Underground coal mine geological disaster early warning method and system and medium

Abstract

The invention discloses an underground coal mine geological disaster early warning method and system and a medium, and relates to the technical field of disaster early warning. The method comprises the following steps: constructing a multi-parameter cooperative monitoring field by arranging multi-modal sensing equipment in an underground coal mine; deploying N geological disaster recognition nodes for the N monitoring fields; the collaborative operation node performs risk identification on the returned data and outputs a geological disaster pre-judgment result; and the early warning management and control middle station receives the spatial topology of the monitoring field, executes disaster evolution prediction on a pre-judgment result according to the spatial topology, and outputs real-time early warning. According to the method, the technical problems of difficulty in real-time cooperative processing of multi-source data and insufficient disaster evolution prediction capability in underground coal mine geological disaster early warning are solved, and the underground coal mine disaster risk is dynamically analyzed through cooperative monitoring of a multi-modal sensor and cooperative work of time sequence data analysis and geological disaster recognition nodes, so that the underground coal mine disaster early warning efficiency is improved. And the accuracy and the response speed of geological disaster early warning are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of disaster early warning, in particular to a coal mine underground geological disaster early warning method, system and medium. BACKGROUND

[0002] The complexity and constantly changing conditions of the geological environment of the coal mine operation surface and the surrounding area make the monitoring and early warning of coal mine underground geological disasters a key to ensuring the safety of mine production. In recent years, with the increase in the depth of coal mines and the progress of mining technology, the risk of geological disasters in coal mines has gradually increased, mainly including coal seam collapse, gas leakage, coal and gas outburst, rock burst, etc. These disasters not only pose a serious threat to the safety of mine workers, but also directly affect the production efficiency and economic benefits of coal mines.

[0003] The existing coal mine underground disaster early warning system has some problems: on the one hand, the traditional monitoring method mainly relies on a single sensor (such as a gas concentration sensor, a stress sensor or a microseismic sensor), and the limitations of data acquisition make it impossible to comprehensively evaluate the geological changes in the coal mine underground; on the other hand, many traditional monitoring systems lack comprehensive analysis and early warning capabilities for multiple disaster factors, and cannot provide timely and accurate early warning signals, resulting in slow response when mine disasters occur and missing the best emergency handling opportunity. SUMMARY

[0004] The present application provides a coal mine underground geological disaster early warning method, system and medium, which solves the technical problems of difficulty in real-time collaborative processing of multi-source data and insufficient disaster evolution prediction capability in coal mine underground geological disaster early warning.

[0005] In view of the above problems, the present application provides a coal mine underground geological disaster early warning method, system and medium.

[0006] The first aspect of the present application provides a coal mine underground geological disaster early warning method, which comprises: constructing a multi-parameter collaborative monitoring field by arranging multi-modal sensing devices in the coal mine underground space, wherein the multi-parameter collaborative monitoring field is composed of N multi-parameter coverage monitoring fields; locally deploying N geological disaster identification nodes for the N multi-parameter coverage monitoring fields, wherein the N geological disaster identification nodes establish bidirectional real-time communication with an early warning control center; cooperatively operating the N geological disaster identification nodes to perform risk identification on N local multi-modal data returned by the N multi-parameter coverage monitoring fields, and outputting N geological disaster prediction results; the early warning control center receives and performs disaster evolution prediction on the N geological disaster prediction results according to the spatial correlation topology of the N multi-parameter coverage monitoring fields, and outputs real-time geological disaster early warning.

[0007] In a second aspect of the present application, a coal mine underground geological disaster early warning system is provided, comprising: a sensor group installation module; a monitoring field construction module; a multi-parameter collaborative monitoring field is constructed by deploying multi-modal sensing devices in the coal mine underground space, wherein the multi-parameter collaborative monitoring field is composed of N multi-parameter coverage monitoring fields; a node deployment module; N geological disaster identification nodes are deployed locally in the N multi-parameter coverage monitoring fields, wherein the N geological disaster identification nodes establish bidirectional real-time communication with an early warning management and control center station; a risk identification module; N local area multi-modal data returned by the N multi-parameter coverage monitoring fields are risk identified by the N geological disaster identification nodes in collaborative operation, and N geological disaster prediction results are output; a disaster prediction module; the early warning management and control center station receives and executes disaster evolution prediction on the N geological disaster prediction results according to the spatial correlation topology of the N multi-parameter coverage monitoring fields, and outputs real-time geological disaster early warning.

[0008] In a third aspect of the present application, a computer readable medium is provided, which stores a computer program, and the program is executed by a processor to implement the coal mine underground geological disaster early warning method provided by the present application.

[0009] The one or more technical solutions provided in the present application have at least the following technical effects or advantages:

[0010] Firstly, a multi-parameter collaborative monitoring field is constructed by deploying multi-modal sensing devices in the coal mine underground space, wherein the multi-parameter collaborative monitoring field is composed of N multi-parameter coverage monitoring fields; then, N geological disaster identification nodes are deployed locally in the N multi-parameter coverage monitoring fields, wherein the N geological disaster identification nodes establish bidirectional real-time communication with an early warning management and control center station; further, N local area multi-modal data returned by the N multi-parameter coverage monitoring fields are risk identified by the N geological disaster identification nodes in collaborative operation, and N geological disaster prediction results are output; finally, the early warning management and control center station receives and executes disaster evolution prediction on the N geological disaster prediction results according to the spatial correlation topology of the N multi-parameter coverage monitoring fields, and outputs real-time geological disaster early warning. The technical problems of difficulty in real-time collaborative processing of multi-source data and insufficient disaster evolution prediction capability in coal mine underground geological disaster early warning are solved, and the technical effect of improving the accuracy and response speed of geological disaster early warning by multi-modal sensor collaborative monitoring, combined with time series data analysis and collaborative work of geological disaster identification nodes, dynamic analysis of disaster risk in the coal mine underground is achieved. BRIEF DESCRIPTION OF DRAWINGS

[0011] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 This is a schematic diagram of the coal mine underground geological disaster early warning method provided in the embodiments of this application;

[0013] Figure 2 This is a schematic diagram of the structure of an underground geological disaster early warning system for coal mines provided in an embodiment of this application.

