A coal mine safety risk identification system based on edge internet of things

Abstract

The application discloses a coal mine safety risk identification system based on edge Internet of Things, which comprises a collection module, a sub-analysis module and a total analysis module; the collection module is used for collecting video data, equipment information, multi-modal environment data and personnel physical sign data in a coal mine underground, obtaining three-dimensional point clouds from the video data, converting the three-dimensional point clouds into a first three-dimensional coordinate system, generating an original three-dimensional model, and transmitting the original three-dimensional model to the total analysis module; the collection module realizes multi-source collection of the video data, the equipment information, the multi-modal environment data and the personnel physical sign data, and generates the original three-dimensional model through three-dimensional reconstruction using the video data; the total analysis module fuses a plurality of original three-dimensional models into a real-time three-dimensional model, realizes three-dimensional visualization of the environment in the coal mine underground, and enables supervisors to intuitively see the spatial structure and equipment distribution in the coal mine underground, thereby being beneficial to improving the spatial situation awareness capability in the coal mine underground.

Description

Technical Field

[0001] This invention relates to the field of coal mine safety monitoring technology, and in particular to a coal mine safety risk identification system based on edge IoT. Background Technology

[0002] The underground working environment in coal mines is complex, with various safety hazards such as gas explosions, water inrushes, and fires, posing a serious threat to the lives of miners. With the development of Internet of Things (IoT) and intelligent sensing technologies, coal mine safety monitoring systems are gradually shifting from manual inspections to automation and intelligence, with various sensors being widely used for real-time acquisition and monitoring of underground environmental parameters.

[0003] Existing coal mine safety monitoring systems typically employ distributed sensor networks, deploying gas sensors, carbon monoxide sensors, temperature sensors, and other equipment at key underground locations to achieve real-time acquisition and early warning of environmental parameters. However, most existing systems display various sensor data independently on the monitoring screen in numerical form, lacking intuitive spatial correlation. This makes it difficult for supervisors to quickly pinpoint the specific location and impact range of risk sources, hindering the formation of a global situational awareness. Furthermore, video surveillance data, equipment operation data, environmental parameter data, and personnel location and vital signs data in existing monitoring systems are often fragmented, stored in different information silos, lacking an effective data fusion mechanism. Threshold-based early warnings for single parameters are prone to false alarms due to sensor malfunctions or environmental fluctuations, while the lack of a multi-dimensional comprehensive analysis mechanism may lead to missed warnings of complex risks. Therefore, the accuracy and reliability of early warning decisions need improvement. Summary of the Invention

[0004] The technical problem solved by this invention is that existing coal mine safety monitoring systems suffer from scattered data, lack of spatial situational awareness, and inaccurate early warning decisions.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a coal mine safety risk identification system based on edge IoT, comprising: a data acquisition module, a sub-analysis module, and a total analysis module;

[0006] The acquisition module is used to acquire video data, equipment information, multimodal environmental data and personnel vital signs data in the coal mine, and to obtain a three-dimensional point cloud from the video data, convert the three-dimensional point cloud into a first three-dimensional coordinate system to generate an original three-dimensional model, transmit the original three-dimensional model to the main analysis module, and input the equipment information, multimodal environmental data and personnel vital signs data into the sub-analysis module.

[0007] The sub-analysis module receives the device information, multimodal environmental data, and personnel vital sign data, and outputs the first early warning information according to the early warning mechanism, and transmits the first early warning information to the main analysis module;

[0008] The overall analysis module receives the original 3D model, converts each original 3D model to the second 3D coordinate system, merges them into a real-time 3D model, marks the first warning information on the real-time 3D model, and outputs the second warning information according to the regional warning mechanism.

[0009] As a preferred embodiment of the coal mine safety risk identification system based on edge IoT described in this invention, the acquisition module is provided in multiple ways, and the acquisition module includes a video acquisition unit, an equipment information acquisition unit and a multimodal environment data acquisition unit respectively;

[0010] The video acquisition unit uses a camera to collect video data from underground coal mines, and the video acquisition areas of each acquisition module have overlapping parts;

[0011] The equipment information acquisition unit uses the wind speed sensor to collect the wind speed information of the ventilator and the flow meter to collect the flow information of the water pump, and records the acquisition time. The equipment information includes the wind speed information of the ventilator and the flow information of the water pump.

[0012] The multimodal environmental data includes methane concentration data, carbon monoxide concentration data, oxygen concentration data, carbon dioxide concentration data, ethylene data, acetylene data, dust concentration data, temperature data, and humidity data.

[0013] The multimodal environmental data acquisition unit includes a gas sensor, a carbon monoxide sensor, an oxygen sensor, a carbon dioxide sensor, an ethylene sensor, an acetylene sensor, a mine dust concentration sensor, and an intrinsically safe mine temperature and humidity sensor.

[0014] The infrared gas sensor is used to collect gas concentration data;

[0015] The carbon monoxide sensor is used to collect carbon monoxide concentration data;

[0016] The oxygen sensor is used to collect oxygen concentration data;

[0017] The carbon dioxide sensor is used to collect carbon dioxide concentration data;

[0018] The ethylene sensor is used to collect ethylene data;

[0019] The acetylene sensor is used to collect acetylene data;

[0020] The mining dust concentration sensor is used to collect dust concentration data;

[0021] The intrinsically safe temperature and humidity sensor for mining is used to collect temperature and humidity data, and to record the collection time of methane concentration data, carbon monoxide concentration data, oxygen concentration data, carbon dioxide concentration data, ethylene data, acetylene data, dust concentration data, temperature data, and humidity data, respectively.

[0022] The vital signs data of the personnel include heart rate information, body temperature information, blood oxygen saturation and fatigue value. The heart rate information, body temperature information, blood oxygen saturation and fatigue value are collected using a smart bracelet, and the collection time of the heart rate information, body temperature information, blood oxygen saturation and fatigue value are recorded respectively.

[0023] As a preferred embodiment of the coal mine safety risk identification system based on edge IoT described in this invention, the first early warning information includes a first-level early warning, a second-level early warning, and a third-level early warning;

[0024] The sub-analysis module includes an equipment information analysis unit, a multimodal environmental data analysis unit, and a personnel vital sign data analysis unit;

[0025] The equipment information analysis unit receives the collected wind speed and flow rate information;

[0026] The multimodal environmental data analysis unit receives data on methane concentration, carbon monoxide concentration, oxygen concentration, carbon dioxide concentration, ethylene concentration, acetylene concentration, dust concentration, temperature, and humidity.

[0027] The personnel vital signs data analysis receives and collects heart rate information, body temperature information, blood oxygen saturation, and fatigue value.

[0028] As a preferred embodiment of the coal mine safety risk identification system based on edge IoT described in this invention, the early warning mechanism specifically includes:

[0029] The equipment information analysis unit, multimodal environment data analysis unit, and personnel vital sign data analysis unit receive equipment information, multimodal environment data, and personnel vital sign data, and perform graded early warnings, outputting first early warning information. The first early warning information includes first-level, second-level, and third-level early warnings for personnel vital sign data, as well as first-level, second-level, and third-level early warnings for equipment information, and first-level, second-level, and third-level early warnings for multimodal environment data.

[0030] As a preferred embodiment of the coal mine safety risk identification system based on edge IoT described in this invention, the first-level, second-level, and third-level early warnings of the personnel vital signs data include:

[0031] The heart rate information is compared with the first heart rate threshold and the second heart rate threshold. If the heart rate information is lower than the first heart rate threshold, it indicates that the heart rate information is a level three warning.

[0032] If the heart rate information is higher than the first heart rate threshold and lower than the second heart rate threshold, and the duration exceeds the first unit of time, it indicates that the heart rate information is a level two warning.

