Ground pressure monitoring and early warning method and system based on multi-source data fusion

Abstract

The present application belongs to the technical field of data processing, and particularly relates to a ground pressure monitoring and early warning method and system based on multi-source data fusion, which comprises the following steps: constructing a time-space distance improved DBSCAN algorithm of microseismic events, clustering the microseismic events by the improved DBSCAN algorithm, and calculating the risk source index of each microseismic event cluster; in a three-dimensional grid space, according to the risk source index, the distance between the microseismic event cluster and the grid unit, and the stress state of the grid unit, obtaining the risk potential energy of each grid unit; combining the current value of the risk potential energy and its change over time to calculate the ground pressure risk index; and performing multi-level early warning according to the ground pressure risk index. The present application performs ground pressure early warning by fusing the microseismic time-space aggregation characteristics, stress field state and risk dynamic evolution trend, thereby improving the accuracy and timeliness of ground pressure disaster early warning.

Description

Technical Field

[0001] This invention relates to the field of data processing technology. More specifically, this invention relates to a ground pressure monitoring and early warning method and system based on multi-source data fusion. Background Technology

[0002] During mineral resource extraction, ground pressure disasters, such as rock bursts, mine tremors, and roof collapses, are among the major hazards threatening mine safety and the lives of personnel. To effectively prevent such disasters, real-time and accurate monitoring and early warning of the stress state and stability of the surrounding rock in the mine are crucial.

[0003] In related technologies, early warning methods for ground pressure disasters often rely on the analysis of a single monitoring data source. For example, by analyzing stress monitoring data, an alarm is triggered when the stress value exceeds a preset empirical threshold; or the risk is assessed solely by analyzing the energy of microseismic events.

[0004] However, the formation of ground pressure hazards is a complex spatiotemporal dynamic process, resulting from the interaction and coupled evolution of rock mass stress fields and microfracture activity. A single data source cannot comprehensively and accurately reflect the evolution of ground pressure hazards. Therefore, existing ground pressure analysis methods using a single data source have limitations. Furthermore, existing technologies often treat microseismic events as isolated risk points in ground pressure risk assessment, neglecting the stress waves they generate that disturb the stability of surrounding rock masses. This disturbance effect is amplified, especially in high-stress areas, leading to risk transmission and concentration. Existing technologies have failed to effectively establish spatial transmission models for this risk. On the other hand, existing ground pressure risk assessments rely solely on current monitoring values, lacking consideration of dynamic risk evolution trends. For example, an area with a high absolute risk value but long-term stability may be less dangerous than an area with a moderate risk value that is rapidly increasing. Therefore, existing technologies fail to adequately capture the temporal dynamic evolution characteristics of risks, affecting the timeliness and accuracy of early warnings. Summary of the Invention

[0005] To address the technical problems mentioned above regarding the use of a single data source for ground pressure monitoring and early warning, and the lack of spatiotemporal feature capture affecting the accuracy of ground pressure early warning, this invention provides solutions in the following aspects.

[0006] In a first aspect, the present invention provides a ground pressure monitoring and early warning method based on multi-source data fusion, comprising: acquiring microseismic data and stress data of various monitoring points within a monitoring area; taking microseismic events within a preset time window as objects, calculating the spatiotemporal distance between events based on the time and spatial location of the microseismic events, and clustering the microseismic events based on the spatiotemporal distance to obtain microseismic event clusters; obtaining a risk source index based on the energy of microseismic events within each microseismic event cluster, the number of microseismic events per unit time, and the average distance of the microseismic event cluster; the average distance of the microseismic event cluster is the mean of the Euclidean distances between the microseismic events within the microseismic event cluster; constructing a three-dimensional grid within the monitoring area to obtain various grid units; for any grid unit, obtaining the risk potential energy of the grid unit based on the distance between the grid unit and each microseismic event cluster, the risk source index of each microseismic event cluster, and the difference between the real-time stress value and the historical stress value at the corresponding position of the grid unit's center point; obtaining the ground pressure risk index of the grid unit based on the risk potential energy of the grid unit at the current moment and the change in risk potential energy between the current moment and the moment before the current moment; and issuing a ground pressure early warning based on the magnitude of the ground pressure risk index.