[0014] Figure labeling: Monitoring field construction module 11, Monitoring field construction module 12, Risk identification module 13, Disaster prediction module 14. Detailed Implementation

[0015] This application provides a method, system, and medium for early warning of geological disasters in coal mines, which solves the technical problems of difficulty in real-time collaborative processing of multi-source data and insufficient ability to predict the evolution of disasters in early warning of geological disasters in coal mines.

[0016] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0017] It should be noted that the terms "comprising" and "having" are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to these processes, methods, products, or devices.

[0018] Example 1, as Figure 1 As shown, this application provides a method for early warning of geological hazards in coal mines, wherein the method includes:

[0019] By deploying multimodal sensing devices in the underground space of a coal mine, a multi-parameter collaborative monitoring field is constructed, wherein the multi-parameter collaborative monitoring field consists of N multi-parameter coverage monitoring fields.

[0020] In one embodiment, in order to realize real-time monitoring and early warning of geological disasters in a coal mine underground, a multi-parameter collaborative monitoring field needs to be constructed by laying out various types of sensing devices in different areas of the mine, wherein the multi-modal sensing devices include microseismic sensors, gas concentration sensors, etc., which can capture changes in different physical quantities simultaneously and reflect the geological and environmental changes in the mine in real time; the multi-parameter collaborative monitoring field refers to a system for comprehensive monitoring of the same geological area in a coal mine underground by multiple sensors, which includes N multi-parameter coverage monitoring fields, and each sensor in each multi-parameter coverage monitoring field is responsible for collecting different types of data to reveal geological changes at different levels in the mine and help identify possible disaster risks.

[0021] Further, by laying out multi-modal sensing devices in the coal mine underground space and constructing a multi-parameter collaborative monitoring field, the method comprises:

[0022] A three-dimensional geomechanical model of the coal mine underground space is constructed, and the working face of the three-dimensional geomechanical model is divided into multiple twin grids based on dynamic risk weights; based on the effective detection radius and directional attenuation characteristics of the microseismic sensor, an initial sensing layout array is selected in the multiple twin grids; the initial sensing layout array is optimized with engineering constraints to compensate for sensor layout deviation, and a microseismic sensing layout array is obtained; by analogy, a gas concentration sensing layout array and a stress monitoring sensor array are analyzed and output; according to the microseismic sensing layout array, the gas concentration sensing layout array and the stress monitoring sensor array, multi-modal sensing devices are deployed in the coal mine underground space to construct and generate the multi-parameter collaborative monitoring field.

[0023] Preferably, first, based on the geological exploration data of the coal mine underground, a three-dimensional geomechanical model reflecting the geological conditions and stress state of the coal mine is constructed using three-dimensional geological modeling software. The model includes the geological stratigraphic position, coal seam distribution, rock mass structure, fault information, mining operation face, etc. of the mine, and can reflect the dynamic changes of the mine working face and surrounding area in real time. The three-dimensional geomechanical model provides a spatial basis for the entire monitoring, ensuring the scientificity and accuracy of sensor placement. Subsequently, by analyzing the geological environment, stress distribution, etc. in the three-dimensional geomechanical model, and according to the dynamic risk weight, the mine working face is divided into multiple twin grids, which represent the risk level of different areas, and the areas with high risk correspond to denser monitoring arrangement. The risk weight is calculated according to the actual situation of the mine, such as changes in ground stress, fluctuations in gas concentration, microseismic activity, etc. By dividing the working face into multiple grids, different monitoring strategies can be adopted for different risk areas. Then, according to the effective detection radius and directional attenuation characteristics of the microseismic sensor, a preliminary sensor placement array is selected in the multiple twin grids. The detection radius of the microseismic sensor will be affected by the environment in the mine, so the coverage of the sensor should be fully considered when placing it. The preliminary sensor array should ensure that it can cover each grid area and effectively detect possible microseismic events to maximize the monitoring range. Based on the initial placement array, considering the complexity of the mine environment (such as sensor installation deviation, mine space limitations, etc.), the particle swarm optimization algorithm (PSO) is used to optimize the sensor placement array with engineering constraints, adjusting the sensor position. The optimization goal is to minimize the monitoring blind area caused by placement errors or mine environmental factors, ensuring the accuracy of sensor placement. Through optimization, the final microseismic sensor placement array is obtained, making the sensor layout more effective and accurate. Similar analysis is also conducted for the placement of gas concentration sensors and stress monitoring sensors. The gas concentration sensor placement array needs to be placed in areas that are easily affected by gas accumulation to timely detect potential gas leakage risks, while the stress monitoring sensor array should be placed in areas that are easily affected by geological changes or mining activities. Finally, based on the optimized microseismic sensor placement array, gas concentration sensor placement array, and stress monitoring sensor placement array, the deployment of multi-modal sensing equipment is carried out in the mine underground. These sensors will work together to achieve comprehensive monitoring of the geological environment in the mine underground. After deployment, the sensors in each area will form a multi-parameter coverage monitoring field, which will constitute a multi-parameter collaborative monitoring field. This multi-parameter collaborative monitoring field can dynamically capture data on microseismic, gas concentration, ground stress, etc. Through data fusion and analysis, it can provide early warning of possible geological disasters, ensuring the safety of mine workers and improving the production efficiency of the mine.

[0024] N geological disaster identification nodes are locally deployed in the N multi-parameter coverage monitoring fields, and the N geological disaster identification nodes establish bidirectional real-time communication with the early warning management and control center.