[0033] If the heart rate information is higher than the second heart rate threshold and lasts for more than the first unit of time, it indicates that the heart rate information is a level one warning, and the second heart rate threshold is higher than the first heart rate threshold;

[0034] The body temperature information is compared with the first body temperature threshold and the second body temperature threshold. If the body temperature information is lower than the first body temperature threshold, it indicates that the body temperature information is a level three warning.

[0035] If the body temperature information is higher than the first body temperature threshold but lower than the second body temperature threshold, and the duration exceeds the second unit of time, it indicates that the body temperature information is a level two warning;

[0036] If the body temperature information is higher than the second body temperature threshold and lasts for more than the second unit of time, it indicates that the body temperature information is a level one warning, and the second body temperature threshold is higher than the first body temperature threshold;

[0037] The blood oxygen saturation information is compared with the first blood oxygen saturation threshold and the second blood oxygen saturation threshold. If the blood oxygen saturation information is lower than the first blood oxygen saturation threshold, it indicates that the blood oxygen saturation information is a level three warning.

[0038] If the blood oxygen saturation information is higher than the first blood oxygen saturation threshold and lower than the second blood oxygen saturation threshold, and the duration exceeds the third unit of time, it indicates that the blood oxygen saturation information is a level two warning.

[0039] If the blood oxygen saturation information is higher than the second blood oxygen saturation threshold and lasts for more than the third unit of time, it indicates that the blood oxygen saturation information is a level one warning, and the second blood oxygen saturation threshold is higher than the first blood oxygen saturation threshold.

[0040] The fatigue value is compared with the first fatigue threshold and the second fatigue threshold. If the fatigue value is lower than the first fatigue threshold, it indicates that the fatigue value is a level three warning.

[0041] If the fatigue value is higher than the first fatigue value threshold but lower than the second fatigue value threshold, and the duration exceeds the fourth unit of time, the fatigue value is considered a level two warning.

[0042] If the fatigue value is higher than the second fatigue value threshold and lasts for more than four units of time, it indicates that the fatigue value is a level one warning, and the second fatigue value threshold is higher than the first fatigue value threshold.

[0043] As a preferred embodiment of the coal mine safety risk identification system based on edge IoT described in this invention, the primary, secondary, and tertiary early warnings of the equipment information include:

[0044] The wind speed information is compared with the first wind speed information threshold and the second wind speed information threshold. If the wind speed information is greater than the first wind speed information threshold, it indicates that the wind speed information is a level three warning.

[0045] If the wind speed information is less than the first wind speed information threshold and greater than the second wind speed information threshold, and the duration exceeds the fifth unit of time, it indicates that the wind speed information is a level two warning.

[0046] If the wind speed information is less than the second wind speed information threshold and the duration exceeds the fifth unit of time, it indicates that the wind speed information is a level one warning, and the second wind speed information threshold is greater than the first wind speed information threshold.

[0047] The traffic information is compared with the first traffic information threshold and the second traffic information threshold. If the traffic information is greater than the first traffic information threshold, it indicates that the traffic information is a level three warning.

[0048] If the traffic flow information is less than the first traffic flow information threshold and greater than the second traffic flow information threshold, and the duration exceeds the sixth unit of time, it indicates that the traffic flow information is a level two warning.

[0049] If the traffic flow information is greater than the second traffic flow information threshold and the duration exceeds the sixth unit of time, it indicates that the traffic flow information is a level one warning, and the second traffic flow information threshold is greater than the first traffic flow information threshold;

[0050] As a preferred embodiment of the coal mine safety risk identification system based on edge IoT described in this invention, the first-level, second-level, and third-level early warnings of the multimodal environmental data include:

[0051] If ethylene and acetylene data are detected, it indicates that the ethylene and acetylene data are at the level of a Level 1 warning.

[0052] The gas concentration is compared with the first gas concentration threshold and the second gas concentration threshold. If the gas concentration information is lower than the first gas concentration threshold, it indicates that the gas concentration information is a level three warning.

[0053] If the gas concentration information is higher than the first gas concentration threshold but lower than the second gas concentration threshold, it indicates that the gas concentration information is at the level of a level two warning.

[0054] If the gas concentration information is higher than the second gas concentration threshold, it indicates that the gas concentration information is at the first level of warning, and the second gas concentration threshold is higher than the first gas concentration threshold;

[0055] The carbon monoxide concentration is compared with the first carbon monoxide concentration threshold and the second carbon monoxide concentration threshold. If the carbon monoxide concentration information is lower than the first carbon monoxide concentration threshold, it indicates that the carbon monoxide concentration information is a level three warning.

[0056] If the carbon monoxide concentration information is higher than the first carbon monoxide concentration threshold but lower than the second carbon monoxide concentration threshold, it indicates that the carbon monoxide concentration information is a level two warning.

[0057] If the carbon monoxide concentration information is higher than the second carbon monoxide concentration threshold, it indicates that the carbon monoxide concentration information is a level one warning, and the second carbon monoxide concentration threshold is higher than the first carbon monoxide concentration threshold.

[0058] The carbon dioxide concentration is compared with a first carbon dioxide concentration threshold and a second carbon dioxide concentration threshold. If the carbon dioxide concentration information is lower than the first carbon dioxide concentration threshold, it indicates that the carbon dioxide concentration information is a level three warning.

[0059] If the carbon dioxide concentration information is higher than the first carbon dioxide concentration threshold but lower than the second carbon dioxide concentration threshold, it indicates that the carbon dioxide concentration information is a level two warning.

[0060] If the carbon dioxide concentration information is higher than the second carbon dioxide concentration threshold, it indicates that the carbon dioxide concentration information is a level one warning, and the second carbon dioxide concentration threshold is higher than the first carbon dioxide concentration threshold;

[0061] The dust concentration is compared with the first dust concentration threshold and the second dust concentration threshold. If the dust concentration information is lower than the first dust concentration threshold, it indicates that the dust concentration information is a level three warning.

[0062] If the dust concentration information is higher than the first dust concentration threshold but lower than the second dust concentration threshold, it indicates that the dust concentration information is a level two warning.

[0063] If the dust concentration information is higher than the second dust concentration threshold, it indicates that the dust concentration information is a level one warning, and the second dust concentration threshold is higher than the first dust concentration threshold;

[0064] The oxygen concentration is compared with the first oxygen concentration threshold and the second oxygen concentration threshold. If the oxygen concentration information is higher than the first oxygen concentration threshold, it indicates that the oxygen concentration information is a level three warning.

[0065] If the oxygen concentration information is lower than the first oxygen concentration threshold but higher than the second oxygen concentration threshold, it indicates that the oxygen concentration information is a level two warning.

[0066] If the oxygen concentration information is lower than the second oxygen concentration threshold, it indicates that the oxygen concentration information is in the first level warning stage, and the second oxygen concentration threshold is lower than the first oxygen concentration threshold.

[0067] The temperature data is compared with a first temperature data threshold and a second temperature data threshold. If the temperature data is lower than the first temperature data threshold, it indicates that the temperature data is a level three warning.

[0068] If the temperature data is higher than the first temperature data threshold and lower than the second temperature data threshold, and the duration exceeds the seventh unit of time, the temperature data is considered a level two warning.

[0069] If the temperature data is higher than the second temperature data threshold and the duration exceeds the seventh unit of time, it indicates that the temperature data is a level one warning, and the second temperature data threshold is higher than the first temperature data threshold.

[0070] The humidity data is compared with the humidity threshold range. If the humidity data is lower or higher than the humidity threshold range, it indicates that the humidity data is a level one warning.

