph based on normalized space coordinates of microseismic monitoring data using two-dimensional Gaussian kernel density estimation method, and the coordinate KDE graph includes xy coordinate KDE graph, xz coordinate KDE graph and yz coordinate KDE graph, so as to realize mining stress evolution path analysis of different two-dimensional dimensions to reflect the migration direction of microseismic in different directions.

[0016] 3. The present application obtains the optical flow vector of each pixel point of the space coordinate gray scale graph based on the adjacent two space coordinate gray scale graphs using Farneback optical flow algorithm, obtains the optical flow vector field graph, and obtains the stress evolution path through the accumulation of the optical flow vector on the optical flow vector field graph of different unit time, realizes dynamic monitoring of stress field, and is convenient for subsequent real-time capture of stress change through multiple indexes, and provides scientific basis for mine pressure disaster warning.

[0017] 4. The present application can identify the stress evolution path in the mining process in real time, provide strong technical support for mine safety production, and through the obtained stress evolution path, it is convenient for subsequent mine enterprises to optimize the design of mining scheme, reasonably adjust the supporting parameters and working face advancing speed, so as to effectively prevent the occurrence of dynamic disasters such as rock burst, and significantly reduce the risk of safety accidents.

[0018] ​In summary, the method of the present application has simple steps and reasonable design, and after generating the coordinate KDE graph based on the normalized spatial coordinates of microseismic monitoring data, the Farneback optical flow algorithm is used to obtain the optical flow vector field graph to quantitatively obtain the path reflecting the mining stress evolution, realize the dynamic tracking of the stress path, and combine the streamline graph and the heat map to intuitively display the stress dissipation path in the mining process, which can capture the rapid changes of the stress field in real time and greatly improve the timeliness of the early warning.

[0019] The technical solutions of the present application will be further described in detail below with reference to the drawings and examples. BRIEF DESCRIPTION OF DRAWINGS

[0020] Figure 1 The method flow chart of the present application is shown. DETAILED DESCRIPTION

[0021] As shown in Figure 1 the present application fuses the mining stress evolution analysis method of optical flow tracking multi-index, which includes the following steps: Step one, obtaining microseismic monitoring data: The microseismic monitoring system is used to monitor the area to be monitored in the mine to obtain the microseismic events in unit time and the occurrence time, spatial coordinates and energy value corresponding to each microseismic event in unit time; and the spatial coordinates are normalized to obtain normalized spatial coordinates; Step two, generating the coordinate KDE graph based on the normalized spatial coordinates of microseismic monitoring data by using the two-dimensional Gaussian kernel density estimation method: Step three, filtering and grayscale processing of the coordinate KDE graph to obtain the spatial coordinate grayscale graph: Step four, based on the adjacent two spatial coordinate grayscale graphs, the optical flow vector of each pixel point of the spatial coordinate grayscale graph is obtained by using the Farneback optical flow algorithm to obtain the optical flow vector field graph; Step five, the coordinates of the optical flow vector field graph are de-normalized to obtain the spatial coordinate optical flow vector field graph; Step six, based on the spatial coordinate optical flow vector field graphs of different unit times, the path reflecting the mining stress evolution is obtained.

[0022] In this embodiment, the spatial coordinates of each microseismic event are normalized by using proportional normalization to obtain the normalized spatial coordinates corresponding to each microseismic event; wherein the normalized spatial coordinates include x-axis normalized coordinates, y-axis normalized coordinates and z-axis normalized coordinates; Step two, the specific process is as follows: Step 201, obtaining the maximum energy value Emax from the energy values corresponding to each microseismic event; Step 202, record each microseismic event in chronological order as the first microseismic event, the second microseismic event, the i-th microseismic event, the (n-1)th microseismic event and the nth microseismic event; wherein i is a positive integer, and 1≤i≤n; Step 203, record the energy value of the i-th microseismic event as , and obtain the weight Wi of the i-th microseismic event according to Wi= / Emax. Step 204, input the weights of the n microseismic events into the computer, and obtain the xy coordinate KDE map by using the two-dimensional Gaussian kernel density estimation method based on the x-axis normalized coordinates and the y-axis normalized coordinates of the n microseismic events. Step 205, input the weights of the n microseismic events into the computer according to the method of step 204, and obtain the xz coordinate KDE map by using the two-dimensional Gaussian kernel density estimation method based on the x-axis normalized coordinates and the z-axis normalized coordinates of the n microseismic events. Step 206, input the weights of the n microseismic events into the computer according to the method of step 204, and obtain the yz coordinate KDE map by using the two-dimensional Gaussian kernel density estimation method based on the y-axis normalized coordinates and the z-axis normalized coordinates of the n microseismic events.

