A roof collapse early warning device for real-time monitoring of mine pressure of intelligent coal mine

Abstract

The application provides a roof collapse real-time monitoring and early warning device for intelligent coal mines, relates to the technical field of mine safety monitoring, and comprises a hierarchical zoning sensing module, a ground intelligent supervision module, an industrial data transmission and convergence module and a three-dimensional visualization display module, wherein the ground intelligent supervision module comprises a core risk index calculation unit, a secondary key area risk assessment unit and a risk correlation and conduction analysis unit; the application performs space-time fusion and collaborative analysis on multi-dimensional information such as stress, displacement and acoustic emission, generates a quantitative roof comprehensive risk index, thereby realizing dynamic and accurate diagnosis of the roof stability state, and changing the traditional lagging alarm mode which relies on a single parameter threshold. Meanwhile, by introducing a risk correlation and conduction analysis mechanism, the spatial transmission precursor of the risk in the underground can be intelligently identified, a leap from isolated point alarm to global linkage early warning is realized, and the advance and reliability of early warning are improved.

Description

Technical Field

[0001] This invention relates to the field of mine safety monitoring technology, and in particular to a real-time monitoring device for mine pressure and an early warning device for roof collapse in intelligent coal mines. Background Technology

[0002] In coal mining, mine pressure manifestation and roof collapse are the main hazards threatening underground safe production. Achieving real-time and accurate monitoring of mine pressure and early warning of roof instability is a core challenge in the construction of smart mine safety. Traditional mine pressure monitoring methods mainly rely on single-type sensors, such as installing hydraulic prop pressure gauges or individual borehole stress gauges in the roadway to monitor stress changes; or setting up roof delamination meters to monitor roof subsidence and displacement.

[0003] However, the above methods have significant limitations: for example, the monitoring methods are isolated, only reflecting changes in single physical quantities such as stress or displacement. Roof instability, however, is the result of the coordinated evolution of multiple fields of information, including stress, deformation, and rock mass fracturing. A single parameter cannot comprehensively and accurately characterize the overall stability of the roof, leading to low accuracy and a high false alarm rate in early warnings. Furthermore, the monitoring methods are passive, typically triggering alarms only when data from monitoring points exceeds preset fixed thresholds. By this time, the disaster has often entered a rapid development phase, resulting in a very short warning window and the loss of valuable time for prevention and response. Moreover, existing systems lack an understanding of the spatial correlation of risks, making it impossible to determine whether an anomaly in one area indicates an increase in risk in adjacent or related areas. The isolated early warning information makes it difficult to support comprehensive risk control decisions.

[0004] In recent years, with the development of sensing technology, advanced methods such as fiber optic gratings, microseismic monitoring, and laser scanning have begun to be applied in mines. However, current research and applications mostly focus on improving the performance or density of single sensors, failing to fundamentally solve the problems of multi-source information fusion and intelligent risk assessment. Distributing high-density, multi-type advanced sensors uniformly throughout the entire mine would be extremely costly and have low engineering feasibility. Furthermore, how to effectively process and fuse massive amounts of heterogeneous monitoring data, extract characteristic information that can characterize the precursors of roof instability, and ultimately transform it into intuitive and operable early warning commands remains a technical bottleneck to be solved. Therefore, this invention proposes a real-time monitoring device for mine pressure and a roof collapse early warning device for intelligent coal mines to address the problems existing in the prior art. Summary of the Invention

[0005] To address the aforementioned problems, the present invention aims to propose a real-time monitoring device for mine pressure and roof collapse early warning for intelligent coal mines. This invention scientifically classifies underground risk areas in coal mines and implements differentiated sensor network deployment, constructing a mine pressure monitoring and roof risk prevention system that integrates zoned sensing, multi-source assessment, correlation transmission, and intelligent early warning, thereby solving the problems of existing technologies.