[0071] As a preferred embodiment of the coal mine safety risk identification system based on edge IoT described in this invention, the system includes: marking a first set of reference points in advance in the underground coal mine and numbering the first set of reference points; marking a second set of reference points at the same location on the original 3D model; and using the same numbering as the first set of reference points.

[0072] The overall analysis module assigns secondary numbering to the acquisition modules in sequence, receives the original 3D models sent by each acquisition module, converts each original 3D model to a second 3D coordinate system, and fuses the original 3D models sent by adjacent secondary-numbered acquisition modules.

[0073] As a preferred embodiment of the coal mine safety risk identification system based on edge IoT described in this invention, the fusion of the original 3D models sent by adjacent numbered acquisition modules includes:

[0074] The original 3D models sent by adjacent secondary numbering acquisition modules are matched with the same primary numbering information. The same primary number is used as the registration point to stitch the two original 3D models into one 3D model. This process continues until all original 3D models are stitched together to form a real-time 3D model.

[0075] After the real-time 3D model is formed, a new original 3D model is received from the acquisition module. The same first number on the new original 3D model and the real-time 3D model is used as the registration point. The new original 3D model is then stitched onto the real-time 3D model to update the real-time 3D model.

[0076] As a preferred embodiment of the coal mine safety risk identification system based on edge IoT described in this invention, wherein: the first early warning information is annotated on the real-time 3D model, and the second early warning information is output according to the regional early warning mechanism, including:

[0077] The same set of original 3D models, secondary numbers and first warning information are treated as a group of regions. Based on the correspondence between the original 3D models, secondary numbers and first warning information, the first warning information is marked on the real-time 3D model, and the second warning information is output.

[0078] The second warning information includes Level 1, Level 2, and Level 3 warnings for personnel vital signs data, Level 1, Level 2, and Level 3 warnings for equipment information, and Level 1, Level 2, and Level 3 warnings for multimodal environment data.

[0079] The first-level, second-level, and third-level warnings of equipment information will be directly output as the first-level, second-level, and third-level warnings of the second warning information of equipment information;

[0080] The Level 1, Level 2, and Level 3 early warnings of personnel vital signs data are integrated and output as the Level 2 early warning information, which includes:

[0081] If any two or more of the heart rate, body temperature, blood oxygen saturation, and fatigue values ​​show a Level II warning, the second warning information output for the personnel's vital signs data will be a Level II warning. Once the personnel's vital signs data reach a Level II warning, the staff member needs to rest and be observed.

[0082] If any one of the heart rate, body temperature, blood oxygen saturation, or fatigue level reaches a Level 3 warning, the second warning information output for the personnel's vital signs data will be a Level 3 warning. After the personnel's vital signs data reaches a Level 3 warning, the staff member needs to undergo further physical examination.

[0083] The first, second, and third level early warnings from multimodal environmental data are integrated and output as the second early warning information, which includes:

[0084] If any two or more of the following data—gas concentration, carbon monoxide concentration, oxygen concentration, carbon dioxide concentration, dust concentration, temperature, and humidity—show a Level II warning, then the multimodal environmental data output for that area will be a Level II warning, indicating that work needs to be suspended for observation.

[0085] If two or more adjacent groups of areas with secondary numbering subsequently issue a Level II warning, the multimodal environmental data output for the affected areas will be a Level III warning, requiring personnel evacuation and work stoppage.

[0086] If any of the following data points triggers a Level 3 warning: methane concentration, carbon monoxide concentration, oxygen concentration, carbon dioxide concentration, ethylene concentration, acetylene concentration, or dust concentration, the multimodal environmental data output for that area will be a Level 3 warning, requiring personnel evacuation and a halt to work.

[0087] The beneficial effects of this invention are as follows: This invention achieves multi-source acquisition of video data, equipment information, multimodal environmental data, and personnel vital signs data through the acquisition module, and uses video data to perform three-dimensional reconstruction to generate an original three-dimensional model. The overall analysis module integrates multiple original three-dimensional models into a real-time three-dimensional model, realizing three-dimensional visualization of the underground coal mine environment. Supervisors can intuitively see the underground spatial structure and equipment distribution, which is conducive to improving the ability to perceive the spatial situation underground in coal mines.

[0088] This invention performs real-time analysis of multi-source data at the edge through a sub-analysis module, and quickly outputs the first warning information according to the early warning mechanism. The main analysis module marks the first warning information on the real-time three-dimensional model, realizing the spatial visualization of safety risks. Regulatory personnel can quickly locate the specific location and impact range of the risk source, effectively shortening the emergency response time.

[0089] This invention comprehensively considers the multi-parameter correlation of a single region and the spatial correlation of adjacent regions through a regional early warning mechanism, realizing a progressive early warning from local anomalies to regional risks. This effectively avoids the false alarms and missed alarms of single-parameter early warning, and significantly improves the accuracy and reliability of early warning decisions.

[0090] This invention achieves accurate registration of multi-view 3D models through a reference point numbering mechanism and ensures the timeliness of real-time 3D models through an incremental update method, providing strong technical support for safe coal mine production. Attached Figure Description

[0091] Figure 1 A basic flowchart of a coal mine safety risk identification system based on edge IoT is provided as an embodiment of the present invention;

[0092] Figure 2 This is a schematic diagram of the process architecture of a coal mine safety risk identification system based on edge IoT, provided as an embodiment of the present invention. Detailed Implementation

[0093] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0094] The system provided in this embodiment of the invention is deployed underground in coal mines and at a surface monitoring center. Multiple acquisition modules and sub-analysis modules are deployed underground. The acquisition modules collect video data, equipment information, multimodal environmental data, and personnel vital sign data through various sensors and cameras. The sub-analysis modules are deployed at underground edge computing nodes to achieve on-site data processing and rapid response. A central analysis module is deployed at the surface monitoring center to receive data from each acquisition module, perform 3D model fusion, and provide comprehensive regional early warning. The system transmits data via mining industrial Ethernet or a 5G network to ensure real-time and reliable data transmission.

[0095] Example 1, Reference Point Figure 1 and Figure 2 As an embodiment of the present invention, a coal mine safety risk identification system based on edge IoT is provided, including: a data acquisition module, a sub-analysis module and a total analysis module;

[0096] The acquisition module is used to collect video data, equipment information, multimodal environmental data and personnel vital signs data in the coal mine. It also extracts three-dimensional point clouds from the video data, converts the three-dimensional point clouds into the first three-dimensional coordinate system, generates the original three-dimensional model, transmits the original three-dimensional model to the main analysis module, and inputs the equipment information, multimodal environmental data and personnel vital signs data into the sub-analysis module.

[0097] In this embodiment, the acquisition module is a set of multi-source data acquisition units deployed at key locations underground in the coal mine to achieve synchronous acquisition of video data, equipment operating parameters, environmental parameters, and personnel vital signs data. Multiple acquisition modules are set up and deployed at key nodes in the underground roadways. Each acquisition module includes a video acquisition unit, an equipment information acquisition unit, a multimodal environmental data acquisition unit, and a personnel vital signs data acquisition unit.

[0098] The system has multiple acquisition modules, which include a video acquisition unit, a device information acquisition unit, and a multimodal environment data acquisition unit.

[0099] The video acquisition unit uses cameras to collect video data from underground coal mines, and the video acquisition areas of each acquisition module overlap.

[0100] In this embodiment, the acquisition module collects underground data in real time using various sensors and cameras, converts the video data into 3D point clouds and generates an original 3D model, transmits other data to the sub-analysis module for edge analysis, and uploads the original 3D model to the main analysis module for fusion processing. The video acquisition unit specifically includes a high-definition camera, a supplementary light, and an image processing chip; the high-definition camera is used to acquire video data from the underground roadway, with a resolution of no less than 1080P and a frame rate of no less than 25fps; the supplementary light provides auxiliary lighting in low-light environments underground; and the image processing chip processes the video data in real time, extracts the 3D point cloud, and generates the original 3D model.