[0007] This invention clusters microseismic events by introducing spatiotemporal distance and calculates a risk source index by combining the energy, frequency, and spatial clustering of these events, enabling accurate identification of microseismic activities with genuine disaster risks. Secondly, by coupling the risk source index with stress state in a three-dimensional grid, risk potential energy is obtained, simulating the transmission and amplification effects of ground pressure risk in rock media, thus adapting to the spatial distribution patterns of risk. By fusing the current absolute value of risk potential energy with its changes over time to obtain the ground pressure risk index, ground pressure early warning can capture the dynamic evolution trend of ground pressure risk. This overcomes the limitations of traditional early warning technologies, which rely on a single data source, ignore the spatiotemporal coupling characteristics of risk, and lack dynamic evolution analysis. It elevates ground pressure risk assessment from isolated event analysis to global spatial analysis and from static threshold judgment to dynamic trend prediction, improving the accuracy, reliability, and timeliness of ground pressure disaster early warning.

[0008] Preferably, the spatiotemporal distance satisfies the following relationship: In the formula, Microseismic events Microseismic events The spacetime distance between them and Microseismic events were detected respectively Monitoring points detected microseismic events The value of the monitoring point on the X-axis in the equipment's spatial coordinate system. and Microseismic events were detected respectively Monitoring points and detected microseismic events The value of the monitoring point on the Y-axis in the equipment's spatial coordinate system. and Microseismic events were detected respectively Monitoring points and detected microseismic events The value of the monitoring point on the Z-axis in the equipment's spatial coordinate system. and Microseismic events Microseismic events The time of occurrence, This is the time weighting coefficient.

[0009] This invention treats time as an independent dimension and introduces a time-weighted coefficient, ensuring that only microseismic events exhibiting proximity in both space and time are considered strongly correlated. This aligns better with the spatiotemporal aggregation of micro-rupture activity during the formation of ground pressure disasters. Compared to traditional methods relying solely on spatial distance, spatiotemporal distance measurement effectively filters out random microseismic events that are spatially close but occur at vastly different times. This allows for more accurate identification of clusters of microseismic events with disaster potential, triggered by localized stress concentrations, providing a solid foundation for the accurate identification of subsequent risk sources.

[0010] Preferably, the clustering of microseismic events based on spatiotemporal distance includes: using spatiotemporal distance as a distance metric and clustering the microseismic events using the DBSCAN algorithm.

[0011] Preferably, the risk source index satisfies the following relationship: In the formula, For the target time Risk source index of microseismic event clusters For the target time The number of microseismic events contained within a microseismic event cluster. For the target time Each microseismic event cluster contains the sum of the energy values ​​of the microseismic events. For the target time The time difference between the first and last microseismic events within a microseismic event cluster. For the first The average distance of a cluster of microseismic events It is an exponential function with the natural constant as its base. This is the spatial clustering sensitivity coefficient. To prevent division by zero parameters.

[0012] This invention obtains a risk source index based on the total energy released by microseismic events, the number of events per unit time, and the average distance of microseismic event clusters. The number of events per unit time reflects the rate of rock mass fracturing, while the average distance of microseismic event clusters reflects the spatial concentration of microseismic events. By integrating the relationships between energy, temporal density, and spatial density of microseismic events, this invention achieves a comprehensive assessment of the disaster-causing potential of microseismic event clusters, making risk assessment more physically meaningful and improving the accuracy of ground pressure early warning.

[0013] Preferably, the step of constructing a three-dimensional grid within the monitoring area to obtain each grid cell includes: obtaining the smallest bounding cube of all monitoring points in the device spatial coordinate system, and dividing the smallest bounding cube into several grid cells of equal size.

[0014] Preferably, the risk potential energy satisfies the following relationship: In the formula, For the target time Risk potential energy of each grid cell For the target time Risk source index of microseismic event clusters For the target time The center point of the first grid cell and the... Euclidean distance between the centroids of a cluster of microseismic events For the first The stress value of each mesh element at the target time. For the first The average stress value of each mesh element at all times within the target time window. For the first The standard deviation of stress values ​​of each mesh element at all times within the target time window. The number of microseismic event clusters at the target time. To prevent division by zero parameters.