[0006] To achieve the objectives of this invention, the invention is implemented through the following technical solution: a real-time monitoring device for mine pressure and an early warning device for roof collapse in intelligent coal mines, comprising:

[0007] The hierarchical and zoned sensing module is used to deploy differentiated sensor networks in underground coal mines based on regional risk levels, forming a three-layer sensing network. The regional risk levels are divided into core risk areas, secondary key monitoring areas, and general monitoring areas.

[0008] The ground-based intelligent monitoring module is used to dynamically assess the risk level of different areas based on the data obtained from the three-layer sensor network, and then trigger early warning information based on the assessment results. It includes a core risk index calculation unit, a secondary key area risk assessment unit, and a risk correlation and transmission analysis unit.

[0009] Industrial data transmission aggregation module, used for data transmission between modules and data access to external systems and equipment;

[0010] The 3D visualization module is used to display collected data, comprehensive risk index, and early warning information in a three-dimensional manner.

[0011] A further improvement is that the hierarchical partitioning perception module includes:

[0012] The core risk area perception unit is used to deploy a first fiber optic stress sensing array, a lidar displacement monitoring unit, and an array-type acoustic emission sensor in the mining face, roadway intersection, and geological structure zone of underground coal mines, so as to simultaneously collect stress time series data, three-dimensional point cloud displacement data, and full waveform acoustic emission data.

[0013] The secondary key monitoring area sensing unit is used to deploy a second fiber optic grating sensor and an acoustic emission event sensor in the main preparation roadway area of ​​the coal mine to collect one-dimensional stress trend data and acoustic emission event parameter data.

[0014] The general monitoring area sensing unit is used to deploy simple delamination instruments in the stable roadway areas of underground coal mines to collect roof delamination displacement alarm data.

[0015] A further improvement is made in that: the core risk index calculation unit generates a first comprehensive risk index for the roof based on the acquired data, and the specific steps are as follows:

[0016] S0: Stress time-series data, roof subsidence velocity data, and acoustic emission event data from the same core risk area are spatiotemporally aligned and matched based on a unified timescale.

[0017] S1: Based on the stress time series data, calculate the stress change gradient and stress concentration coefficient at each measuring point in the monitoring area, and generate a stress field deterioration index;

[0018] S2: Based on the top plate settlement velocity data, calculate the settlement acceleration, identify the delamination and crack propagation rates, and generate displacement field instability index.

[0019] S3: Based on the full waveform acoustic emission data, calculate the cumulative event energy per unit time, and statistically analyze the b-value and its changing trend of the acoustic emission event energy-frequency relationship to generate a rupture field activity index;

[0020] S4: Normalize the stress field deterioration index, displacement field instability index, and fracture field activity index to a preset scale, and perform weighted summation using a weighted summation method to output the first roof comprehensive risk index. Then, based on the statistical characteristics of the first roof comprehensive risk index and historical disaster data of the mining area, dynamically associate a set of increasing three-level early warning thresholds for the index.

[0021] A further improvement is made in that: the secondary key area risk assessment unit generates a second risk level identifier based on the acquired data, and the specific steps are as follows:

[0022] A1: Extract the stress change rate of key measuring points from one-dimensional stress trend data, and then extract the cumulative energy of the event from acoustic emission event parameter data;

[0023] A2: Establish a time series of stress change rate and cumulative energy of events, and analyze its abrupt change trend;

[0024] A3: Establish a correlation model between parameter trends and risk levels based on historical data of core risk areas;

[0025] A4: Compare the trend with the correlation model, and based on the comparison results, output a discrete second risk level indicator representing the degree of risk increase.

[0026] A further improvement is that the risk correlation and transmission analysis unit is used to perform risk transmission trigger analysis and risk correlation determination, wherein the specific method of risk transmission trigger analysis is as follows:

[0027] When roof delamination displacement alarm data is received, starting from the alarm location, the secondary key monitoring area and core risk area in the upstream direction are determined as the area to be verified based on the mine roadway topology. Then, the historical monitoring data of the area to be verified within the first preset time before the alarm time are retrieved, and the changing trend of key parameters in the historical monitoring data is analyzed. If the changing trend shows that the risk level has been rising before the alarm time, a preliminary confirmation signal of risk transmission is generated.