[0101] In one embodiment, the video data is processed using a multi-view stereo vision (MVS) algorithm. A projection matrix is ​​constructed using the intrinsic parameter matrix and extrinsic homogeneous transformation matrix obtained from high-definition camera calibration. Corresponding points in images from different viewpoints are determined through feature matching. Then, the three-dimensional coordinates are calculated using the principle of triangulation to analyze the multi-view video sequence. First, image feature points such as the tunnel outline and equipment edges are extracted. Then, the SIFT feature matching algorithm is used to find corresponding points between images from different viewpoints. The depth information of the feature points is calculated based on the camera calibration parameters to generate a depth map. After fusing the depth maps from different viewpoints, a video 3D model containing the underground coal mine is constructed, providing a data foundation for the subsequent original 3D model. A 3D point cloud is then extracted from the video 3D model. After obtaining the data, the origin, coordinate axis directions, and scale factor of the original 3D model's coordinate system are acquired. A coordinate transformation matrix is ​​established to transform the point cloud coordinates of the video 3D model to the corresponding spatial coordinate system of the original 3D model. The Iterative Closest Point (ICP) algorithm is preferred, with a preset maximum number of iterations of 100 and a distance convergence threshold of 1e-6 meters. By minimizing the sum of squared Euclidean distances between the point cloud data in the video 3D model's coordinate system and the point cloud data in the original 3D model's coordinate system, the 3D point cloud data in the video 3D model is registered with the original 3D model. The registered point cloud data is then merged with the mesh structure of the original 3D model. Data redundancy is eliminated through surface fitting, and finally, an original 3D model containing the video 3D model data is generated.

[0102] The video acquisition areas of each acquisition module have overlapping parts, with the proportion of overlapping areas being no less than 30%, to ensure that the original 3D models of adjacent acquisition modules have common feature areas, providing a registration basis for subsequent model fusion.

[0103] The equipment information acquisition unit uses a wind speed sensor to collect the wind speed information of the ventilator and a flow meter to collect the flow rate information of the water pump, and records the acquisition time. The equipment information includes the wind speed information of the ventilator and the flow rate information of the water pump.

[0104] In this embodiment, the equipment information acquisition unit specifically includes a wind speed sensor and a flow meter. The wind speed sensor is installed at the outlet of the ventilation fan's duct to collect the fan's wind speed information, with a measurement range of 0-30 m / s and a measurement accuracy of not less than ±0.3 m / s. The flow meter is installed at the outlet pipe of the water pump to collect the pump's flow rate information, with a measurement range determined according to the pump specifications and a measurement accuracy of not less than ±1%. The equipment information includes the fan's wind speed information and the water pump's flow rate information, and records the acquisition time with a time accuracy of not less than 1 second.

[0105] Multimodal environmental data includes methane concentration data, carbon monoxide concentration data, oxygen concentration data, carbon dioxide concentration data, ethylene data, acetylene data, dust concentration data, temperature data, and humidity data;

[0106] The multimodal environmental data acquisition unit includes a gas sensor, a carbon monoxide sensor, an oxygen sensor, a carbon dioxide sensor, an ethylene sensor, an acetylene sensor, a mine dust concentration sensor, and an intrinsically safe temperature and humidity sensor for mining.

[0107] Infrared gas sensors are used to collect gas concentration data;

[0108] A carbon monoxide sensor is used to collect carbon monoxide concentration data;

[0109] Oxygen sensors are used to collect oxygen concentration data;

[0110] Carbon dioxide sensors are used to collect carbon dioxide concentration data;

[0111] Ethylene sensors are used to collect ethylene data;

[0112] The acetylene sensor is used to collect acetylene data;

[0113] Mining dust concentration sensors are used to collect dust concentration data;

[0114] The intrinsically safe temperature and humidity sensor for mining is used to collect temperature and humidity data, and to record the collection time of methane concentration data, carbon monoxide concentration data, oxygen concentration data, carbon dioxide concentration data, ethylene data, acetylene data, dust concentration data, temperature data, and humidity data, respectively.

[0115] In this preferred embodiment, the gas sensor is used to collect gas concentration data. An infrared gas sensor is used, with a measurement range of 0-100% LEL and a measurement accuracy of not less than ±3% FS. The carbon monoxide sensor is used to collect carbon monoxide concentration data, with a measurement range of 0-1000 ppm and a measurement accuracy of not less than ±5% FS. The oxygen sensor is used to collect oxygen concentration data, with a measurement range of 0-30% VOL and a measurement accuracy of not less than ±2% FS. The carbon dioxide sensor is used to collect carbon dioxide concentration data, with a measurement range of 0-5% VOL and a measurement accuracy of not less than ±3% FS. The ethylene sensor uses… The following sensors are used to collect ethylene data: a measurement range of 0-100 ppm with an accuracy of ±5%FS; an acetylene sensor for collecting acetylene data with a measurement range of 0-100 ppm with an accuracy of ±5%FS; a mining dust concentration sensor for collecting dust concentration data with a measurement range of 0-1000 mg / m³ with an accuracy of ±10%; and an intrinsically safe temperature and humidity sensor for collecting temperature and humidity data with a temperature measurement range of -20℃ to +60℃ and an accuracy of ±0.5℃, and a humidity measurement range of 0-100%RH with an accuracy of ±3%RH. The data collected by each sensor is recorded at the acquisition time with a time accuracy of at least 1 second, enabling time-series correlation of the data.

[0116] Personnel vital signs data include heart rate, body temperature, blood oxygen saturation, and fatigue level. These data are collected using a smart bracelet, and the collection time for each data point is recorded.

[0117] In this embodiment, the smart bracelet is worn on the worker's wrist and incorporates a heart rate sensor, temperature sensor, blood oxygen sensor, and accelerometer. The heart rate sensor collects heart rate information, with a measurement range of 30-200 bpm and an accuracy of at least ±2 bpm. The temperature sensor collects body temperature information, with a measurement range of 35℃-42℃ and an accuracy of at least ±0.1℃. The blood oxygen sensor collects blood oxygen saturation, with a measurement range of 70%-100% and an accuracy of at least ±2%. The accelerometer collects motion data, and fatigue values ​​are calculated through motion data analysis. The fatigue value is an index of fatigue level calculated by analyzing parameters such as the worker's motion data, heart rate variability, and activity duration. The value ranges from 0-100, with higher values ​​indicating higher fatigue levels. Each vital sign data point is recorded with its collection time and the worker's identification, enabling temporal and personnel-related correlation of data.

[0118] In practice, the acquisition module adopts a modular design, with each acquisition unit operating independently. Power is supplied by an intrinsically safe mine power supply, and data is transmitted to the sub-analysis module via an RS485 bus or wireless network. The deployment location of the acquisition modules is determined based on the underground roadway layout, with the distance between adjacent acquisition modules not exceeding 100 meters to ensure that the overlapping areas of the video acquisition regions meet the requirements for model fusion.

[0119] The sub-analysis module receives equipment information, multimodal environmental data, and personnel vital sign data, and outputs the first early warning information according to the early warning mechanism, and transmits the first early warning information to the main analysis module;

[0120] In this embodiment, the sub-analysis module is a data analysis unit deployed on the downhole edge computing node. The sub-analysis module corresponds one-to-one with the acquisition module. The sub-analysis module is used to realize real-time analysis and early warning of equipment information, multimodal environmental data and personnel vital signs data.