[0023] In this embodiment, step 204 has the following specific process: Step 2041, obtain the probability density function estimated at the point (x, y) based on the weights of the n microseismic events. , and ; wherein, is the bandwidth, is the kernel function, and the kernel function is a Gaussian function; represents any x-axis and y-axis coordinates, xi represents the x-axis normalized coordinate of the i-th microseismic event, and yi represents the y-axis normalized coordinate of the i-th microseismic event. Step 2042, set the value range of the x-axis as 0-1, and the value range of the y-axis as 0-1, and divide the entire region of the x-axis coordinates and the y-axis coordinates into a two-dimensional grid. Step 2043, obtain the coordinates of each grid , and input into the probability density function to obtain the probability density of each grid. Step 2044, map the probability density of each grid to a color heat map to obtain the xy coordinate KDE map.

[0024] In this embodiment, step three has the following specific process: ​Step 301, Gaussian filtering is performed on the xy coordinate KDE graph, the xz coordinate KDE graph and the yz coordinate KDE graph respectively by using a computer to obtain a filtered xy coordinate KDE graph, a filtered xz coordinate KDE graph and a filtered yz coordinate KDE graph. Step 302, gray processing is performed on the filtered xy coordinate KDE graph, the filtered xz coordinate KDE graph and the filtered yz coordinate KDE graph respectively by using a computer to obtain an xy coordinate gray graph, an xz coordinate gray graph and a yz coordinate gray graph.

[0025] In this embodiment, step four has the following specific process: Step 401, the microseismic monitoring data in the next unit time is processed according to the method of step two and step three to obtain an xy coordinate gray graph of the next unit time, and the xy coordinate gray graph of the current unit time and the xy coordinate gray graph of the next unit time are recorded as two adjacent xy coordinate gray graphs; Step 402, the Farneback optical flow algorithm is used to process the two adjacent xy coordinate gray graphs by using a computer to obtain the optical flow vector of each pixel point on the xy coordinate gray graph of the current unit time; Step 403, the optical flow vector of each pixel point is used to draw an optical flow vector arrow on the xy coordinate gray graph of the current unit time by using a computer to obtain an xy normalized coordinate optical flow vector field graph; Step 404, the xz normalized coordinate optical flow vector field graph is obtained according to the method of step 401 to step 403; Step 405, the yz normalized coordinate optical flow vector field graph is obtained according to the method of step 401 to step 403.

[0026] In this embodiment, step five has the following specific process: Step 501, the horizontal coordinate and the vertical coordinate in the xy normalized coordinate optical flow vector field graph are de-normalized to obtain an xy space coordinate optical flow vector field graph; Step 502, the horizontal coordinate and the vertical coordinate in the xz normalized coordinate optical flow vector field graph are de-normalized to obtain an xz space coordinate optical flow vector field graph; Step 503, the horizontal coordinate and the vertical coordinate in the yz normalized coordinate optical flow vector field graph are de-normalized to obtain a yz space coordinate optical flow vector field graph.