[0028] The specific methods for determining risk correlation are as follows:

[0029] The system continuously receives the second risk level indicator. When the second risk level indicator reaches the preset increase threshold, the corresponding secondary key monitoring area is determined to be a risk active area. Then, based on the mine roadway topology and mine pressure propagation law, the adjacent core risk area within the mining disturbance influence range of the risk active area is determined, and a risk advance warning signal is generated.

[0030] A further improvement is that the ground-based intelligent monitoring module also includes a tiered early warning triggering unit, the triggering mechanism of which is as follows:

[0031] When the comprehensive risk index of the first roof exceeds the current level threshold in its dynamically associated three-level early warning threshold, a corresponding early warning instruction is generated. This early warning instruction includes a unique identifier of the early warning target area, an early warning level identifier determined based on the threshold exceeded, and a corresponding early warning action.

[0032] When a risk shift warning signal is received, an extended warning covering the core risk area and secondary key monitoring area indicated by the signal is triggered.

[0033] When a preliminary confirmation signal of risk transmission is received from the risk association and transmission analysis unit, a system-level inspection alarm is triggered.

[0034] The further improvement is that the warning action is divided into three groups, wherein the first warning action is to mark the corresponding core risk area as the first concern state in the three-dimensional visualization display module;

[0035] The second early warning action is to mark the corresponding core risk area as the second alert state in the 3D visualization display module and activate the local audible and visual alarm in that area.

[0036] The third early warning action involves marking the corresponding core risk area as the third dangerous state in the 3D visualization module, activating the local audible and visual alarm in that area, and sending a suggested instruction to evacuate personnel and lock down equipment in the area.

[0037] A further improvement is that the 3D visualization module includes a 3D model management unit, which is used to load, maintain, and draw 3D models based on mine geology and roadway data;

[0038] The risk index visualization unit is used to render and display the first roof comprehensive risk index of each core risk zone in the form of a heat map in the three-dimensional tunnel model.

[0039] The risk level visualization unit is used to display the second risk level identifier of each key monitoring area in the form of isosurfaces in the three-dimensional tunnel model;

[0040] The alarm point visualization unit is used to mark the location of alarm points issued by the general monitoring area in the three-dimensional tunnel model with a highlighted flashing icon;

[0041] A risk conduction visualization unit, configured to dynamically draw and mark the risk conduction path calculated by the risk association and conduction analysis unit in a three-dimensional roadway model.

[0042] Beneficial effects of the present invention are as follows: the present invention performs spatiotemporal fusion and collaborative analysis on multi-dimensional information such as stress, displacement and acoustic emission, and generates a quantitative comprehensive roof risk index, thereby realizing dynamic and accurate diagnosis of roof stability status, and changing the traditional delayed alarm mode that relies on a single parameter threshold. Meanwhile, by introducing a risk association and conduction analysis mechanism, the present invention can intelligently identify precursors of spatial risk transmission underground, realize the leap from isolated point alarming to global coordinated early warning, and improve the advance time and reliability of early warning. In addition, by adopting the hierarchical and zonal arrangement, the overall deployment and maintenance cost of the device is greatly reduced on the premise of ensuring high-precision monitoring of the core area. Finally, complex multi-source data and evaluation results are presented in an integrated and intuitive manner through a three-dimensional visualization platform, which comprehensively improves the active prevention and control level and intelligent management efficiency of coal mine roof disasters. Description of Drawings

[0043] Figure 1 is a schematic flow diagram of the present invention. Detailed Description

[0044] In order to deepen the understanding of the present invention, the present invention will be further detailed below with reference to embodiments. The embodiments are only used to explain the present invention, and do not constitute a limitation to the protection scope of the present invention.

[0045] According to Figure 1 as shown, this embodiment proposes a real-time mine pressure monitoring and roof collapse early warning device for intelligent coal mines, which includes a hierarchically and zonally arranged perception module deployed underground, an industrial data transmission and aggregation module covering the underground and aboveground areas, a ground intelligent supervision module located in a ground monitoring center, and a three-dimensional visual display module.