[0121] The first level of early warning information includes Level 1, Level 2, and Level 3 warnings;

[0122] The sub-analysis module includes an equipment information analysis unit, a multimodal environmental data analysis unit, and a personnel vital signs data analysis unit;

[0123] The equipment information analysis unit receives the collected wind speed and flow rate information; the equipment information analysis unit is used to analyze the fan wind speed information to determine the operating status of the ventilation system, and to analyze the pump flow rate information to determine the operating status of the drainage system.

[0124] The multimodal environmental data analysis unit receives data on methane concentration, carbon monoxide concentration, oxygen concentration, carbon dioxide concentration, ethylene concentration, acetylene concentration, dust concentration, temperature, and humidity.

[0125] The system receives and collects vital signs data, including heart rate, body temperature, blood oxygen saturation, and fatigue levels.

[0126] The early warning mechanism specifically includes:

[0127] The equipment information analysis unit, multimodal environment data analysis unit, and personnel vital sign data analysis unit receive equipment information, multimodal environment data, and personnel vital sign data, and perform graded early warnings, outputting first early warning information. The first early warning information includes first-level, second-level, and third-level early warnings for personnel vital sign data, as well as first-level, second-level, and third-level early warnings for equipment information and multimodal environment data.

[0128] In this embodiment, the first early warning information refers to the early warning result output by the sub-analysis module according to the early warning mechanism, which is used to indicate the degree of abnormality in the status of equipment, environment, or personnel. The first early warning information includes level one, level two, and level three early warnings; among the level one, level two, and level three early warnings in the first and second early warning information: level one early warning indicates the highest risk level, requiring immediate emergency measures; level two early warning indicates a medium risk level, requiring close monitoring and preventive measures; and level three early warning indicates a lower risk level, requiring monitoring and recording.

[0129] The Level 1, Level 2, and Level 3 early warning systems for personnel vital signs data include:

[0130] The heart rate information is compared with the first heart rate threshold and the second heart rate threshold. If the heart rate information is lower than the first heart rate threshold, it indicates that the heart rate information is a level three warning, which means that the heart rate is normal. In this embodiment, only the case of excessively high heart rate is monitored, because the narrow and noisy working environment in the coal mine can cause the heart rate to be too high. An excessively high heart rate indicates that the current worker may be tense, under too much pressure, and in a state of high physiological load. Prolonged exposure to this state can lead to work errors and thus cause safety problems.

[0131] If the heart rate information is higher than the first heart rate threshold but lower than the second heart rate threshold, and the duration exceeds the first unit of time, it indicates that the heart rate information is a level two warning; that is, the heart rate is slightly abnormal, and the current condition of the staff member needs to be monitored.

[0132] If the heart rate information is higher than the second heart rate threshold and lasts for more than the first unit of time, it indicates that the heart rate information is a level one warning. The second heart rate threshold is higher than the first heart rate threshold. This indicates that the staff member is facing some emergencies, may be nervous, or under too much pressure. Prolonged tension and excessive pressure can lead to mistakes in work and thus cause safety problems.

[0133] The body temperature information is compared with a first body temperature threshold and a second body temperature threshold. If the body temperature information is lower than the first body temperature threshold, it indicates a level three warning; otherwise, it indicates a normal body temperature. Excessively high body temperature can lead to physiological dysfunction, decreased cognitive ability, frequent operational errors, increased accident risk, and severely reduced work efficiency.

[0134] If the body temperature is higher than the first body temperature threshold but lower than the second body temperature threshold, and the duration exceeds the second unit of time, it indicates a level two warning; it indicates a slightly abnormal body temperature, possibly due to excessive heat.

[0135] If the body temperature information is higher than the second body temperature threshold and lasts for more than the second unit of time, it indicates that the body temperature information is a level one warning, and the second body temperature threshold is higher than the first body temperature threshold;

[0136] Blood oxygen saturation information is compared with a first blood oxygen saturation threshold and a second blood oxygen saturation threshold. If the blood oxygen saturation information is lower than the first blood oxygen saturation threshold, it indicates a level three warning. Oxygen is fundamental to maintaining brain function. Sufficient oxygen supply can maintain good judgment and awareness, helping miners accurately identify risks and respond correctly. Low blood oxygen saturation indicates hypoxia, which may be due to insufficient oxygen underground or excessive inhalation of harmful gases. Hypoxia will rapidly impair a miner's ability to escape and save themselves.

[0137] If the blood oxygen saturation information is higher than the first blood oxygen saturation threshold and lower than the second blood oxygen saturation threshold, and the duration exceeds the third unit of time, it indicates that the blood oxygen saturation information is a level two warning.

[0138] If the blood oxygen saturation information is higher than the second blood oxygen saturation threshold and lasts for more than the third unit of time, it indicates that the blood oxygen saturation information is a level one warning, and the second blood oxygen saturation threshold is higher than the first blood oxygen saturation threshold.

[0139] The fatigue value is compared with the first fatigue threshold and the second fatigue threshold. If the fatigue value is lower than the first fatigue threshold, it indicates that the fatigue value is a level three warning; there is no obvious feeling of fatigue.

[0140] If the fatigue value is higher than the first fatigue value threshold but lower than the second fatigue value threshold, and the duration exceeds the fourth unit of time, the fatigue value is considered a level two warning, indicating mild fatigue.

[0141] If the fatigue value is higher than the second fatigue threshold and lasts for more than four units of time, it indicates a level one warning for fatigue, where the second fatigue threshold is higher than the first fatigue threshold, indicating severe fatigue.

[0142] In practice, the first heart rate threshold can be set to 50 bpm, and the second heart rate threshold can be set to 120 bpm; the first body temperature threshold can be set to 37.0℃, and the second body temperature threshold can be set to 38.0℃; the first blood oxygen saturation threshold can be set to 90%, and the second blood oxygen saturation threshold can be set to 95%; the first fatigue threshold can be set to 40, and the second fatigue threshold can be set to 70. The first to fourth unit times can be set according to the actual situation, generally 5-30 minutes.

[0143] The equipment information warnings at levels one, two, and three include:

[0144] The wind speed information is compared with the first wind speed information threshold and the second wind speed information threshold. If the wind speed information is greater than the first wind speed information threshold, it indicates that the wind speed information is a level three warning; at this time, the ventilation capacity is normal.

[0145] If the wind speed information is less than the first wind speed information threshold and greater than the second wind speed information threshold, and the duration exceeds the fifth unit of time, it indicates that the wind speed information is a level two warning; it indicates that there may be a problem with the fan. The cross-section of the underground roadway can be considered relatively constant. A decrease in wind speed indicates a decrease in ventilation capacity, and it is necessary to send someone to investigate.

[0146] If the wind speed information is lower than the second wind speed information threshold and this condition persists for more than five units of time, it indicates a Level 1 warning. If the second wind speed information threshold is higher than the first wind speed information threshold, it means the ventilation fan has malfunctioned and must be stopped immediately. Personnel must be dispatched to investigate, because once ventilation is lost in a coal mine, methane and harmful gases will accumulate beyond limits within minutes. If the backup ventilation fan fails to start automatically or the air door malfunctions, the mine will face the dual threats of suffocation and explosion, posing a significant danger to lives.

[0147] The traffic flow information is compared with the first traffic flow information threshold and the second traffic flow information threshold. If the traffic flow information is greater than the first traffic flow information threshold, it indicates that the traffic flow information is a level three warning.

[0148] If the flow rate is less than the first flow rate threshold but greater than the second flow rate threshold, and the duration exceeds the sixth unit of time, it indicates that the flow rate is a level two warning; it indicates that the drainage system's drainage capacity is insufficient, and personnel need to be dispatched to investigate.