[0027] In this embodiment, step six has the following specific process: Step 601, record the optical flow vector of the jth pixel point on the xy space coordinate optical flow vector field map of the current unit time as (Vx(j), Vy(j)); wherein, Vx(j) represents the velocity component of the jth pixel point in the x-axis direction, and Vy(j) represents the velocity component of the jth pixel point in the y-axis direction; Step 602, accumulate the optical flow vectors of the jth pixel point on the xy space coordinate optical flow vector field maps of different unit times to obtain the stress evolution path of the jth pixel point on the xy space coordinate optical flow vector field map; Step 603, repeatedly perform steps 601 to 602 multiple times to obtain the stress evolution paths of each pixel point on the xy space coordinate optical flow vector field map; Step 604, according to , obtain the flow velocity amplitude of the jth pixel point ; according to , obtain the optical flow intensity corresponding to the xy space coordinate optical flow vector field map of the current unit time ; wherein, 1≤j≤J, and J represents the total number of pixel points; select the median probability density from the probability density corresponding to the xy space coordinate optical flow vector field map of the current unit time; record the pixel points on the xy space coordinate optical flow vector field map of the current unit time which are greater than the optical flow intensity and have a probability density greater than the median probability density as the stress concentration region of the current unit time; Step 605, filter out the stress evolution paths of the stress concentration region from the stress evolution paths of each pixel point on the xy space coordinate optical flow vector field maps of different unit times to obtain the stress evolution paths of the xy space stress concentration region; Step 606, obtain the stress evolution paths of the xz space stress concentration region according to the method of steps 601 to 605; Step 607, obtain the stress evolution paths of the yz space stress concentration region according to the method of steps 601 to 605.

[0028] In this embodiment, after step six, multi-index parameter acquisition is further performed, which is specifically as follows: Step 701, according to , obtain the vector angle of the jth pixel point ; and according to ; obtain the optical flow vector concentration ; Step 702, according to , obtain the total energy data of the microseismic event corresponding to the xy space coordinate optical flow vector field map of the current unit time ; according to This yields the average energy value corresponding to the optical flow vector field map of the xy spatial coordinates at the current unit time. ; Step 703, according to The energy dissipation rate corresponding to the optical flow vector field diagrams of two adjacent xy spatial coordinates is obtained. ;in, This represents the total energy data of the microseismic event corresponding to the xy spatial coordinate optical flow vector field map in the next unit of time. Indicates the current unit of time. Indicates a unit of time; Step 704, according to This yields the information entropy corresponding to the optical flow vector field map of the xy spatial coordinates at the current unit time. ;in, This represents the coordinates of the j-th pixel in the KDE corresponding to the xy coordinates of the current unit of time. The probability density at that location; Step 705: Obtain the minimum total cost of transferring the probability distribution corresponding to each pixel in the xy spatial coordinate optical flow vector field map of the current unit time to the probability distribution corresponding to the xy spatial coordinate optical flow vector field map of the next unit time as the first-order Wasserstein distance. Step 706: Repeat steps 701 to 705 multiple times to obtain the relevant parameters of the xz space coordinate optical flow vector field map; Step 707: Repeat steps 701 to 705 multiple times to obtain the relevant parameters of the optical flow vector field map in yz space coordinates.

[0029] In this embodiment, the optical flow vector concentration is used to measure the dispersion of local stress direction and can be used to indicate whether the stress direction tends to be stable. An increase in the optical flow vector concentration indicates that the structural disturbance is enhanced and may be accompanied by stress reconstruction. The energy dissipation rate represents the local energy dissipation characteristics and is used to describe the release rate of microseismic events per unit time. An increase in the energy dissipation rate indicates that the system is active and dissipation is enhanced, which may indicate plastic failure or local instability. A decrease in the energy dissipation rate indicates that the system is approaching a steady state or entering a period of energy accumulation.

[0030] Information entropy, as a statistical physical index of the evolution of microseismic density field, is used to quantitatively analyze the clustering or dispersion of microseismic events at different unit times. The higher the information entropy, the more discrete the distribution of microseismic events and the more disordered the internal rupture activity of the system, which may be in the stress adjustment or active development stage of microcrack network. On the other hand, a decrease in information entropy may correspond to the clustering of microseismic activity, indicating that the stress concentration area tends to be stable.

[0031] The first-order Wasserstein distance is introduced to quantitatively characterize the degree of change of the spatial distribution density of microseismic events in a continuous time period, so as to reveal the dynamic evolution characteristics of the stress state in the coal rock mass. The greater the first-order Wasserstein distance, the more intense the migration of the density center, which can be used to represent the rapid evolution of the stress redistribution or dissipation concentration area.

[0032] In the embodiment, the monitored mine area is a working face and a roadway; in actual use, the coordinate system of the spatial coordinates is not limited, as long as the actual monitoring requirements are met. The spatial coordinates are three-dimensional spatial coordinates.