[0046] For the hierarchically and zonally arranged perception module, a three-layer sensing network layout is implemented according to the underground spatial positions and risk levels. Specifically, it includes:

[0047] a core risk area perception unit, configured to arrange a first fiber Bragg grating stress sensing array, a laser radar displacement monitoring unit and an array acoustic emission sensor in mining working faces, roadway intersections and geological structure areas underground coal mines, so as to synchronously collect stress time-series data, three-dimensional point cloud displacement data and full-waveform acoustic emission data;

[0048] a secondary key monitoring area perception unit, configured to arrange a second fiber Bragg grating sensor and an acoustic emission event sensor in main preparation roadway areas underground coal mines, so as to collect one-dimensional stress trend data and acoustic emission event parameter data;

[0049] The general monitoring area sensing unit is used to deploy simple delamination instruments in the stable roadway areas of underground coal mines to collect roof delamination displacement alarm data.

[0050] For the core risk area sensing units, they are deployed in high-risk areas such as the fully mechanized mining face, roadway intersections, and geological structural zones. Fiber Bragg grating (FBG) stress sensors are installed in a 10m×10m grid on the roof and sidewalls to form an array, monitoring stress (MPa) and its rate of change in real time. An explosion-proof two-dimensional lidar combined with a high-definition infrared camera is deployed every 30 meters to obtain roof subsidence velocity (mm / h) and delamination crack information through point cloud data analysis. An array of acoustic emission (AE) sensors is deployed in key rock strata to capture full-waveform acoustic emission signals for calculating event energy and b-values. All data undergoes preliminary processing and packaging at an intrinsically safe data acquisition substation.

[0051] The secondary key monitoring area sensing units are deployed in main transport and return air roadways. They employ linear series FBG sensors, deployed at 20-meter intervals along the centerline of the roadway roof to monitor stress trends. Acoustic emission event sensors are also used, deployed at 50-meter intervals to monitor event rate and cumulative energy. The acquired data is accessed through an intrinsically safe data acquisition substation.

[0052] Finally, the sensing units in the general monitoring area are deployed in stable mining roadways, chambers, and other similar areas. Mechanical roof separation meters are used, installed at 100-meter intervals. When the roof separation value exceeds a set threshold (10mm), an alarm is generated. The data can be accessed via low-power IoT nodes or manual inspection terminals.

[0053] The industrial data transmission aggregation module is used for data transmission between modules within the device and for data access to external systems and equipment. It uses an industrial gigabit Ethernet ring network as the backbone, supplemented by Wi-Fi 6 Mesh networks in key areas. All data acquisition substations connect to the network via wired or wireless means and are equipped with a precise clock synchronization protocol to ensure the spatiotemporal synchronization of all sensed data. Correspondingly, the data is aggregated to a ground-based industrial server and stored in real-time and historical databases for easy retrieval later.

[0054] The ground-based intelligent monitoring module runs on a high-performance ground server cluster. It is used to dynamically assess the risk level of different areas based on data acquired from a three-layer sensor network, and then trigger early warning information based on the assessment results. It includes a core risk index calculation unit, a secondary key area risk assessment unit, and a risk correlation and transmission analysis unit.

[0055] Specifically, the core risk index calculation unit generates the first comprehensive risk index of the roof based on the acquired data. The specific steps are as follows:

[0056] S0: Retrieve stress time series data, roof subsidence velocity data and acoustic emission event list data of the same core risk area within the same time window from the real-time database, and align them with time and space coordinates.

[0057] S1: Based on the stress time series data, calculate the average gradient of stress change at all FBG measuring points in the region over the past hour, and identify the maximum stress concentration factor (the ratio of local stress to regional average stress). By weighting, obtain an index value of 0 to 100, which is the stress field deterioration index.

[0058] S2: Based on the top plate sinking velocity data, calculate the sinking acceleration, and combine it with the crack propagation rate from image recognition. Then, obtain an index value of 0 to 100 through a normalization model. This index value is the displacement field instability index.