[0025] In one embodiment, each multi-parameter coverage monitoring field locally deploys a geological disaster identification node, which serves as the core part of data processing and risk analysis, responsible for receiving and processing data from the respective monitoring field. Specifically, each geological disaster identification node can obtain real-time sensor data of the local monitoring field and process it through internal microseismic activity anomaly detection channels, gas concentration gradient warning channels, and stress mutation identification channels to identify signals such as microseismic activity, gas anomalies, and stress mutations that may indicate geological disasters. These geological disaster identification nodes not only can independently perform preliminary disaster warning tasks, but also can timely return analysis results to the early warning management and control center. The early warning management and control center is responsible for aggregating data from various nodes and making global disaster evolution predictions. To ensure real-time response, each geological disaster identification node establishes bidirectional real-time communication with the early warning management and control center. In this way, in addition to feeding back disaster prediction results to the center, the nodes can also receive instructions and configurations from the center, ensuring that the nodes can be flexibly adjusted according to real-time changes in the geological environment, thereby effectively avoiding or reducing disaster losses.

[0026] Further, N geological disaster identification nodes are locally deployed in the N multi-parameter coverage monitoring fields, and the method comprises:

[0027] After the first multi-parameter coverage monitoring field is locally called for risk events, risk deviation threshold aggregation is performed to obtain a dynamic energy baseline, a concentration mutation threshold, and a structure instability critical value. The dynamic energy baseline, the concentration mutation threshold, and the structure instability critical value are excluded from the local risk events to obtain a plurality of sample risk data, wherein the sample risk data includes sample rock mass fracture dynamics sequence, sample gas migration trend field, sample local stress evolution tensor field, and sample risk situation grade. The dynamic energy baseline, the concentration mutation threshold, and the structure instability critical value are mapped and loaded to the microseismic activity anomaly detection channel, the gas concentration gradient warning channel, and the stress mutation identification channel as single-dimensional risk judgment conditions. The plurality of sample risk data are used as training data to perform model parameter adjustment of the comprehensive risk decision channel, complete the localization of the standard disaster identification node, and obtain the first geological disaster identification node.

[0028] Optionally, in the first multi-parameter coverage monitoring field, historical disaster events that have occurred in the monitoring field are extracted from historical monitoring data, such as historical microseismic events, gas concentration overrun events, stress mutation events, etc., and these historical disaster events will form a local risk event set. Subsequently, by performing aggregation calculation on the local risk event set, a risk deviation threshold aggregation operation is performed. Specifically, the energy of the historical microseismic events is statistically calculated by using a sliding window, and a dynamic threshold is calculated by using a quantile method (95% quantile) to determine a dynamic energy baseline, which is used to judge the background energy level of microseismic activity and can help identify whether there is abnormal stratum movement; based on the gas concentration gradient in the gas concentration overrun event, a wavelet transform is used to detect the mutation point, and the mean value of the historical mutation gradient and the sum of multiple standard deviations (usually 1.5) are used as the concentration mutation threshold, which is used to analyze the change of gas concentration and help judge whether there is a risk of gas leakage or accumulation; based on the anchor rod pressure gauge instability precursor data in the stress mutation event, the upper limit of the confidence interval is calculated by using the Hoeffding inequality to obtain a structure instability threshold, which can help identify whether there is a risk of rock mass structure instability in the mine. Then, the data corresponding to the dynamic energy baseline, the concentration mutation threshold and the structure instability threshold are removed from the local risk event set to prevent them from interfering with the identification of other risk factors. The data after removal will form a plurality of sample risk data, which include a sample rock mass fracture dynamics sequence, a sample gas migration trend field, a sample local stress evolution tensor field and a sample risk situation level. The sample rock mass fracture dynamics sequence includes an energy-time sequence of microseismic events, the sample gas migration trend field includes a spatial gradient field of gas concentration distribution, the sample local stress evolution tensor field includes a principal stress direction change matrix of stress monitoring points, and the risk situation level includes an artificially labeled event consequence level. Then, the obtained dynamic energy baseline, concentration mutation threshold and structure instability threshold are used as single-dimensional risk judgment conditions and are mapped and loaded into a microseismic activity anomaly detection channel, a gas concentration gradient early warning channel and a stress mutation identification channel. The microseismic activity anomaly detection channel is used to detect whether the microseismic data exceeds the preset dynamic energy baseline to judge whether there are signs of abnormal stratum movement or collapse. The gas concentration gradient early warning channel is used to monitor the mutation of gas concentration, and if the concentration exceeds the preset threshold, a gas leakage warning is issued. The stress mutation identification channel is used to monitor the change of ground stress, and if the stress change exceeds the set instability threshold, it is determined that there is a risk of structure instability or rock burst in the mine.Finally, the obtained sample risk data is input into the algorithm model (such as a neural network) of the comprehensive risk decision channel as training data, and through steps such as forward propagation, loss calculation, back propagation, and parameter optimization, the parameters are gradually adjusted, so that the comprehensive risk decision channel can accurately assess disaster risks and provide decision support based on multi-dimensional risk data, help to comprehensively analyze the outputs of multiple channels, and ensure the accuracy of risk judgment. After completing this process, the first geological disaster identification node with local disaster identification capability is finally formed, which can judge the potential disaster risk in the mine in real time based on local sensor data and multiple internal channels, and provide data support for mine safety management.

[0029] The N geological disaster identification nodes are cooperatively operated to identify risks of N local multi-modal data returned by the N multi-parameter coverage monitoring fields, and N geological disaster prediction results are output.

[0030] In one embodiment, the local multi-modal data collected by the N multi-parameter coverage monitoring fields includes microseismic signals, gas concentration, stress changes and other sensor data, which can comprehensively reflect the geological and environmental changes in the mine. Subsequently, the collected data is returned to the corresponding geological disaster identification nodes, which are cooperatively operated, and at the same time, the received local multi-modal data is analyzed according to the internal microseismic activity anomaly detection channel, the gas concentration gradient warning channel and the stress mutation identification channel, and if necessary, the comprehensive risk decision channel is also used for analysis, so as to obtain the geological disaster prediction results of each multi-parameter coverage monitoring field. These geological disaster prediction results not only reflect the disaster risk state of each region, but also provide data support for subsequent disaster evolution prediction, early warning and emergency response, thereby enhancing the predictability and response ability of mine safety management.