[0015] This invention obtains risk potential energy through a risk source index and local stress state. It simulates the physical phenomenon that the influence of a microseismic risk source on its surrounding rock mass decreases with increasing distance using a distance attenuation term. An amplification factor is constructed by introducing the difference between the historical average stress of real-time stress, reflecting the amplification effect of high stress concentration areas on risk. By organically combining isolated risk sources with a continuous stress field, risk assessment is no longer limited to the risk source itself but extends to the entire area it may affect, thereby identifying potentially high-risk areas far from microseismic event clusters but under high stress.

[0016] Preferably, the method for obtaining the real-time stress value at the location corresponding to the center point of the grid cell includes: obtaining the stress value at the center point of the grid cell using a Kriging interpolation algorithm based on the stress values ​​at each monitoring point.

[0017] Preferably, the ground pressure risk index satisfies the following relationship: In the formula, For the target time Ground pressure risk index for each grid cell For the target time Risk potential energy of each grid cell The moment before the target time Risk potential energy of each grid cell It is the hyperbolic tangent function.

[0018] This invention achieves dynamic assessment of ground pressure risk by obtaining a ground pressure risk index based on the current value and change of risk potential energy. This invention not only focuses on the absolute magnitude of risk potential energy but also pays closer attention to areas where risk potential energy is increasing by incorporating changes in risk potential energy at adjacent time points. This amplifies the ground pressure risk index in areas where risk potential energy is rapidly rising, enabling earlier identification of areas with drastically deteriorating ground pressure risk, thus buying time for preventative measures and improving the sensitivity of early warning systems.

[0019] Preferably, the ground pressure early warning based on the magnitude of the ground pressure risk index includes: obtaining the 95th percentile, 99th percentile, 99.9th percentile, and 99.99th percentile of the ground pressure risk index; triggering a level four alarm when the ground pressure risk index of any grid cell is between the 95th and 99th percentiles; triggering a level three alarm when the ground pressure risk index of any grid cell is between the 99th and 99.9th percentiles; triggering a level two alarm when the ground pressure risk index of any grid cell is between the 99.9th and 99.99th percentiles; and triggering a level one alarm when the ground pressure risk index of any grid cell exceeds the 99.99th percentile.

[0020] Secondly, the present invention provides a ground pressure monitoring and early warning system based on multi-source data fusion, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned ground pressure monitoring and early warning method based on multi-source data fusion is implemented.

[0021] By adopting the above technical solution, the above-mentioned ground pressure monitoring and early warning method based on multi-source data fusion is generated into a computer program and stored in a memory so that it can be loaded and executed by a processor. In this way, a terminal device can be made based on the memory and the processor for convenient use.

[0022] The beneficial effects of this invention are as follows: By clustering microseismic events based on their combined temporal and spatial distances, it more accurately identifies the sources of hazardous microseismic activity; by combining the energy, frequency, and spatial concentration of microseismic event clusters to assess risk sources, it makes the risk level classification of risk sources more reliable; by coupling the influence of each risk source on the corresponding location of the grid cell with the stress state of the rock mass in a three-dimensional grid space, it reveals the spatial distribution and transmission law of ground pressure risk throughout the monitoring area, rather than only assessing isolated risk points; finally, by comprehensively assessing the current value of the risk and its growth rate, the early warning system can identify areas of rapidly deteriorating risk earlier. This invention solves the problem of treating risk as an isolated static event in traditional technologies, realizing the transformation of ground pressure disaster monitoring from point-like monitoring to field-like analysis, and from static assessment to dynamic prediction, comprehensively improving the accuracy and reliability of early warning. Attached Figure Description

[0023] Figure 1 This is a flowchart illustrating the ground pressure monitoring and early warning method based on multi-source data fusion in this invention;

[0024] Figure 2 This is a schematic diagram illustrating the effect of microseismic event clustering based on spatiotemporal distance in this invention;

[0025] Figure 3 This is a schematic diagram illustrating the effect of existing algorithms on microseismic event clustering based on spatial distance;

[0026] Figure 4 This is a schematic diagram illustrating the alarm level distribution in this invention. Detailed Implementation

[0027] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0028] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0029] This invention discloses a ground pressure monitoring and early warning method based on multi-source data fusion, referring to... Figure 1 This includes steps S1-S5:

[0030] S1, Preprocessing of ground pressure monitoring data, and construction of equipment spatial coordinate system.