[0059] S3: Based on the full waveform acoustic emission data, calculate the cumulative energy of events per unit time, and statistically analyze the b-value and its changing trend of the acoustic emission event energy-frequency relationship to generate a rupture field activity index. That is, statistically analyze the cumulative energy of acoustic emission events in the past 10 minutes, and calculate the rate of change of the b-value (the slope of the event energy-frequency distribution curve) in this period relative to the previous period, and comprehensively obtain an index value of 0 to 100.

[0060] S4: The stress field deterioration index, displacement field instability index, and fracture field activity index are normalized to a preset scale, and a weighted summation method is used to output the first comprehensive roof risk index. Then, based on the statistical characteristics of the first comprehensive roof risk index and historical disaster data of the mining area, a set of increasing three-level early warning thresholds are dynamically associated with this index. Specifically, the entropy weight method is used to dynamically determine the weights of the three indicators of stress, displacement, and acoustic emission (e.g., 0.4, 0.3, and 0.3 respectively). After normalizing the three indicators, a weighted summation is performed (the index values ​​from 0 to 100 are the normalized results) to obtain the first comprehensive roof risk index (R) from 0 to 100. Based on the historical monitoring data and microseismic event records of the mining area over the past year, the R value is divided into three intervals through statistical distribution: R < 30 is low risk, 30 ≤ R < 50 is medium risk, and R ≥ 50 is high risk. Correspondingly, the first early warning threshold is set to 30, the second early warning threshold to 50, and the third early warning threshold to 70.

[0061] The secondary risk assessment unit generates a second risk level label based on the acquired data. The specific steps are as follows:

[0062] A1: Extract the stress change rate of key measuring points from one-dimensional stress trend data, that is, extract the stress change rate of key measuring points in the past 30 minutes from the linear FBG data of the secondary key area, and then extract the cumulative energy of the event (past 30 minutes) from the acoustic emission event parameter data.

[0063] A2: Establish a 5-minute moving average sequence for the two parameters mentioned above. When both parameters show an increase of more than 10% over three consecutive periods, it is determined to be a "coordinated upward trend".

[0064] A3: Establish a correlation model between parameter trends and risk levels based on historical data of core risk areas;

[0065] A4: Compare the trend with the correlation model, and based on the comparison results, output a discrete second risk level label representing the degree of risk increase. That is, input the trend characteristics obtained from the current analysis into the model, and the model outputs a discrete second risk level label, which is L1-stable, L2-concerned, and L3-increased.

[0066] The risk correlation and transmission analysis unit is used to perform risk transmission trigger analysis and risk correlation determination. The specific method of risk transmission trigger analysis is as follows:

[0067] When a roof delamination displacement alarm is received, the secondary key monitoring area and core risk area upstream of the alarm location are determined as the areas to be verified based on the mine roadway topology. Then, historical monitoring data of the areas to be verified within the first preset time period before the alarm time are retrieved. The changing trends of key parameters in the historical monitoring data (the temporal change characteristics of the second risk level identifier from the secondary key monitoring area sensing unit and the first roof comprehensive risk index from the core risk area sensing unit) are analyzed. If the changing trend shows that the risk level has been rising before the alarm time, a preliminary confirmation signal of risk transmission is generated.

[0068] Specifically, when an alarm from a stratification analyzer is received in a general monitoring area (e.g., G-05), the system uses the location of G-05 as the starting point and, based on the roadway topology digital map, automatically determines the secondary key monitoring section (e.g., S-02) and the core risk area (e.g., C-01) within a 50-meter range upstream. Historical data (second risk level identifier and first roof comprehensive risk index) for S-02 and C-01 over the past two hours are retrieved. Analysis reveals that one hour before the alarm, the identifier of S-02 changed from "L1" to "L2," and the R index of C-01 rose from 25 to 38. Therefore, it is determined that the upstream risk had already increased before the alarm, generating a preliminary confirmation signal of risk transmission, along with the location information of G-05, S-02, and C-01.