[0031] Further, the N geological disaster identification nodes are cooperatively operated to identify risks of N local multi-modal data returned by the N multi-parameter coverage monitoring fields, and N geological disaster prediction results are output, the method comprising:

[0032] The multi-index mean of the first local multi-modal data is solved, and a first real-time rock mass fracture dynamics sequence, a first real-time gas migration trend field and a first real-time ground stress evolution tensor field are output; the first real-time rock mass fracture dynamics sequence, the first real-time gas migration trend field and the first real-time ground stress evolution tensor field are loaded to the microseismic activity anomaly detection channel, the gas concentration gradient early warning channel and the stress mutation identification channel of the first geological disaster identification node, and a single-dimensional comparison result group is output; if any result of the single-dimensional comparison result group is an index deviation, the deviated data is directly output as a first geological disaster prediction result; if all the single-dimensional comparison results are passed, the comprehensive risk decision channel is activated to perform risk fusion analysis on the first real-time rock mass fracture dynamics sequence, the first real-time gas migration trend field and the first real-time ground stress evolution tensor field, and the first geological disaster prediction result is output.

[0033] Optionally, first, multi-index mean value solving is performed on the multi-sensor data in the first local multi-modal data, including microseismic energy data, gas concentration data, and stress monitoring data, etc. By calculating the mean value of microseismic energy, the first real-time rock mass fracture dynamics sequence can be obtained, which can reflect whether there is abnormal rock stratum activity in the mine and provide a basis for subsequent disaster risk judgment. By calculating the mean value of gas concentration, the first real-time gas migration trend field can be obtained, which can reveal the potential risk of gas leakage or accumulation. By calculating the mean value of stress monitoring data, the first real-time geo-stress evolution tensor field can be obtained, which reflects the change of stress in the mine and provides important data support for structural disasters (such as rock burst, collapse, etc.). Subsequently, the first real-time rock mass fracture dynamics sequence, the first real-time gas migration trend field and the first real-time geo-stress evolution tensor field are loaded into the microseismic activity anomaly detection channel, the gas concentration gradient early warning channel and the stress mutation identification channel of the first geological disaster identification node respectively. By comparing with the dynamic energy baseline, the concentration mutation threshold and the structure instability critical value, the single-dimensional comparison result set of each channel can be obtained, which reflects whether each channel deviates. If the single-dimensional comparison result of a certain channel is index deviation, that is, the data exceeds the preset dynamic energy baseline, concentration mutation threshold or structure instability critical value, the deviated data is directly output as the first geological disaster prediction result of the first local multi-modal data. On the contrary, if the comparison results of all channels meet the normal threshold value, that is, there is no deviation, the comprehensive risk decision channel will be activated. The channel will perform risk fusion analysis on the first real-time rock mass fracture dynamics sequence, the first real-time gas migration trend field and the first real-time geo-stress evolution tensor field transmitted by the microseismic activity anomaly detection channel, the gas concentration gradient early warning channel and the stress mutation identification channel, and output the first geological disaster prediction result. The result is based on the comprehensive evaluation of all data sources and provides an overall risk situation level. This result can help mine managers to early warning and take timely measures for emergency treatment, reducing the impact of potential disasters.

[0034] The pre-warning management and control middle platform receives and executes disaster evolution prediction on the N geological disaster prediction results according to the spatial correlation topology of the N multi-parameter coverage monitoring fields, and outputs real-time geological disaster warning.

[0035] In one embodiment, the early warning management and control platform is responsible for receiving N geological disaster prediction results from various geological disaster identification nodes, which reflect the disaster risk states of different multi-parameter coverage monitoring fields. After receiving these prediction results, the early warning management and control platform performs disaster risk coupling superposition of adjacent monitoring fields according to the spatial correlation topology, and then performs disaster evolution prediction according to the superposition results to determine the disaster evolution path. Finally, the early warning management and control platform generates a real-time geological disaster warning by combining all analysis results, which includes disaster risk assessment results of each area in the mine and indicates the location and level of possible disasters, so as to ensure that mine workers can learn about the potential disaster risks in time and maximize the safety of mine workers.

[0036] Further, the method further comprises:

[0037] Before the deployment of the edge computing node, the standard disaster identification node is tested for effective coverage accuracy to obtain a multi-parameter fusion coverage scale; based on the multi-parameter fusion coverage scale, the multi-parameter collaborative monitoring field is divided into the N multi-parameter coverage monitoring fields; and according to the spatial distribution correlation relationship of the N multi-parameter coverage monitoring fields in the multi-parameter collaborative monitoring field, the spatial correlation topology is constructed.

[0038] Optionally, before deploying the geological disaster identification node as an edge computing node, the standard disaster identification node will be tested for coverage accuracy. The standard disaster identification node is a trained geological disaster identification node. During the coverage accuracy test, the standard disaster identification node will be deployed in an experimental mine or a simulation platform (loaded with a three-dimensional geomechanical model), and a coordinate-known signal emitter array will be set up around the space, and the emission of simulated signals will be carried out through the signal emitter array. Subsequently, according to the simulation results, the proportion of the area where the node can accurately detect disaster events to the test field is calculated to obtain the spatial coverage rate, the time difference from signal emission to node triggering of the alarm is calculated to obtain the time response delay, and the ratio of the number of correct signal recognition to the total number of signal emission is calculated to obtain the parameter detection rate. Through the summary of these calculated coverage accuracy evaluation indexes, the multi-parameter fusion coverage scale is obtained. Then, according to the obtained multi-parameter fusion coverage scale, the entire underground coal mine multi-parameter collaborative monitoring field is divided into N multi-parameter coverage monitoring fields, and the size and range of each coverage monitoring field are determined based on the time response delay of the sensor and the spatial coverage rate to ensure that each new coverage monitoring field can efficiently and comprehensively monitor multiple types of monitoring, and ensure that the sensors in each monitoring field can cover the key areas in the mine. After completing the segmentation of the multi-parameter coverage monitoring field, based on the spatial distribution relationship of these monitoring fields, a spatial correlation topology is constructed, which reveals the spatial relationship between different monitoring fields, as well as the geological features and risk propagation paths of different areas in the mine, which can help to achieve more accurate geological disaster prediction and quickly identify the evolution path of the disaster when it occurs, and timely warn the mine workers to ensure the safety management of the mine.