[0031] During the occurrence of ground pressure disasters, dynamic evolution of data such as stress and microseismic events occurs. Therefore, mine ground pressure monitoring and early warning can be carried out by comprehensively analyzing stress and microseismic data at various times. However, different data have different dimensions, and it is necessary to unify the dimensions of data from different sources in order to achieve effective fusion and analysis of multi-source data. Therefore, this invention first preprocesses the multi-source ground pressure monitoring data and constructs the equipment spatial coordinate system. Ground pressure monitoring and early warning are carried out using the preprocessed data of each point in the equipment spatial coordinate system.

[0032] Specifically, based on the actual conditions of the mine monitoring area, each monitoring point in the mine monitoring area is determined. At each monitoring point in the mine monitoring area, monitoring equipment such as microseismic sensors and stress gauges are deployed to collect data on stress and microseismic events at each monitoring point at various times. The data in each dimension are linearly normalized, and the mine ground pressure is monitored using the linearly normalized data.

[0033] Furthermore, taking the center of the main shaft opening of the mine as the origin of the equipment spatial coordinate system, taking the direction from west to east as the X-axis, the direction from south to north as the Y-axis, and the vertical upward direction as the Z-axis, an equipment spatial coordinate system is formed, and the coordinates of each monitoring point in the equipment spatial coordinate system are obtained.

[0034] S2. Construct a spatiotemporal distance and improve the DBSCAN algorithm through the spatiotemporal distance. Obtain each microseismic event cluster by collecting microseismic events from each monitoring point at each time and by improving the DBSCAN algorithm. Obtain the risk source index of each microseismic event cluster based on the microseismic events in the microseismic event cluster.

[0035] It should be noted that isolated, sporadic microseismic events are common in mining operations. However, when microseismic activity exhibits significant aggregation in both time and space, localized areas of the mine may experience high stress concentration and accelerated propagation of internal rock fissures, indicating a potential ground pressure disaster. Therefore, a comprehensive analysis of microseismic events across both temporal and spatial dimensions is necessary for accurate ground pressure early warning. This invention constructs a spatiotemporal distance model and improves the DBSCAN algorithm using this model. By analyzing microseismic events collected from various monitoring points at different times and applying the improved DBSCAN algorithm, microseismic event clusters are obtained. Finally, a risk source index is derived for each microseismic event cluster based on the microseismic events within that cluster.

[0036] Specifically, a time window is set with any given moment as the target moment. The time window contains all moments within 24 hours prior to the target moment, where 24 hours is the length of the time window. The implementers can adjust the size of the time window according to the actual situation. For any two microseismic events that occur within the target moment's time window, the spatiotemporal distance between the two microseismic events is obtained.

[0037] It should be added that microseismic events are collected by microseismic sensors. Each microseismic event includes the occurrence of the microseismic event, its spatial location, and the energy of the vibration.

[0038] Specifically, the spatiotemporal distance satisfies the following relationship:

[0039]

[0040] In the formula, Microseismic events Microseismic events The spacetime distance between them and Microseismic events were detected respectively Monitoring points and detected microseismic events The value of the monitoring point on the X-axis in the equipment's spatial coordinate system. and Microseismic events were detected respectively Monitoring points and detected microseismic events The value of the monitoring point on the Y-axis in the equipment's spatial coordinate system. and Microseismic events were detected respectively Monitoring points and detected microseismic events The value of the monitoring point on the Z-axis in the equipment's spatial coordinate system. and Microseismic events Microseismic events The time of occurrence, The time weighting coefficient is used in this embodiment. The value is 0.2, and the implementers can adjust the time weighting coefficient according to the actual situation.

[0041] in, This represents the spatial distance between two microseismic events. The larger the value, the farther apart the two microseismic events are, and the greater the spatiotemporal distance between them; the smaller the value, the closer the two microseismic events are, and the smaller the spatiotemporal distance between them. This represents the temporal distance between two microseismic events. A larger value indicates a greater time difference between the events, and thus a greater spatial-temporal distance between them; conversely, a smaller value indicates a smaller time difference, and thus a smaller spatial-temporal distance. Because... and They have different dimensions, therefore, through the time weighting coefficient right Adjustments were made to reduce the impact of the different dimensions of the two on the spatiotemporal distance.