[0069] The specific methods for determining risk correlation are as follows:

[0070] The system continuously receives the second risk level indicator. When the second risk level indicator reaches the preset increase threshold, the corresponding secondary key monitoring area is determined to be a risk active area. Then, based on the mine roadway topology and mine pressure propagation law, the adjacent core risk area within the mining disturbance influence range of the risk active area is determined, and a risk advance warning signal is generated.

[0071] Specifically, all secondary critical area markers are continuously monitored. When the marker for secondary critical area S-03 remains "L3-Rising" for more than 15 minutes, based on the mine pressure propagation model (e.g., based on mining direction and rock strata structure), it is determined that the stress may be transferring to the adjacent core area C-02. A risk advance warning signal is then generated, containing the information: "Active Risk Area: S-03; Potential Impact Core Area: C-02".

[0072] The ground-based intelligent monitoring module also includes a tiered early warning triggering unit, which contains three sets of triggering mechanisms, including:

[0073] The first triggering mechanism is used for tiered early warning in the core area: when the comprehensive risk index of the first roof exceeds the current level threshold of its dynamically associated three-level early warning threshold, a corresponding early warning instruction is generated. This early warning instruction includes a unique identifier of the target area, an early warning level identifier determined according to the exceeded threshold (Level 1, Level 2, and Level 3, corresponding to the three early warning thresholds respectively), and a corresponding early warning action. Specifically, if the R index of the core area (e.g., number C-03) rises to 35 (exceeding the first early warning threshold of 30), an early warning instruction is generated: {Target Area: C-03, Early Warning Level: Level 1, Action: Mark as Attention};

[0074] If the R index of C-03 rises to 55 (exceeding the second warning threshold of 50), a warning instruction is generated: {Target area: C-03, Warning level: Level 2, Action: Mark as alert state, activate the C-03 area audible and visual alarm};

[0075] If the R index of C-03 rises to 75 (exceeding the third warning threshold of 70), a warning instruction will be generated: {Target area: C-03, Warning level: Level 3, Action: Mark as dangerous state, activate alarm, and send the instruction "Recommend evacuating personnel from the C-03 area and locking equipment" to the dispatch center}.

[0076] The second triggering mechanism is used for extended early warning: when a risk shift warning signal is received, an extended early warning covering the core risk area and secondary key monitoring area indicated by the signal is triggered. Specifically, upon receiving a risk shift warning signal regarding "S-03 affecting C-02", an extended early warning instruction is immediately generated, stating: "Attention: The risk in secondary key area S-03 has increased and may affect core area C-02. Please strengthen monitoring of C-02 and S-03."

[0077] The third triggering mechanism is used for system inspection alarms: When a preliminary confirmation signal of risk transmission regarding "G-05 alarm transmission to S-02 and C-01" is received, a system-level inspection alarm is triggered, which reads: "General area G-05 alarm, the risk in the upstream S-02 and C-01 areas has increased in the early stage, and the risk transmission is suspected. Please immediately send personnel to inspect the S-02 and C-01 areas."

[0078] The 3D visualization module includes a 3D model management unit, which is used to load, maintain, and draw 3D models based on mine geology and roadway data, serving as the visualization base.

[0079] The risk index visualization unit is used to render and display the first comprehensive roof risk index of each core risk zone in the form of a heatmap within a 3D tunnel model. It reads the R-index of each core zone from the database and overlays a semi-transparent heatmap onto the corresponding area in the 3D model. The color gradually changes from blue (low risk) to red (high risk). When a warning instruction is received for a specific core zone, this unit dynamically changes the visual style of the heatmap for that area according to the warning level in the instruction. For example, at level one, the heatmap outline flashes yellow; at level two, it flashes rapidly orange; and at level three, it turns red with a pulsed halo.