[0039] Further, according to the spatial distribution and correlation relationship of the N multi-parameter coverage monitoring fields in the multi-parameter collaborative monitoring field, the spatial correlation topology is constructed, and the method comprises:

[0040] According to the geological structure connectivity characteristics of the N multi-parameter coverage monitoring fields in the underground space of the coal mine, a geological connectivity topology is constructed; according to the stress transmission path of the N multi-parameter coverage monitoring fields in the underground space of the coal mine, a stress transmission topology is constructed; and the geological connectivity topology and the stress transmission topology are fused to obtain the spatial correlation topology.

[0041] Optionally, in order to establish the geological connection relationship between each monitoring area in the coal mine underground, the three-dimensional geomechanical model is used to determine the stratum unit, coal seam thickness, lithological characteristics, fault and joint structure information of each monitoring site, and then N multi-parameter coverage monitoring fields are taken as nodes, and their spatial positions (XYZ coordinates) and coverage ranges are marked. Subsequently, using the region growing algorithm, taking the monitoring field node as the starting point, searching for the connected region along the extension direction of the structural surface (such as the strike of the fault), and calculating the ratio of the extension length of the structural surface to the spacing of the monitoring field, the ratio of the opening degree of the structural surface to the block size of the rock mass, and multiplying the two ratios to obtain the structural surface connectivity index, wherein the opening degree of the structural surface refers to the separation distance of the rock blocks on both sides of the fracture surface (such as fault, joint, fissure) in the rock mass, which can be obtained by subtracting the initial undisturbed displacement from the relative displacement of the rock mass on both sides of the structural surface; the block size of the rock mass refers to the size distribution characteristics of the independent block formed after the rock mass is cut by the structural surface, which can be obtained by summing the product of the particle size classification and the mass percentage of all blocks, and then dividing the sum by 100. When the structural surface connectivity index is greater than the connectivity threshold, it indicates that the structural surface is connected, at this time, the monitoring field is taken as the node of the graph, the connection relationship is taken as the edge, and the weight of the edge is the connectivity index, thereby generating a geological connection topology graph, which can be used to analyze the risk transmission channel under the geological conditions. In addition, based on the three-dimensional geomechanical model, the stress tensor and principal stress direction of each monitoring field are extracted, and a stress influence radius is set for each multi-parameter monitoring field, i.e. the spatial range that may be affected by stress change. If there is an overlapping area between the stress fields of two monitoring fields, it is considered that they have a stress transmission relationship, at this time, the monitoring field is taken as the node, the transmission relationship is taken as the edge, and the weight of the edge is the ratio of the stress difference to the spacing of the monitoring field, thereby generating a stress transmission topology graph, which is used to reveal the propagation logic and influence path of the dynamic stress underground. Finally, the geological connection topology graph and the stress transmission topology graph are aligned, fused through node consistency (i.e. the same monitoring field), and the comprehensive strength of risk propagation is represented by weighted average, thereby generating a spatial correlation topology graph, which reflects the comprehensive correlation between each monitoring area in the coal mine underground, and can be used to dynamically simulate the diffusion path of risk between monitoring fields, providing scientific data support and logical foundation for coal mine underground geological disaster warning.

[0042] Further, the standard disaster identification node includes a microseismic activity anomaly detection channel, a gas concentration gradient warning channel, a stress mutation identification channel, and a comprehensive risk decision channel, wherein the microseismic activity anomaly detection channel, the gas concentration gradient warning channel, and the stress mutation identification channel are connected in parallel, and the output ends of the microseismic activity anomaly detection channel, the gas concentration gradient warning channel, and the stress mutation identification channel are merged into the comprehensive risk decision channel.

[0043] Optionally, the standard disaster identification node internally comprises a microseismic activity anomaly detection channel, a gas concentration gradient early warning channel, a stress mutation identification channel and a comprehensive risk decision channel, wherein the microseismic activity anomaly detection channel, the gas concentration gradient early warning channel and the stress mutation identification channel exist in a parallel manner, so that these channels can simultaneously analyze the received data, reducing the time consumed by analysis. Moreover, the output ends of the microseismic activity anomaly detection channel, the gas concentration gradient early warning channel and the stress mutation identification channel all flow into the comprehensive risk decision channel, so that when the microseismic activity anomaly detection channel, the gas concentration gradient early warning channel and the stress mutation identification channel do not identify deviations, the comprehensive risk decision channel can perform comprehensive analysis according to the real-time rock mass fracture dynamics sequence, the real-time gas migration trend field and the real-time geostress evolution tensor field received from the three channels, further identifying geological disasters.

[0044] Further, the early warning management and control platform receives and according to the spatial correlation topology of the N multi-parameter coverage monitoring fields, performs disaster evolution prediction on the N geological disaster prediction results, and outputs real-time geological disaster early warning, the method comprising:

[0045] According to the spatial correlation topology and the N geological disaster prediction results, the topologically adjacent multi-parameter coverage monitoring fields are subjected to disaster risk coupling superposition of the same type of prediction results to obtain N corrected disaster prediction results; the N corrected disaster prediction results are subjected to disaster conduction space path evolution analysis to output a continuous high-risk unit chain; and the continuous high-risk unit chain is matched with a continuous risk level according to the monitoring field span, wherein the continuous high-risk unit chain and the continuous risk level constitute the real-time geological disaster early warning.