[0149] If the flow rate exceeds the second flow rate threshold and lasts for more than six units of time, it indicates a Level 1 warning. The second flow rate threshold is greater than the first flow rate threshold, indicating a problem with the drainage system. Work must be stopped immediately and personnel must be dispatched to investigate. This is because a pump or power line failure can lead to insufficient drainage capacity. When the water flow suddenly increases (water inrush accident), it cannot be discharged in time, which will directly cause the well to flood and endanger lives.

[0150] In practice, the first wind speed threshold can be set to 2 m / s, and the second wind speed threshold can be set to 8 m / s; the first and second flow rate thresholds are determined according to the pump specifications. The fifth and sixth unit times can be set to 10-60 minutes.

[0151] The Level 1, Level 2, and Level 3 early warning systems for multimodal environmental data include:

[0152] If ethylene and acetylene data are detected, it indicates that the ethylene and acetylene data are at the level of a Level 1 warning.

[0153] The gas concentration is compared with the first gas concentration threshold and the second gas concentration threshold. If the gas concentration information is lower than the first gas concentration threshold, it indicates that the gas concentration information is a level three warning. Since ethylene and acetylene are important indicator gases for coal mine fires and explosions, once detected, the highest level warning is triggered, indicating that there may be a risk of coal spontaneous combustion or fire.

[0154] If the gas concentration information is higher than the first gas concentration threshold but lower than the second gas concentration threshold, it indicates that the gas concentration information is at the level of a level two warning.

[0155] If the gas concentration information is higher than the second gas concentration threshold, it indicates that the gas concentration information is at the first level of warning. If the second gas concentration threshold is higher than the first gas concentration threshold, the gas concentration is seriously exceeded and there is a risk of explosion.

[0156] The carbon monoxide concentration is compared with the first and second carbon monoxide concentration thresholds. If the carbon monoxide concentration is lower than the first carbon monoxide concentration threshold, it indicates that the carbon monoxide concentration is at level three warning. Carbon monoxide is produced by coal combustion or oxidation reaction. If the carbon monoxide concentration is too high, it indicates that there may be a fire in the coal mine.

[0157] If the carbon monoxide concentration information is higher than the first carbon monoxide concentration threshold but lower than the second carbon monoxide concentration threshold, it indicates that the carbon monoxide concentration information is a level two warning.

[0158] If the carbon monoxide concentration information is higher than the second carbon monoxide concentration threshold, it indicates that the carbon monoxide concentration information is at the first level warning level, and the second carbon monoxide concentration threshold is higher than the first carbon monoxide concentration threshold.

[0159] The carbon dioxide concentration is compared with the first carbon dioxide concentration threshold and the second carbon dioxide concentration threshold. If the carbon dioxide concentration information is lower than the first carbon dioxide concentration threshold, it indicates that the carbon dioxide concentration information is a level three warning.

[0160] If the carbon dioxide concentration information is higher than the first carbon dioxide concentration threshold but lower than the second carbon dioxide concentration threshold, it indicates that the carbon dioxide concentration information is a level two warning.

[0161] If the carbon dioxide concentration information is higher than the second carbon dioxide concentration threshold, it indicates that the carbon dioxide concentration information is a level one warning. If the second carbon dioxide concentration threshold is higher than the first carbon dioxide concentration threshold, there is a risk of asphyxiation due to the carbon dioxide concentration.

[0162] The dust concentration is compared with the first dust concentration threshold and the second dust concentration threshold. If the dust concentration information is lower than the first dust concentration threshold, it indicates that the dust concentration information is a level three warning.

[0163] If the dust concentration information is higher than the first dust concentration threshold but lower than the second dust concentration threshold, it indicates that the dust concentration information is a level two warning.

[0164] If the dust concentration information is higher than the second dust concentration threshold, it indicates that the dust concentration information is at the first level of warning. If the second dust concentration threshold is higher than the first dust concentration threshold, the dust is seriously exceeding the standard and there is a risk of explosion.

[0165] The oxygen concentration is compared with the first oxygen concentration threshold and the second oxygen concentration threshold. If the oxygen concentration information is higher than the first oxygen concentration threshold, it indicates that the oxygen concentration information is a level three warning.

[0166] If the oxygen concentration information is lower than the first oxygen concentration threshold but higher than the second oxygen concentration threshold, it indicates that the oxygen concentration information is a level two warning.

[0167] If the oxygen concentration information is lower than the second oxygen concentration threshold, it indicates that the oxygen concentration information is in the first level of warning, and the second oxygen concentration threshold is lower than the first oxygen concentration threshold; the oxygen is severely insufficient and there is a risk of suffocation.

[0168] The temperature data is compared with the first temperature data threshold and the second temperature data threshold. If the temperature data is lower than the first temperature data threshold, it indicates that the temperature data is a level three warning.

[0169] If the temperature data is higher than the first temperature data threshold and lower than the second temperature data threshold, and the duration exceeds the seventh unit of time, the temperature data is considered a level two warning.

[0170] If the temperature data is higher than the second temperature data threshold and the duration exceeds the seventh unit of time, it indicates that the temperature data is a Level 1 warning, the second temperature data threshold is higher than the first temperature data threshold; the temperature is severely abnormal, and there may be spontaneous combustion of coal.

[0171] Compare the humidity data with the humidity threshold range. If the humidity data is lower or higher than the humidity threshold range, it indicates that the humidity data is a level 1 warning.

[0172] In this embodiment, the sub-analysis module summarizes the first early warning information of various types of data and transmits it to the main analysis module of the ground monitoring center through the mining industrial Ethernet or 5G network.

[0173] In specific implementation, the first gas concentration threshold can be set to 0.5%, and the second gas concentration threshold can be set to 1.0%; the first carbon monoxide concentration threshold can be set to 24 ppm, and the second carbon monoxide concentration threshold can be set to 50 ppm; the first carbon dioxide concentration threshold can be set to 0.5%, and the second carbon dioxide concentration threshold can be set to 1.5%; the first dust concentration threshold can be set to 10 mg / m³, and the second dust concentration threshold can be set to 50 mg / m³; the first oxygen concentration threshold can be set to 20%, and the second oxygen concentration threshold can be set to 18%; the first temperature threshold can be set to 30℃, and the second temperature threshold can be set to 40℃; the humidity threshold range can be set to 40%-80%RH. The seventh unit time can be set to 30-120 minutes.

[0174] The overall analysis module receives the original 3D model, converts each original 3D model to the second 3D coordinate system, merges them into a real-time 3D model, marks the first warning information on the real-time 3D model, and outputs the second warning information according to the regional warning mechanism.

[0175] In this embodiment, the overall analysis module is deployed in the integrated analysis unit of the ground monitoring center to realize the fusion of multiple original 3D models, spatial labeling of early warning information, and comprehensive regional early warning. A first set of reference points is pre-marked underground in the coal mine and numbered. A second set of reference points is then marked at the same locations on the original 3D model, using the same numbering system as the first set.

[0176] In this embodiment, the underground reference point marking is done manually in the coal mine when the system deployment phase is completed, while the model reference point marking is done automatically when the original 3D model is generated.

[0177] The first set of reference points is pre-marked underground in the coal mine and assigned a number. These reference points are made using reflective markers and are deployed on the tunnel walls, roof, or equipment surfaces. The spacing between adjacent reference points is determined based on the tunnel length and the camera's field of view, generally set to 10-50 meters. The initial numbering uses numerical or alphanumeric codes, such as A1, A2, A3... or 1, 2, 3..., to ensure the uniqueness of the numbers.