[0080] The risk level visualization unit is used to display the second risk level identifiers of each secondary key monitoring area in the form of isosurfaces within a 3D tunnel model. Specifically, it reads the second risk level identifiers of each secondary key area from the database and renders them as isosurfaces of different colors (e.g., green, yellow, orange) in the 3D model. When a risk advance warning signal is received, this unit draws a highlighted red arrow connecting the secondary key area (e.g., S-03) and the core area (e.g., C-02) indicated by the signal.

[0081] The alarm point visualization unit is used to mark the locations of alarm points issued by the general monitoring area with bright, flashing icons in the 3D tunnel model. For example, the alarm location of the separation instrument in the general monitoring area (number G-05) is marked as a bright, flashing red icon in the 3D model.

[0082] The risk transmission visualization unit is used to dynamically draw and mark the risk transmission path calculated by the risk association and transmission analysis unit in the 3D tunnel model. Specifically, based on the calculation results of the risk association and transmission analysis unit, the risk transmission path is dynamically drawn in the 3D model (such as a dashed arrow from G-05 to S-02 and then to C-01). When a preliminary confirmation signal of risk transmission is received, the unit renders the path as a bold, flashing red dashed line and highlights the areas to be verified (S-02, C-01) in the signal with a yellow outline.

[0083] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the present invention without departing from its framework and scope of application, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A real-time monitoring device for mine pressure and an early warning system for roof collapse in intelligent coal mines, characterized in that: include: The hierarchical and zoned sensing module is used to deploy differentiated sensor networks in underground coal mines based on regional risk levels, forming a three-layer sensing network. The regional risk levels are divided into core risk areas, secondary key monitoring areas, and general monitoring areas. The ground-based intelligent monitoring module is used to dynamically assess the risk level of different areas based on the data obtained from the three-layer sensor network, and then trigger early warning information based on the assessment results. It includes a core risk index calculation unit, a secondary key area risk assessment unit, and a risk correlation and transmission analysis unit. Industrial data transmission aggregation module, used for data transmission between modules and data access to external systems and equipment; The 3D visualization module is used to display collected data, comprehensive risk index, and early warning information in a three-dimensional manner.

2. The real-time monitoring device for mine pressure and roof collapse early warning for intelligent coal mines according to claim 1, characterized in that: The hierarchical and partitioned sensing module includes: The core risk area perception unit is used to deploy a first fiber optic stress sensing array, a lidar displacement monitoring unit, and an array-type acoustic emission sensor in the mining face, roadway intersection, and geological structure zone of underground coal mines, so as to simultaneously collect stress time series data, three-dimensional point cloud displacement data, and full waveform acoustic emission data. The secondary key monitoring area sensing unit is used to deploy a second fiber optic grating sensor and an acoustic emission event sensor in the main preparation roadway area of ​​the coal mine to collect one-dimensional stress trend data and acoustic emission event parameter data. The general monitoring area sensing unit is used to deploy simple delamination instruments in the stable roadway areas of underground coal mines to collect roof delamination displacement alarm data.

3. The real-time monitoring device for mine pressure and roof collapse early warning for intelligent coal mines according to claim 1, characterized in that: The core risk index calculation unit generates the first comprehensive roof risk index based on the acquired data. The specific steps are as follows: S0: The stress time series data, roof subsidence velocity data and acoustic emission event data from the same core risk area are spatiotemporally aligned and matched based on a unified time scale; S1: Based on the stress time series data, calculate the stress change gradient and stress concentration coefficient at each measuring point in the monitoring area, and generate a stress field deterioration index; S2: Based on the top plate settlement velocity data, calculate the settlement acceleration, identify the delamination and crack propagation rates, and generate displacement field instability index. S3: Based on the full waveform acoustic emission data, calculate the cumulative event energy per unit time, and statistically analyze the b-value and its changing trend of the acoustic emission event energy-frequency relationship to generate a rupture field activity index; S4: Normalize the stress field deterioration index, displacement field instability index, and fracture field activity index to a preset scale, and perform weighted summation using a weighted summation method to output the first roof comprehensive risk index. Then, based on the statistical characteristics of the first roof comprehensive risk index and historical disaster data of the mining area, dynamically associate a set of increasing three-level early warning thresholds for the index.