[0046] Optionally, according to the spatial correlation topology, the adjacent matrix is used to mark the topological adjacent nodes, the weight in the adjacent node is calculated by ratio with the total weight of all adjacent monitoring fields, the superposition weight of the adjacent monitoring field is obtained, the superposition (weighted calculation) of the geological disaster prediction results of the same disaster type in the adjacent monitoring field is coupled using the superposition weight of the adjacent monitoring field, and N modified disaster prediction results are obtained. Then, from the N modified disaster prediction results, a plurality of monitoring fields with the highest risk situation level are selected as seed nodes in the disaster conduction space path evolution analysis, and then the seed nodes are used as the starting point to traverse the adjacent nodes (preferably the nodes with high edge weight) layer by layer, and the path length and covered nodes are recorded. Subsequently, the paths in which all the nodes in the path have a modified risk situation level greater than or equal to a preset risk level are retained, and low-risk nodes are removed, and then the collinear paths (such as A→B→C and A→B→D, which are collinear in the A→B segment) are merged to reduce redundancy, thereby obtaining a continuous high-risk unit chain, which is composed of topologically adjacent monitoring fields with a modified risk situation level greater than or equal to a preset risk level. Then, the chain length of the continuous high-risk unit chain is compared with the risk level mapping table to match the continuous risk level corresponding to the current chain length, which will be combined with the continuous high-risk unit chain as the real-time geological disaster warning result and pushed to the ground command center and underground operation personnel to ensure the safety of the mine operation personnel.

[0047] In summary, the embodiments of the present application have at least the following technical effects:

[0048] Firstly, a plurality of multi-modal sensing devices are arranged in the coal mine underground space to construct a multi-parameter collaborative monitoring field, wherein the multi-parameter collaborative monitoring field is composed of N multi-parameter coverage monitoring fields; subsequently, N geological disaster identification nodes are locally deployed in the N multi-parameter coverage monitoring fields, wherein the N geological disaster identification nodes establish bidirectional real-time communication with the early warning management and control center; further, the N geological disaster identification nodes are collaboratively operated to perform risk identification on the N local area multi-modal data returned by the N multi-parameter coverage monitoring fields, and output N geological disaster prediction results; finally, the early warning management and control center receives and performs disaster evolution prediction on the N geological disaster prediction results according to the spatial correlation topology of the N multi-parameter coverage monitoring fields, and outputs real-time geological disaster warning. The technical problems of real-time collaborative processing of multi-source data and insufficient disaster evolution prediction capability in coal mine underground geological disaster warning are solved, and the technical effects of dynamic analysis of disaster risk in the coal mine underground by multi-modal sensor collaborative monitoring, combined with time series data analysis and collaborative work of geological disaster identification nodes, and improvement of the accuracy and response speed of geological disaster warning are achieved.

[0049] Embodiment two, based on the same inventive concept as the coal mine underground geological disaster warning method in the foregoing embodiments, asFigure 2 As shown, the present application provides a coal mine underground geological disaster early warning system, wherein the system comprises: a monitoring field construction module 11: a multi-parameter collaborative monitoring field is constructed by deploying multi-modal sensing devices in the coal mine underground space, wherein the multi-parameter collaborative monitoring field is composed of N multi-parameter coverage monitoring fields; a node deployment module 12: N geological disaster identification nodes are locally deployed for the N multi-parameter coverage monitoring fields, wherein the N geological disaster identification nodes establish bidirectional real-time communication with the early warning management and control center; a risk identification module 13: the N local area multi-modal data returned by the N multi-parameter coverage monitoring fields are risk identified by cooperatively operating the N geological disaster identification nodes, and N geological disaster prediction results are output; a disaster prediction module 14: the early warning management and control center receives and executes disaster evolution prediction on the N geological disaster prediction results according to the spatial correlation topology of the N multi-parameter coverage monitoring fields, and outputs real-time geological disaster early warning.

[0050] Further, the monitoring field construction module 11 is used to execute the following method:

[0051] A three-dimensional geomechanical model of the coal mine underground space is constructed, and the working face of the three-dimensional geomechanical model is divided into a plurality of twin grids based on a dynamic risk weight; based on the effective detection radius and directional attenuation characteristics of the microseismic sensor, an initial sensing arrangement array is selected in the plurality of twin grids; the initial sensing arrangement array is subjected to engineering constraint optimization to compensate for sensor arrangement deviation, thereby obtaining a microseismic sensing arrangement array; by analogy, a gas concentration sensing arrangement array and a stress monitoring sensor array are analyzed and output; the multi-modal sensing device is deployed in the coal mine underground space according to the microseismic sensing arrangement array, the gas concentration sensing arrangement array and the stress monitoring sensor array, thereby constructing the multi-parameter collaborative monitoring field.

[0052] Further, the node deployment module 12 is used to execute the following method:

[0053] After the first multi-parameter coverage monitoring field is locally called for a risk event, a risk deviation threshold is aggregated, thereby obtaining a dynamic energy baseline, a concentration mutation threshold and a structure instability critical value; the dynamic energy baseline, the concentration mutation threshold and the structure instability critical value are excluded from the local risk event, thereby obtaining a plurality of sample risk data, wherein the sample risk data includes a sample rock mass fracture dynamics sequence, a sample gas migration trend field, a sample local stress evolution tensor field and a sample risk situation grade; the dynamic energy baseline, the concentration mutation threshold and the structure instability critical value are mapped and loaded to the microseismic activity anomaly detection channel, the gas concentration gradient early warning channel and the stress mutation identification channel as a single-dimensional risk judgment condition; the plurality of sample risk data are used as training data to model parameterize the comprehensive risk decision channel, thereby completing the localization of the standard disaster identification node, and obtaining a first geological disaster identification node.

[0054] Further, the risk identification module 13 is used to execute the following method:

[0055] The multi-index mean of the first local multi-modal data is solved, and a first real-time rock mass fracture dynamics sequence, a first real-time gas migration trend field and a first real-time geostress evolution tensor field are output; the first real-time rock mass fracture dynamics sequence, the first real-time gas migration trend field and the first real-time geostress evolution tensor field are loaded to the microseismic activity anomaly detection channel, the gas concentration gradient early warning channel and the stress mutation identification channel of the first geological disaster identification node, and a single-dimensional comparison result group is output; if any result of the single-dimensional comparison result group is an index deviation, the deviated data is directly output as a first geological disaster prediction result; if all the single-dimensional comparison results are passed, the comprehensive risk decision channel is activated to perform risk fusion analysis on the first real-time rock mass fracture dynamics sequence, the first real-time gas migration trend field and the first real-time geostress evolution tensor field, and the first geological disaster prediction result is output.