[0042] Furthermore, spatiotemporal distance is used instead of the distance metric in the existing DBSCAN algorithm. The microseismic events within the time window are clustered using DBSCAN to obtain each microseismic event cluster. The mean Euclidean distance between the spatial locations of each microseismic event within the microseismic event cluster is obtained. This mean is used as the average distance of the microseismic event cluster. The risk source index of each microseismic event cluster is obtained based on the number of microseismic events contained in the microseismic event cluster, the energy value of the microseismic events, and the average distance of the microseismic events.

[0043] For example, Figure 2 This is a diagram illustrating the microseismic event clustering results based on spatiotemporal distance according to the present invention. Figure 3 The figure shows the effect of microseismic event clustering based on spatial distance. As can be seen from the figure, existing algorithms for microseismic event clustering based on spatial distance can only obtain microseismic events that are spatially close, and incorrectly classify microseismic events that are spatially close but occurred at different times into a single cluster. This obscures the true evolution process of ground pressure risk. In contrast, the present invention clusters microseismic events based on spatiotemporal distance, which can effectively distinguish microseismic events that are not sequential in time and avoid misjudgment caused by the spatial proximity of microseismic events. This enables a deeper and more physically meaningful division of event activity, providing more accurate input for subsequent ground pressure risk assessment.

[0044] Specifically, the risk source index satisfies the following relationship:

[0045] ;

[0046] In the formula, For the target time Risk source index of microseismic event clusters For the target time The number of microseismic events contained within a microseismic event cluster. For the target time Each microseismic event cluster contains the sum of the energy values ​​of the microseismic events. For the target time The time difference between the first and last microseismic events within a microseismic event cluster. For the first The average distance of a cluster of microseismic events It is an exponential function with the natural constant as its base. The spatial clustering sensitivity coefficient is used in this embodiment. The value is 0.1. Implementers can adjust this value according to the actual situation. To prevent division by zero parameters, in this embodiment... The value is 0.001, and the implementers can adjust it according to the actual situation. The value of .

[0047] in, The larger the number, the more likely it is to be the first. The higher the energy released by the vibrations corresponding to the microseismic events within a cluster of microseismic events, the higher the energy released by the vibrations corresponding to the microseismic events within that cluster. A cluster of microseismic events may correspond to a larger scale of rock mass fracturing and a greater risk of ground pressure. The larger the risk source index of a cluster of microseismic events, the greater the risk source index. The smaller the number, the more likely it is to be the first The lower the energy released by the vibrations corresponding to the microseismic events within a cluster of microseismic events, the better. A cluster of microseismic events may correspond to a smaller scale of rock mass fracturing and a smaller risk of ground pressure. The smaller the risk source index of a cluster of microseismic events, the better.

[0048] This represents the risk coefficient of microseismic events over time. The larger the number, the higher the value. The more densely packed the microseismic events within a cluster of microseismic events, the more unstable the rock mass becomes. The larger the risk source index of a cluster of microseismic events, the greater the risk source index. The smaller the number, the higher the value. The sparser the occurrence of microseismic events in a cluster of microseismic events, the more stable the state of the rock mass. The smaller the risk source index of a cluster of microseismic events, the better.

[0049] This represents the risk coefficient of microseismic events in the spatial dimension. The smaller the value, the more spatially concentrated the microseismic events are, and the more likely the rock mass is to undergo unstable changes. The larger the risk source index of a cluster of microseismic events, the greater the risk source index. The larger the value, the more sparse the microseismic events are in space, and the more stable the rock mass is. The smaller the risk source index of a cluster of microseismic events, the better.

[0050] S3. Construct a three-dimensional mesh in the equipment space coordinate system, and obtain the risk potential energy based on the risk source index and stress monitoring value.

[0051] It should be noted that the risks generated by microseismic event clusters are not limited to their origin but propagate like waves to the surrounding rock mass. This propagation is not uniform but is significantly affected by the stress state of the rock mass along the path. High-stress areas amplify or accelerate the propagation of risks, while low-stress areas may attenuate them. Therefore, this invention obtains the risk potential energy based on the risk source index and stress monitoring values.