4. The real-time monitoring device for mine pressure and roof collapse early warning for intelligent coal mines according to claim 1, characterized in that: The secondary key area risk assessment unit generates a second risk level identifier based on the acquired data. The specific steps are as follows: A1: Extract the stress change rate of key measuring points from one-dimensional stress trend data, and then extract the cumulative energy of the event from acoustic emission event parameter data; A2: Establish a time series of stress change rate and cumulative energy of events, and analyze its abrupt change trend; A3: Establish a correlation model between parameter trends and risk levels based on historical data of core risk areas; A4: Compare the trend with the correlation model, and based on the comparison results, output a discrete second risk level indicator representing the degree of risk increase.

5. A real-time monitoring device for mine pressure and roof collapse early warning for intelligent coal mines according to claim 4, characterized in that: The risk association and transmission analysis unit is used to perform risk transmission trigger analysis and risk association determination, wherein the specific method of risk transmission trigger analysis is as follows: When roof delamination displacement alarm data is received, starting from the alarm location, the secondary key monitoring area and core risk area in the upstream direction are determined as the area to be verified based on the mine roadway topology. Then, the historical monitoring data of the area to be verified within the first preset time before the alarm time are retrieved, and the changing trend of key parameters in the historical monitoring data is analyzed. If the changing trend shows that the risk level has been rising before the alarm time, a preliminary confirmation signal of risk transmission is generated. The specific methods for determining risk correlation are as follows: The system continuously receives the second risk level indicator. When the second risk level indicator reaches the preset increase threshold, the corresponding secondary key monitoring area is determined to be a risk active area. Then, based on the mine roadway topology and mine pressure propagation law, the adjacent core risk area within the mining disturbance influence range of the risk active area is determined, and a risk advance warning signal is generated.

6. The real-time monitoring device for mine pressure and roof collapse early warning for intelligent coal mines according to claim 5, characterized in that: The ground-based intelligent monitoring module also includes a tiered early warning triggering unit, the triggering mechanism of which is as follows: When the comprehensive risk index of the first roof exceeds the current level threshold in its dynamically associated three-level early warning threshold, a corresponding early warning instruction is generated. This early warning instruction includes a unique identifier of the early warning target area, an early warning level identifier determined based on the threshold exceeded, and a corresponding early warning action. When a risk shift warning signal is received, an extended warning covering the core risk area and secondary key monitoring area indicated by the signal is triggered. When a preliminary confirmation signal of risk transmission is received from the risk association and transmission analysis unit, a system-level inspection alarm is triggered.

7. A real-time monitoring device for mine pressure and roof collapse early warning for intelligent coal mines according to claim 6, characterized in that: The warning actions are divided into three groups, the first of which is to mark the corresponding core risk area as the first concern in the three-dimensional visualization display module; The second early warning action is to mark the corresponding core risk area as the second alert state in the 3D visualization display module and activate the local audible and visual alarm in that area. The third early warning action involves marking the corresponding core risk area as the third dangerous state in the 3D visualization module, activating the local audible and visual alarm in that area, and sending a suggested instruction to evacuate personnel and lock down equipment in the area.

8. A real-time monitoring device for mine pressure and roof collapse early warning for intelligent coal mines according to claim 1, characterized in that: The 3D visualization module includes a 3D model management unit, which is used to load, maintain, and draw 3D models based on mine geology and roadway data; The risk index visualization unit is used to render and display the first roof comprehensive risk index of each core risk zone in the form of a heat map in the three-dimensional tunnel model. The risk level visualization unit is used to display the second risk level identifier of each key monitoring area in the form of isosurfaces in the three-dimensional tunnel model; The alarm point visualization unit is used to mark the location of alarm points issued by the general monitoring area in the three-dimensional tunnel model with a highlighted flashing icon; The risk transmission visualization unit is used to dynamically draw and mark the risk transmission paths calculated by the risk association and transmission analysis unit in the three-dimensional tunnel model.

Images