[0056] Further, the disaster prediction module 14 is used to execute the following method:

[0057] Before the edge computing node is deployed, an effective coverage precision test is performed on the standard disaster identification node, and a multi-parameter fusion coverage scale is obtained; based on the multi-parameter fusion coverage scale, the multi-parameter collaborative monitoring field is divided into the N multi-parameter coverage monitoring fields; according to the spatial distribution correlation relationship of the N multi-parameter coverage monitoring fields in the multi-parameter collaborative monitoring field, the spatial correlation topology is constructed.

[0058] Further, the disaster prediction module 14 is used to execute the following method:

[0059] According to the geological structure connectivity characteristics of the N multi-parameter coverage monitoring fields in the underground space of the coal mine, a geological connectivity topology is constructed; according to the stress transmission path of the N multi-parameter coverage monitoring fields in the underground space of the coal mine, a stress transmission topology is constructed; and the geological connectivity topology and the stress transmission topology are fused to obtain the spatial correlation topology.

[0060] Further, the disaster prediction module 14 is used to execute the following method:

[0061] The standard disaster identification node includes a microseismic activity anomaly detection channel, a gas concentration gradient early warning channel, a stress mutation identification channel and a comprehensive risk decision channel, wherein the microseismic activity anomaly detection channel, the gas concentration gradient early warning channel and the stress mutation identification channel are connected in parallel, and the output ends of the microseismic activity anomaly detection channel, the gas concentration gradient early warning channel and the stress mutation identification channel are connected to the comprehensive risk decision channel.

[0062] Further, the disaster prediction module 14 is configured to execute the following method:

[0063] According to the spatial correlation topology and the N geological disaster prediction results, disaster risk coupling superposition of the same type prediction results is performed on the topological adjacent multi-parameter coverage monitoring field to obtain N modified disaster prediction results; disaster conduction space path evolution analysis is performed on the N modified disaster prediction results to output a continuous high-risk unit chain; and the continuous high-risk unit chain is matched with a continuous risk level according to the monitoring field span, wherein the continuous high-risk unit chain and the continuous risk level constitute the real-time geological disaster warning.

[0064] In the third embodiment, based on the same inventive concept as the coal mine underground geological disaster warning method in the foregoing embodiments, the application provides a medium, and the medium stores a computer program. When the processor executes the computer program, the following steps are implemented: a multi-parameter collaborative monitoring field is constructed by arranging multi-modal sensing devices in a coal mine underground space, wherein the multi-parameter collaborative monitoring field is composed of N multi-parameter coverage monitoring fields; N geological disaster identification nodes are locally deployed on the N multi-parameter coverage monitoring fields, wherein the N geological disaster identification nodes establish bidirectional real-time communication with a warning control center; the N geological disaster identification nodes are collaboratively operated to perform risk identification on N local multi-modal data returned by the N multi-parameter coverage monitoring fields, and output N geological disaster prediction results; and the warning control center receives the N geological disaster prediction results and performs disaster evolution prediction on the N geological disaster prediction results according to the spatial correlation topology of the N multi-parameter coverage monitoring fields, and outputs a real-time geological disaster warning.

[0065] It should be noted that the above-mentioned sequence of the embodiments of the application is only for description, and does not represent the advantages and disadvantages of the embodiments. The above describes a specific embodiment of the present application. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multi-task processing and parallel processing are possible or can be advantageous.

[0066] The above only describes the preferred embodiments of the application and does not limit the application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the application shall be included in the protection scope of the application.

[0067] The specification and drawings are merely exemplary of the application, and any and all modifications, variations, combinations or equivalents that are within the scope of the application are considered to be covered by the application. Obviously, those skilled in the art can make various modifications and variations to the application without departing from the scope of the application. Thus, if these modifications and variations of the application belong to the scope of the application and its equivalents, the application is intended to include these modifications and variations.

Claims

1. A coal mine underground geological disaster early warning method, characterized in that, The method comprises: By arranging multi-modal sensing devices in the underground space of the coal mine, a multi-parameter collaborative monitoring field is constructed, wherein the multi-parameter collaborative monitoring field is composed of N multi-parameter coverage monitoring fields; N geological disaster identification nodes are locally deployed for the N multi-parameter coverage monitoring fields, wherein the N geological disaster identification nodes establish bidirectional real-time communication with the early warning management and control center; The N geological disaster identification nodes are collaboratively operated to perform risk identification on the N local area multi-modal data returned by the N multi-parameter coverage monitoring fields, and output N geological disaster prediction results; The early warning management and control center receives and, according to the spatial correlation topology of the N multi-parameter coverage monitoring fields, performs disaster evolution prediction on the N geological disaster prediction results, and outputs real-time geological disaster warning.

2. The coal mine underground geological disaster early warning method of claim 1, wherein, By arranging multi-modal sensing devices in the underground space of the coal mine, a multi-parameter collaborative monitoring field is constructed, and the method comprises: A three-dimensional geomechanical model of the underground space of the coal mine is constructed, and the working face of the three-dimensional geomechanical model is divided into a plurality of twin grids based on dynamic risk weights; Based on the effective detection radius and directional attenuation characteristics of the microseismic sensor, an initial sensing arrangement array is selected in the plurality of twin grids; The initial sensing arrangement array is subjected to engineering constraint optimization to compensate for the sensing arrangement deviation, and a microseismic sensing arrangement array is obtained; By analogy, a gas concentration sensing arrangement array and a stress monitoring sensing array are analyzed and output; According to the microseismic sensing arrangement array, the gas concentration sensing arrangement array and the stress monitoring sensing array, multi-modal sensing device deployment is performed in the underground space of the coal mine, and the multi-parameter collaborative monitoring field is constructed and generated.