[0052] Specifically, obtain the smallest bounding cube of all monitoring points in the equipment's spatial coordinate system, and divide this smallest bounding cube into several grid cells of equal size. For example, the smallest bounding cube can be divided into 5-meter grid cells. 5 meters A 5-meter grid cell is used. The stress value at the center point of each grid cell is obtained using the Kriging interpolation algorithm based on the stress sensor readings at each monitoring point. The mean and standard deviation of the stress values ​​at the center point of each grid cell are also obtained over the time window. The distance between the center point of each grid cell and the centroid of each microseismic event cluster is also obtained. The risk potential energy of each grid cell is obtained based on the risk source index of all microseismic event clusters, the distance between the centroid of the microseismic event cluster and the center point of the grid cell, and the mean and standard deviation of the stress values ​​at the center point of the grid cell over the time window. In this embodiment, the grid cell size is 5 meters. 5 meters The size of the grid unit is 5 meters, and the implementers can determine the size of the grid unit according to the actual situation of the mining area.

[0053] Specifically, the risk potential energy satisfies the following relationship:

[0054] ;

[0055] In the formula, For the target time Risk potential energy of each grid cell For the target time Risk source index of microseismic event clusters For the target time The center point of the first grid cell and the... Euclidean distance between the centroids of a cluster of microseismic events For the first The stress value of each mesh element at the target time. For the first The average stress value of each mesh element at all times within the target time window. For the first The standard deviation of stress values ​​of each mesh element at all times within the target time window. The number of microseismic event clusters at the target time. To prevent division by zero parameters, in this embodiment... The value is 0.001, and the implementers can adjust it according to the actual situation. The value of .

[0056] in, The larger the number, the more likely it is to be the first. The cluster of microseismic events and the first The farther away the first grid cell is, the more distant the second grid cell becomes. The vibration at the location corresponding to the first microseismic event cluster affects the first... The smaller the influence of the position of each grid cell; The smaller the number, the more likely it is to be the first The cluster of microseismic events and the first The closer the nth grid cell is, the more likely the nth grid cell is to be found. The vibration at the location corresponding to the first microseismic event cluster affects the first... The greater the influence of the location corresponding to each grid cell, the greater the impact; therefore, with As a risk source index The risk potential energy is calculated using the distance decay term; The larger the value, the more likely there is a distance from the first. The high-pressure risk source near the first grid cell affects the first If the first grid cell has an effect, then the second... The greater the risk potential energy of each grid cell; The smaller the value, the less likely it is that there is no high-pressure risk source affecting the first. If the first grid cell has an effect, then the second... The smaller the risk potential energy of each grid cell.

[0057] The larger the number, the more likely it is to be the first. The greater the stress value at the location corresponding to the first grid cell is compared to the historical overall level, the greater the stress value at the first grid cell location. The more likely the location of the first mesh element is to be in a state of stress concentration, the more likely the second mesh element is to be in a state of stress concentration. The more likely the location of a grid cell is to be at risk of ground pressure, the higher the probability of the risk. The greater the risk potential energy of each grid cell; The smaller the number, the more likely it is to be the first The smaller the stress value at the location corresponding to the first grid cell is compared to the historical overall level, the better. The more likely the location corresponding to the first mesh element is to be in a state of stress relaxation, the more likely the second mesh element is to be in a state of stress relaxation. The more likely the location corresponding to a grid cell is to be free from ground pressure risk, the more likely the first grid cell is to be the location of the second grid cell. The smaller the risk potential energy of each grid cell.

[0058] S4. Obtain the ground pressure risk index based on the risk potential energy of each grid cell and the change in risk potential energy at adjacent times.

[0059] It should be noted that the level of ground pressure risk depends not only on its current absolute value, but also on its rate of increase. An area with a low but rapidly rising risk value may be far more dangerous than an area with a high but long-term stable risk value. Therefore, this invention obtains the ground pressure risk index based on the risk potential energy of each grid cell and the change in risk potential energy at adjacent time points.

[0060] Specifically, the difference between the risk potential energy of each grid cell at two adjacent moments is obtained, and the ground pressure risk index is obtained based on the risk potential energy of the grid cell and the difference between the risk potential energy of the grid cell at two adjacent moments.