[0178] At the same locations on the original 3D model, a second set of reference points was marked, using the same numbering as the first set. The second set of reference points automatically detected reflective markers in the video data using an image recognition algorithm and was then marked to ensure correspondence with the actual reference points downhole.

[0179] The overall analysis module assigns secondary numbering to the acquisition modules in sequence. The secondary numbering reflects the spatial order of the acquisition modules in the roadway, such as C1, C2, C3... or 1, 2, 3...

[0180] The overall analysis module assigns secondary numbering to the acquisition modules in sequence, receives the original 3D models sent by each acquisition module, converts each original 3D model to a secondary 3D coordinate system, and fuses the original 3D models sent by adjacent secondary numbered acquisition modules.

[0181] The fusion of the original 3D models sent by adjacent acquisition modules includes:

[0182] The original 3D models sent by adjacent secondary numbering acquisition modules are matched with the same primary numbering information. The same primary number is used as the registration point to stitch the two original 3D models into one 3D model. This process continues until all original 3D models are stitched together to form a real-time 3D model.

[0183] After the real-time 3D model is formed, the new original 3D model sent by the acquisition module is received. The same number on the new original 3D model and the real-time 3D model is used as the registration point. The new original 3D model is then stitched onto the real-time 3D model to update the real-time 3D model.

[0184] In this embodiment, the fusion method of the original 3D model includes two stages: initial fusion and incremental update. The initial fusion is completed when the system starts up, and the incremental update continues during subsequent operation.

[0185] Initial Fusion Phase: The main analysis module receives the original 3D models sent by each acquisition module and transforms each original 3D model into a second 3D coordinate system; the second 3D coordinate system is a global coordinate system with the coal mine shaft or a designated location as its origin. The original 3D models sent by adjacent acquisition modules with secondary numbering are then fused. Specifically, this involves matching the same primary numbering information in the original 3D models sent by adjacent acquisition modules with secondary numbering, using the same primary number as a registration point, calculating the coordinate transformation matrix between the two models using the registration point, and stitching the two original 3D models into a single 3D model. This process is repeated for all adjacent original 3D models until all original 3D models are stitched together, forming a real-time 3D model.

[0186] Incremental Update Phase: After the real-time 3D model is formed, the overall analysis module continuously receives new original 3D models sent by the acquisition module. The same primary identification number on the new original 3D model and the real-time 3D model is used as a registration point. The coordinate transformation matrix is ​​calculated using the registration point, and the new original 3D model is then stitched onto the real-time 3D model to update it. Incremental updates ensure that the real-time 3D model reflects the latest state of the underground coal mine environment.

[0187] In practice, the model fusion uses the Iterative Closest Point (ICP) algorithm for fine registration, with a registration accuracy of no less than 5 centimeters. The update cycle of the real-time 3D model is determined based on the data upload frequency of the acquisition module, and is generally set to 1-10 seconds.

[0188] The first early warning information is marked on the real-time 3D model, and the second early warning information is output according to the regional early warning mechanism, including:

[0189] The same set of original 3D models, secondary numbers, and first warning information are grouped into one region. Based on the correspondence between the original 3D models, secondary numbers, and first warning information, the first warning information is annotated on the real-time 3D model, and the second warning information is output. In this embodiment, the spatial annotation of the first warning information is achieved by establishing the correspondence between the original 3D models, secondary numbers, and first warning information. The same set of original 3D models, secondary numbers, and first warning information are grouped into one region; each region corresponds to the monitoring range of one acquisition module, and the region boundary is determined based on the overlapping portion of the video acquisition areas of adjacent acquisition modules. Based on the correspondence between the original 3D models, secondary numbers, and first warning information, the first warning information is annotated on the corresponding spatial location on the real-time 3D model. The annotation method uses color coding and icon display: Level 1 warnings are annotated in red, Level 2 warnings in orange, and Level 3 warnings in green. The annotation content includes the warning type (equipment, environment, personnel), warning level, warning parameters, and warning time.

[0190] The second set of early warning information includes Level 1, Level 2, and Level 3 early warnings for personnel vital signs data, Level 1, Level 2, and Level 3 early warnings for equipment information, and Level 1, Level 2, and Level 3 early warnings for multimodal environmental data.

[0191] The first, second, and third level warnings of equipment information are directly output as the first, second, and third level warnings of the second level warning information of equipment information. The integration rule for the second level warning information of equipment information is as follows: the first, second, and third level warnings of equipment information are directly output as the first, second, and third level warnings of the second level warning information of equipment information, without additional integration. This is because the equipment is relatively independent.

[0192] The Level 1, Level 2, and Level 3 early warnings of personnel vital signs data are integrated and output as the Level 2 early warning information, which includes:

[0193] If any two or more of the heart rate, body temperature, blood oxygen saturation, and fatigue values ​​show a Level II warning, the second warning information output for the personnel's vital signs data will be a Level II warning. Once the personnel's vital signs data reach a Level II warning, the staff member needs to rest and be observed.

[0194] If any one of the heart rate, body temperature, blood oxygen saturation, or fatigue level reaches a Level 3 warning, the second warning information output for the personnel's vital signs data will be a Level 3 warning. After the personnel's vital signs data reaches a Level 3 warning, the staff member needs to undergo further physical examination.

[0195] The first, second, and third level early warnings from multimodal environmental data are integrated and output as the second early warning information, which includes:

[0196] If any two or more of the following data—gas concentration, carbon monoxide concentration, oxygen concentration, carbon dioxide concentration, dust concentration, temperature, and humidity—show a Level II warning, then the multimodal environmental data output for that area will be a Level II warning, indicating that work needs to be suspended for observation.

[0197] If two or more adjacent groups of areas with secondary numbering subsequently issue a Level II warning, the multimodal environmental data output for the affected areas will be a Level III warning, requiring personnel evacuation and work stoppage.

[0198] If any of the following data points—gas concentration, carbon monoxide concentration, oxygen concentration, carbon dioxide concentration, ethylene concentration, acetylene concentration, or dust concentration—shows a Level 3 warning, then the multimodal environmental data output for that area will be a Level 3 warning, requiring personnel evacuation and work stoppage.

[0199] When anomalies occur in the multimodal environmental data of a certain area:

[0200] The multimodal environmental data analysis unit of the sub-analysis module monitors the environmental parameters of the area in real time. If any two or more of the following data—gas concentration, carbon monoxide concentration, oxygen concentration, carbon dioxide concentration, dust concentration, temperature, and humidity—show a Level II warning, the multimodal environmental data output for that area will be a Level II warning. The main analysis module will mark the area in orange on the real-time 3D model and issue a stop-work observation command. If any of the following data shows a Level I warning, the multimodal environmental data output for that area will be a Level I warning. The main analysis module will mark the area in red on the real-time 3D model and issue personnel evacuation and stop-work commands. When ethylene or acetylene gas is detected, a Level I warning will be triggered immediately, and the fire emergency response plan will be activated, notifying relevant personnel to conduct investigation and handling.

[0201] When environmental parameters from multiple adjacent regions show abnormal correlation:

[0202] The regional early warning unit of the overall analysis module continuously monitors the early warning status of each region. If two or more adjacent groups of regions with secondary numbering subsequently issue a Level II early warning, the multimodal environmental data output for the affected region will be a Level I early warning. The overall analysis module will mark the affected region in red on the real-time 3D model and issue personnel evacuation and work stoppage instructions. This scenario reflects the spatial diffusion trend of abnormal underground environment. By linking early warnings in adjacent regions, potential large-scale risks can be identified in advance, buying valuable time for emergency response. In specific implementation, when the gas concentration in two adjacent regions both show a Level II early warning, it is determined that there may be a ventilation system malfunction or abnormal gas outburst. The early warning level will be immediately upgraded, and the ventilation department will be notified to investigate.