3. The coal mine underground geological disaster early warning method of claim 2, further comprising: Before deployment of the edge computing node, an effective coverage precision test is performed on the standard disaster identification node to obtain a multi-parameter fusion coverage scale; Based on the multi-parameter fusion coverage scale, the multi-parameter collaborative monitoring field is divided into the N multi-parameter coverage monitoring fields; According to the spatial distribution correlation relationship of the N multi-parameter coverage monitoring fields in the multi-parameter collaborative monitoring field, the spatial correlation topology is constructed.

4. The coal mine underground geological disaster early warning method of claim 3, wherein, According to the spatial distribution correlation relationship of the N multi-parameter coverage monitoring fields in the multi-parameter collaborative monitoring field, the spatial correlation topology is constructed, and the method comprises: According to the geological structure connectivity characteristics of the N multi-parameter coverage monitoring fields in the underground space of the coal mine, a geological connectivity topology is constructed; According to the stress transmission path of the N multi-parameter coverage monitoring fields in the underground space of the coal mine, a stress transmission topology is constructed; The geological connectivity topology and the stress transmission topology are fused to obtain the spatial correlation topology.

5. The coal mine underground geological disaster early warning method of claim 3, wherein, The standard disaster identification node comprises a microseismic activity anomaly detection channel, a gas concentration gradient early warning channel, a stress mutation identification channel and a comprehensive risk decision channel, wherein the microseismic activity anomaly detection channel, the gas concentration gradient early warning channel and the stress mutation identification channel are connected in parallel, and the output ends of the microseismic activity anomaly detection channel, the gas concentration gradient early warning channel and the stress mutation identification channel are connected to the comprehensive risk decision channel.

6. The coal mine underground geological disaster early warning method of claim 5, wherein, N geological disaster identification nodes are locally deployed on the N multi-parameter coverage monitoring fields, and the method comprises: After a local risk event is called on the first multi-parameter coverage monitoring field, risk deviation threshold aggregation is performed to obtain a dynamic energy baseline, a concentration mutation threshold and a structure instability critical value; The dynamic energy baseline, the concentration mutation threshold and the structure instability critical value are removed from the local risk event to obtain a plurality of sample risk data, wherein the sample risk data comprises a sample rock mass fracture dynamics sequence, a sample gas migration trend field, a sample local stress evolution tensor field and a sample risk situation grade; The dynamic energy baseline, the concentration mutation threshold and the structure instability critical value are mapped and loaded to the microseismic activity anomaly detection channel, the gas concentration gradient early warning channel and the stress mutation identification channel as a single-dimensional risk judgment condition; The plurality of sample risk data are taken as training data to perform model parameter adjustment of the comprehensive risk decision channel, and the localization of the standard disaster identification node is completed to obtain a first geological disaster identification node.

7. The coal mine underground geological disaster early warning method of claim 6, wherein, The N geological disaster identification nodes are cooperatively operated to perform risk identification on N local area multi-modal data returned by the N multi-parameter coverage monitoring fields, and N geological disaster prediction results are output, and the method comprises: Multi-index mean value solving is performed on the first local area multi-modal data to output a first real-time rock mass fracture dynamics sequence, a first real-time gas migration trend field and a first real-time local stress evolution tensor field; The first real-time rock mass fracture dynamics sequence, the first real-time gas migration trend field and the first real-time local stress evolution tensor field are loaded to the microseismic activity anomaly detection channel, the gas concentration gradient early warning channel and the stress mutation identification channel of the first geological disaster identification node to output a single-dimensional comparison result group; If any result of the single-dimensional comparison result group is an index deviation, the deviated data is directly output as a first geological disaster prediction result; If all the single-dimensional comparison results are passed, the comprehensive risk decision channel is activated to perform risk fusion analysis on the first real-time rock mass fracture dynamics sequence, the first real-time gas migration trend field and the first real-time local stress evolution tensor field, and the first geological disaster prediction result is output.

8. The coal mine underground geological disaster early warning method of claim 3, wherein, The early warning management middle station receives and, according to a spatial correlation topology of the N multi-parameter coverage monitoring fields, performs disaster evolution prediction on the N geological disaster prediction results to output real-time geological disaster early warning, and the method comprises: According to the spatial correlation topology and the N geological disaster prediction results, disaster risk coupling superposition of the same type of prediction results is performed on topologically adjacent multi-parameter coverage monitoring fields to obtain N corrected disaster prediction results; Disaster conduction space path evolution analysis is performed on the N corrected disaster prediction results to output a continuous high-risk unit chain; The monitoring field span of the continuous high-risk unit chain is matched with a continuous risk grade, wherein the continuous high-risk unit chain and the continuous risk grade constitute the real-time geological disaster early warning.

9. The coal mine underground geological disaster early warning system, characterized in that, A system for implementing the coal mine underground geological disaster early warning method of any one of claims 1-8, the system comprising: The monitoring field construction module: a multi-modal sensing device is arranged in the underground space of the coal mine to construct a multi-parameter collaborative monitoring field, wherein the multi-parameter collaborative monitoring field is composed of N multi-parameter coverage monitoring fields; The node deployment module: N geological disaster identification nodes are deployed locally in the N multi-parameter coverage monitoring fields, wherein the N geological disaster identification nodes establish bidirectional real-time communication with the early warning and control center station; The risk identification module: the N geological disaster identification nodes are cooperatively operated to identify the risks of N local area multi-modal data returned by the N multi-parameter coverage monitoring fields, and output N geological disaster prediction results; The disaster prediction module: the early warning and control center station receives and executes disaster evolution prediction on the N geological disaster prediction results according to the spatial correlation topology of the N multi-parameter coverage monitoring fields, and outputs real-time geological disaster warning.

10. A computer readable medium having stored thereon a computer program, characterized in that The program is executed by the processor to realize the coal mine underground geological disaster warning method in any one of claims 1-8.

Images