[0033] In the embodiment, the coordinate KDE diagram includes an xy coordinate KDE diagram, an xz coordinate KDE diagram and a yz coordinate KDE diagram, so as to realize the mining stress evolution path analysis of different two-dimensional dimensions to reflect the microseismic migration direction in different directions.

[0034] In the embodiment, in actual use, according to the needs of the coal mine, the stress evolution path required by other requirements can be obtained from the spatial coordinate optical flow vector field diagram of different unit times.

[0035] In the embodiment, the unit time is a day, which can be adjusted according to actual requirements.

[0036] In the embodiment, in actual use, the spatial coordinates of each microseismic event are normalized by using the proportional normalization method to obtain the spatial normalized coordinates corresponding to each microseismic event; when the proportional normalization method is used, the maximum value of the x-axis coordinate, the maximum value of the y-axis coordinate and the maximum value of the z-axis coordinate are obtained, and the x-axis coordinate of the spatial coordinates of each microseismic event / the maximum value of the x-axis coordinate, the y-axis coordinate / the maximum value of the y-axis coordinate and the z-axis coordinate / the maximum value of the z-axis coordinate are normalized. Conversely, the maximum value of the x-axis coordinate, the maximum value of the y-axis coordinate and the maximum value of the z-axis coordinate are multiplied to perform denormalization.

[0037] In the embodiment, in actual use, is the bandwidth, and the value is 0.1.

[0038] In the embodiment, in actual use, the size of each grid is 1 / 600 to ensure the clarity of the subsequent spatial coordinate gray scale diagram.

[0039] In the embodiment, it should be noted that step 2044 specifically comprises: based on the probability density of each grid, the probability density is mapped to a color heat map by using python in the PyCharm compiler by using a computer to obtain the xy coordinate KDE diagram.

[0040] In the embodiment, in actual use, the optical flow vector of each pixel point is the velocity vector of each pixel point.

[0041] In the embodiment, in actual use, the microseismic phenomenon is closely related to the mechanical behavior of the rock. Under the action of external stress, the rock will deform, and when the energy accumulates to a certain critical point, it will trigger the rapid release and propagation of elastic waves or stress waves in the surrounding rock mass. By studying the spatiotemporal distribution law of rock failure, we can infer the development trend of macroscopic failure and further understand the potential disaster activity law, providing an important basis for disaster risk early warning.

[0042] In summary, the method of the present application is simple in steps and reasonable in design. After generating the coordinate KDE graph based on the normalized spatial coordinates of the microseismic monitoring data, the Farneback optical flow algorithm is used to obtain the optical flow vector field graph to quantitatively obtain the path reflecting the stress evolution caused by mining, realize the dynamic tracking of the stress path, and combine the streamline graph and the heat map to intuitively display the stress dissipation path in the mining process. It can capture the rapid changes of the stress field in real time, greatly improving the timeliness of the early warning.

[0043] The above is only a preferred embodiment of the present application, and does not limit the present application. Any simple modification, change and equivalent structural change made according to the technical essence of the present application to the above embodiment are still within the protection scope of the technical solution of the present application.

Claims

1. A mining stress evolution analysis method fusing optical flow tracking and multi-indexes, characterized in that, The method comprises the following steps: Step one, obtaining microseismic monitoring data: Using a microseismic monitoring system to monitor the area of the mine to be monitored, obtaining microseismic events in a unit of time and occurrence time, spatial coordinates and energy value corresponding to each microseismic event in a unit of time in a unit of time; and normalizing the spatial coordinates to obtain normalized spatial coordinates; Step two, generating a coordinate KDE map based on the normalized spatial coordinates of the microseismic monitoring data using a two-dimensional Gaussian kernel density estimation method: Step three, filtering and grayscale processing of the coordinate KDE map to obtain a spatial coordinate grayscale map: Step four, based on two adjacent spatial coordinate grayscale maps, using the Farneback optical flow algorithm to obtain the optical flow vector of each pixel point of the spatial coordinate grayscale map to obtain an optical flow vector field map; Step five, performing inverse normalization on the coordinates of the optical flow vector field map to obtain a spatial coordinate optical flow vector field map; Step six, based on the spatial coordinate optical flow vector field maps of different units of time, obtaining a path reflecting the evolution of mining stress.