[0061] Specifically, the ground pressure risk index satisfies the following relationship:

[0062] ;

[0063] In the formula, For the target time Ground pressure risk index for each grid cell For the target time Risk potential energy of each grid cell The moment before the target time Risk potential energy of each grid cell It is the hyperbolic tangent function.

[0064] in, The larger the number, the more likely it is to be the first. The more likely the risk potential energy at the location corresponding to a grid cell is to be increasing rapidly, then the... The more likely the location of the first grid cell is to be at risk of ground pressure, the more likely the second grid cell is to be at risk of ground pressure. The higher the ground pressure risk index of a grid cell, the greater the risk. The smaller the number, the more likely it is to be the first The more likely the risk potential energy at the location corresponding to the first grid cell is decreasing, the more likely the risk potential energy at the second grid cell is decreasing. The less likely the location corresponding to the first grid cell is to be subject to ground pressure risk, the more likely the first grid cell is to be subject to ground pressure risk. The smaller the ground pressure risk index of each grid cell, the better. Therefore, through... right Further adjustments will be made to comprehensively consider both the absolute and relative magnitudes of the risk potential energy, thereby achieving accurate early warning for ground pressure monitoring.

[0065] It should be noted that when the target time is... Risk potential of each grid cell When the value is 0, it indicates that there is no risk potential energy transmitted from microseismic events in the grid cell at the target time. Therefore, at this time, let the target time be... The ground pressure risk index for each grid cell is also set to 0 to avoid the potential problem of dividing by zero during the calculation of the ground pressure risk index.

[0066] S5. Conduct ground pressure early warning based on the ground pressure risk index.

[0067] Specifically, during the operation of the ground pressure monitoring system, the ground pressure risk index is obtained at each moment and for each grid unit. The 95th, 99th, 99.9th, and 99.99th percentiles of the ground pressure risk index are also obtained. When the ground pressure risk index of any grid unit is between the 95th and 99th percentiles, a Level IV alarm is triggered to alert attention to ground pressure changes. When the ground pressure risk index of any grid unit is between the 99th and 99.9th percentiles, a Level III alarm is triggered, indicating the presence of ground pressure anomalies. When the ground pressure risk index of any grid unit is between the 99.9th and 99.99th percentiles, a Level II alarm is triggered, and ground pressure disaster prevention measures should be taken. When the ground pressure risk index of any grid unit exceeds the 99.99th percentile, a Level I alarm is triggered, indicating that a ground pressure disaster has occurred, and the highest level alarm is issued.

[0068] It should be noted that percentiles are a commonly used concept in statistics. For example, the 95th percentile means that when all data are arranged from smallest to largest, 95% of the data are less than or equal to the value. The other 5% of the data are greater than the numerical value. Then the value This represents the 95th percentile of all data; the 99th, 99.9th, and 99.99th percentiles are calculated similarly and will not be elaborated upon here.

[0069] For example, Figure 4 This is an alarm level distribution map in the present invention. The map shows the alarm levels of some grid cells. The location of ground pressure risk can be clearly seen through the alarm level distribution map, and preventive measures can be taken according to the location of ground pressure risk.

[0070] This invention also discloses a ground pressure monitoring and early warning system based on multi-source data fusion, including a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, the ground pressure monitoring and early warning method based on multi-source data fusion according to this invention is implemented.