2. The method according to claim 1, wherein the method is a fusion of optical flow tracking and multi-indexed mining-induced stress evolution analysis. The spatial coordinates of each microseismic event are normalized by proportional normalization to obtain normalized spatial coordinates corresponding to each microseismic event; wherein the normalized spatial coordinates include x-axis normalized coordinates, y-axis normalized coordinates and z-axis normalized coordinates; Step two, the specific process is as follows: Step 201, obtaining the maximum energy value Emax from the energy values corresponding to each microseismic event; Step 202, sequentially recording each microseismic event as the first microseismic event, the second microseismic event, the i-th microseismic event, the n-th microseismic event according to the occurrence time; wherein i is a positive integer, and 1≤i≤n; Step 203, record the energy value of the i-th microseismic event as and according to Wi= / Emax, get the weight Wi of the i-th microseismic event; Step 204, inputting the weight of n microseismic events into a computer, based on the x-axis normalized coordinates and y-axis normalized coordinates of n microseismic events, using a two-dimensional Gaussian kernel density estimation method to obtain an xy coordinate KDE map; Step 205, inputting the weight of n microseismic events into a computer according to the method of step 204, based on the x-axis normalized coordinates and z-axis normalized coordinates of n microseismic events, using a two-dimensional Gaussian kernel density estimation method to obtain an xz coordinate KDE map; Step 206, inputting the weight of n microseismic events into a computer according to the method of step 204, based on the y-axis normalized coordinates and z-axis normalized coordinates of n microseismic events, using a two-dimensional Gaussian kernel density estimation method to obtain a yz coordinate KDE map.

3. The method according to claim 2, wherein the method is a fusion of optical flow tracking and multi-indexed mining-induced stress evolution analysis. Step 204, the specific process is as follows: Step 2041, obtaining a probability density function based on the weights of the n microseismic events at the estimated probability density function , and ; wherein, is a bandwidth, is a kernel function, and the kernel function is a Gaussian function; represents any x-axis and y-axis coordinates, xi represents the x-axis normalized coordinate of the i th microseismic event, and yi represents the y-axis normalized coordinate of the i th microseismic event. Step 2042, setting the value range of the x-axis to 0-1 and the value range of the y-axis to 0-1, and dividing the entire region of the x-axis coordinates and y-axis coordinates into a two-dimensional grid; Step 2043, obtaining coordinates of each grid and inputting the probability density function to obtain the probability density of each grid; Step 2044, mapping the probability density of each grid to a color heat map to obtain an xy coordinate KDE map.

4. The method according to claim 2, wherein the method is a fusion of optical flow tracking and multi-indexed mining-induced stress evolution analysis. Step three, the specific process is as follows: Step 301, performing Gaussian filtering on the xy coordinate KDE map, the xz coordinate KDE map and the yz coordinate KDE map respectively using a computer to obtain a filtered xy coordinate KDE map, a filtered xz coordinate KDE map and a filtered yz coordinate KDE map; Step 302, using a computer to perform gray processing on the filtered xy coordinate KDE map, the filtered xz coordinate KDE map and the filtered yz coordinate KDE map respectively, to obtain an xy coordinate gray map, an xz coordinate gray map and a yz coordinate gray map.

5. The method according to claim 4, wherein the method is a fusion of optical flow tracking and multi-indexed mining-induced stress evolution analysis. Step four, the specific process is as follows: Step 401, according to the method of step two and step three, the microseismic monitoring data in the next unit time is processed to obtain the xy coordinate gray map of the next unit time, and the xy coordinate gray map of the current unit time and the xy coordinate gray map of the next unit time are recorded as adjacent two xy coordinate gray maps; Step 402, using a computer to process the adjacent two xy coordinate gray maps by using Farneback optical flow algorithm to obtain the optical flow vector of each pixel point on the xy coordinate gray map of the current unit time; Step 403, using a computer to draw optical flow vector arrows on the xy coordinate gray map of the current unit time based on the optical flow vector of each pixel point, to obtain an xy normalized coordinate optical flow vector field map; Step 404, according to the method of step 401 to step 403, an xz normalized coordinate optical flow vector field map is obtained; Step 405, according to the method of step 401 to step 403, a yz normalized coordinate optical flow vector field map is obtained.

6. The method according to claim 5, wherein the method is a fusion of optical flow tracking and multi-indexed mining-induced stress evolution analysis. Step five, the specific process is as follows: Step 501, the horizontal coordinate and the vertical coordinate in the xy normalized coordinate optical flow vector field map are de-normalized to obtain an xy space coordinate optical flow vector field map; Step 502, the horizontal coordinate and the vertical coordinate in the xz normalized coordinate optical flow vector field map are de-normalized to obtain an xz space coordinate optical flow vector field map; Step 503, the horizontal coordinate and the vertical coordinate in the yz normalized coordinate optical flow vector field map are de-normalized to obtain a yz space coordinate optical flow vector field map.

7. The method according to claim 6, wherein the method is a fusion optical flow tracking multi-index mining stress evolution analysis method. Step six, the specific process is as follows: Step 601, the optical flow vector of the jth pixel point on the xy space coordinate optical flow vector field map of the current unit time is recorded as (Vx(j), Vy(j)); wherein, Vx(j) represents the velocity component of the jth pixel point in the x-axis direction, and Vy(j) represents the velocity component of the jth pixel point in the y-axis direction; Step 602, the optical flow vectors of the jth pixel point on the xy space coordinate optical flow vector field maps of different unit times are accumulated to obtain the stress evolution path of the jth pixel point on the xy space coordinate optical flow vector field map; Step 603, steps 601 to 602 are repeated multiple times to obtain the stress evolution paths of each pixel point on the xy space coordinate optical flow vector field map; Step 604, according to , the flow rate amplitude of the jth pixel point is obtained ; according to , the optical flow intensity corresponding to the optical flow vector field diagram of the xy spatial coordinates in the current unit time is obtained ; wherein, 1≤j≤J, J represents the total number of pixel points; The median probability density is selected from the probability density corresponding to the xy space coordinate optical flow vector field map of the current unit time; pixels with probability density greater than the median probability density are marked as the stress concentration region in the current unit time. pixels with probability density greater than the median probability density are marked as the stress concentration region in the current unit time. Step 605, the stress evolution paths of the stress concentration region are screened out from the stress evolution paths of each pixel point on the xy space coordinate optical flow vector field maps of different unit times, and recorded as the stress evolution paths of the xy space stress concentration region; Step 606, according to the method of step 601 to step 605, the stress evolution paths of the xz space stress concentration region are obtained; Step 607, according to the method of step 601 to step 605, the stress evolution paths of the yz space stress concentration region are obtained. Step 607, according to the method of step 601 to step 605, the stress evolution path of the yz space stress concentration area is obtained.

8. The method according to claim 7, wherein the method is a fusion optical flow tracking multi-index mining stress evolution analysis method. After step six, multi-index parameter acquisition is also performed, which is as follows: Step 701, according to , the vector angle of the jth pixel point ; and according to ; obtaining a concentration of optical flow vectors ; Step 702, according to , get the total energy data of microseismic events corresponding to the xy space coordinate optical flow vector field diagram of the current unit time ; according to , get the energy mean value corresponding to the xy space coordinate optical flow vector field diagram of the current unit time ; Step 703, according to , the energy dissipation rate corresponding to the adjacent two xy spatial coordinate optical flow vector field map is obtained ; wherein, represents the total energy data of the microseismic event corresponding to the xy spatial coordinate optical flow vector field map of the next unit time, represents the current unit time, represents the unit time; Step 704, according to , the information entropy corresponding to the xy space coordinate optical flow vector field diagram of the current unit time is obtained ; wherein, represents the probability density of the jth pixel point coordinate at the current unit time corresponding to the xy coordinate KDE ; Step 705, the minimum total cost of the probability distribution corresponding to each pixel point of the xy space coordinate optical flow vector field diagram of the current unit time transferring to the probability distribution corresponding to the xy space coordinate optical flow vector field diagram of the next unit time is obtained as the first order Wasserstein distance; Step 706, steps 701 to 705 are repeated multiple times to obtain the related parameters of the xz space coordinate optical flow vector field diagram; Step 707, steps 701 to 705 are repeated multiple times to obtain the related parameters of the yz space coordinate optical flow vector field diagram.

Images