[0071] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

Claims

1. A ground pressure monitoring and early warning method based on multi-source data fusion, characterized in that, include: Acquire microseismic data and stress data at various monitoring points within the monitoring area; Taking microseismic events within a preset time window as the object, the spatiotemporal distance between events is calculated based on their occurrence time and spatial location. The DBSCAN algorithm is improved by using spatiotemporal distance instead of the distance metric used in existing DBSCAN technologies. Microseismic events are then clustered based on this spatiotemporal distance to obtain microseismic event clusters. A risk source index is obtained based on the energy of microseismic events within each cluster, the number of microseismic events per unit time, and the average distance of the microseismic event cluster. In the formula, For the target time Risk source index of microseismic event clusters For the target time The number of microseismic events contained within a microseismic event cluster. For the target time Each microseismic event cluster contains the sum of the energy values ​​of the microseismic events. For the target time The time difference between the first and last microseismic events within a microseismic event cluster. For the first The average distance of a cluster of microseismic events It is an exponential function with the natural constant as its base. This is the spatial clustering sensitivity coefficient. To prevent division by zero parameters; the average distance of the microseismic event cluster is the average of the Euclidean distances between each microseismic event within the cluster; Spatiotemporal distance satisfies the following relationship: In the formula, Microseismic events Microseismic events The spatiotemporal distance between them and Microseismic events were detected respectively Monitoring points detected microseismic events The value of the monitoring point on the X-axis in the equipment's spatial coordinate system. and Microseismic events were detected respectively Monitoring points detected microseismic events The value of the monitoring point on the Y-axis in the equipment's spatial coordinate system. and Microseismic events were detected respectively Monitoring points detected microseismic events The value of the monitoring point on the Z-axis in the equipment's spatial coordinate system. and Microseismic events Microseismic events The time of occurrence, This is the time weighting coefficient; By treating time as an independent dimension and introducing a time weighting coefficient, we ensure that only microseismic events that exhibit proximity in both spatial and temporal dimensions are considered strongly correlated. A three-dimensional grid is constructed within the monitoring area to obtain each grid cell. For any grid cell, the risk potential energy of the grid cell is obtained based on the distance between the grid cell and each microseismic event cluster, the risk source index of each microseismic event cluster, and the difference between the real-time stress value and the historical stress value at the corresponding position of the grid cell center point. Risk potential energy satisfies the following relationship: In the formula, For the target time Risk potential energy of each grid cell For the target time The center point of the first grid cell and the... Euclidean distance between the centroids of a cluster of microseismic events For the first The stress value of each mesh element at the target time. For the first The average stress value of each mesh element at all times within the target time window. For the first The standard deviation of stress values ​​of each mesh element at all times within the target time window. The number of microseismic event clusters at the target time; The ground pressure risk index of the grid cell is obtained based on the risk potential energy of the grid cell at the current moment and the change in risk potential energy between the current moment and the previous moment. In the formula, For the target time Ground pressure risk index for each grid cell The moment before the target time Risk potential energy of each grid cell It is the hyperbolic tangent function; Ground pressure early warning is issued based on the magnitude of the ground pressure risk index.

2. The ground pressure monitoring and early warning method based on multi-source data fusion according to claim 1, characterized in that, The clustering of microseismic events based on spatiotemporal distance includes: using spatiotemporal distance as a distance metric and clustering microseismic events using the DBSCAN algorithm.

3. The ground pressure monitoring and early warning method based on multi-source data fusion according to claim 1, characterized in that, The step of constructing a three-dimensional mesh within the monitoring area to obtain each mesh unit includes: obtaining the smallest bounding cube of all monitoring points in the device's spatial coordinate system, and dividing the smallest bounding cube into several mesh units of equal size.

4. The ground pressure monitoring and early warning method based on multi-source data fusion according to claim 1 or 3, characterized in that, The method for obtaining the real-time stress value at the location corresponding to the center point of the grid cell includes: obtaining the stress value at the center point of the grid cell using a Kriging interpolation algorithm based on the stress values ​​at each monitoring point.

5. The ground pressure monitoring and early warning method based on multi-source data fusion according to claim 1, characterized in that, The method of issuing ground pressure early warnings based on the magnitude of the ground pressure risk index includes: Obtain the 95th, 99th, 99.9th, and 99.99th percentiles of the ground pressure risk index; trigger a Level 4 alarm when the ground pressure risk index of any grid cell is between the 95th and 99th percentiles; trigger a Level 3 alarm when the ground pressure risk index of any grid cell is between the 99th and 99.9th percentiles; trigger a Level 2 alarm when the ground pressure risk index of any grid cell is between the 99.9th and 99.99th percentiles; trigger a Level 1 alarm when the ground pressure risk index of any grid cell exceeds the 99.99th percentile.

6. A ground pressure monitoring and early warning system based on multi-source data fusion, characterized in that, include: A processor and a memory, wherein the memory stores computer program instructions that, when executed by the processor, implement the ground pressure monitoring and early warning method based on multi-source data fusion according to any one of claims 1-5.

Images