[0203] When the vital signs data of the workers are abnormal:

[0204] The personnel vital signs data analysis unit of the sub-analysis module monitors the vital signs parameters of each worker in real time. If any two or more of the heart rate, body temperature, blood oxygen saturation, and fatigue value show a Level II warning, the second warning information for the personnel vital signs data is output as a Level II warning. The main analysis module marks the worker's location in orange on the real-time 3D model and issues a rest and observation instruction. If any of the heart rate, body temperature, blood oxygen saturation, and fatigue value reaches a Level I warning, the second warning information for the personnel vital signs data is output as a Level I warning. The main analysis module marks the worker's location in red on the real-time 3D model and issues an instruction for further physical examination. In specific implementation, when a worker's blood oxygen saturation remains below 90% and their heart rate remains above 120 bpm, it is determined that the worker may be experiencing hypoxia or physical discomfort. The worker is immediately notified to stop working and go to a safe area to rest, and the on-site safety officer is notified to pay attention.

[0205] When the ventilator or water pump malfunctions:

[0206] The equipment information analysis unit of the sub-analysis module monitors equipment operating parameters in real time. If a level three warning is issued for wind speed or flow rate information, the system records the warning information and notifies equipment maintenance personnel to pay attention; if a level two warning is issued, the system issues an equipment inspection command; if a level one warning is issued, an equipment fault alarm is issued and backup equipment is activated. In specific implementation, when the fan wind speed is consistently below 2 m / s, it is determined that the ventilation capacity is insufficient and may affect the underground air quality. The ventilation department is immediately notified to check the fan operating status and the integrity of the ventilation duct.

[0207] This embodiment achieves synchronous acquisition of multi-source data through an acquisition module, real-time edge analysis through a sub-analysis module, and comprehensive regional early warning through a general analysis module, forming a complete coal mine safety risk identification system. The system annotates early warning information on a real-time 3D model, realizing spatial visualization of safety risks, allowing supervisors to intuitively see the location and distribution range of risk sources. The regional early warning mechanism comprehensively considers the multi-parameter correlation of a single region and the spatial correlation of adjacent regions, achieving progressive early warning from local anomalies to regional risks, effectively improving the accuracy and reliability of early warning decisions.

[0208] Example 2

[0209] The reference diagram illustrates another embodiment of the present invention, which is based on the previous embodiment. The difference between this embodiment and the previous embodiment is that it provides a coal mine safety risk identification system based on edge IoT. To verify and illustrate the technical effects of the method, this embodiment uses a traditional technical solution to compare and test with the method of the present invention, and compares the test results using scientific demonstration methods to verify the real effect of the method.

[0210] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0211] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A coal mine safety risk identification system based on edge IoT, characterized in that, include: The module consists of a data acquisition module, a sub-analysis module, and a comprehensive analysis module. The acquisition module is used to acquire video data, equipment information, multimodal environmental data and personnel vital signs data in the coal mine, and to obtain a three-dimensional point cloud from the video data, convert the three-dimensional point cloud into a first three-dimensional coordinate system to generate an original three-dimensional model, transmit the original three-dimensional model to the main analysis module, and input the equipment information, multimodal environmental data and personnel vital signs data into the sub-analysis module. The sub-analysis module receives the device information, multimodal environmental data, and personnel vital sign data, and outputs the first early warning information according to the early warning mechanism, and transmits the first early warning information to the main analysis module; The overall analysis module receives the original 3D model, converts each original 3D model to the second 3D coordinate system, merges them into a real-time 3D model, marks the first warning information on the real-time 3D model, and outputs the second warning information according to the regional warning mechanism.

2. The coal mine safety risk identification system based on edge IoT as described in claim 1, characterized in that: The acquisition module is provided in multiple ways, and the acquisition module includes a video acquisition unit, a device information acquisition unit and a multimodal environment data acquisition unit respectively; The video acquisition unit uses a camera to collect video data from underground coal mines, and the video acquisition areas of each acquisition module have overlapping parts; The equipment information acquisition unit uses the wind speed sensor to collect the wind speed information of the ventilator and the flow meter to collect the flow information of the water pump, and records the acquisition time. The equipment information includes the wind speed information of the ventilator and the flow information of the water pump. The multimodal environmental data includes methane concentration data, carbon monoxide concentration data, oxygen concentration data, carbon dioxide concentration data, ethylene data, acetylene data, dust concentration data, temperature data, and humidity data. The multimodal environmental data acquisition unit includes a gas sensor, a carbon monoxide sensor, an oxygen sensor, a carbon dioxide sensor, an ethylene sensor, an acetylene sensor, a mine dust concentration sensor, and an intrinsically safe mine temperature and humidity sensor. The infrared gas sensor is used to collect gas concentration data; The carbon monoxide sensor is used to collect carbon monoxide concentration data; The oxygen sensor is used to collect oxygen concentration data; The carbon dioxide sensor is used to collect carbon dioxide concentration data; The ethylene sensor is used to collect ethylene data; The acetylene sensor is used to collect acetylene data; The mining dust concentration sensor is used to collect dust concentration data; The intrinsically safe temperature and humidity sensor for mining is used to collect temperature and humidity data, and to record the collection time of methane concentration data, carbon monoxide concentration data, oxygen concentration data, carbon dioxide concentration data, ethylene data, acetylene data, dust concentration data, temperature data, and humidity data, respectively. The vital signs data of the personnel include heart rate information, body temperature information, blood oxygen saturation and fatigue value. The heart rate information, body temperature information, blood oxygen saturation and fatigue value are collected using a smart bracelet, and the collection time of the heart rate information, body temperature information, blood oxygen saturation and fatigue value are recorded respectively.

3. The coal mine safety risk identification system based on edge IoT as described in claim 2, characterized in that: The first warning information includes Level 1 warning, Level 2 warning, and Level 3 warning; The sub-analysis module includes an equipment information analysis unit, a multimodal environmental data analysis unit, and a personnel vital sign data analysis unit; The equipment information analysis unit receives the collected wind speed and flow rate information; The multimodal environmental data analysis unit receives data on methane concentration, carbon monoxide concentration, oxygen concentration, carbon dioxide concentration, ethylene concentration, acetylene concentration, dust concentration, temperature, and humidity. The personnel vital signs data analysis receives and collects heart rate information, body temperature information, blood oxygen saturation, and fatigue value.

4. The coal mine safety risk identification system based on edge IoT as described in claim 3, characterized in that: The aforementioned early warning mechanism specifically includes: The equipment information analysis unit, multimodal environment data analysis unit, and personnel vital sign data analysis unit receive equipment information, multimodal environment data, and personnel vital sign data, and perform graded early warnings, outputting first early warning information. The first early warning information includes first-level, second-level, and third-level early warnings for personnel vital sign data, as well as first-level, second-level, and third-level early warnings for equipment information, and first-level, second-level, and third-level early warnings for multimodal environment data.

5. The coal mine safety risk identification system based on edge IoT as described in claim 4, characterized in that: The Level 1, Level 2, and Level 3 early warning systems for the personnel vital signs data include: The heart rate information is compared with the first heart rate threshold and the second heart rate threshold. If the heart rate information is lower than the first heart rate threshold, it indicates that the heart rate information is a level three warning. If the heart rate information is higher than the first heart rate threshold and lower than the second heart rate threshold, and the duration exceeds the first unit of time, it indicates that the heart rate information is a level two warning. If the heart rate information is higher than the second heart rate threshold and lasts for more than the first unit of time, it indicates that the heart rate information is a level one warning, and the second heart rate threshold is higher than the first heart rate threshold; The body temperature information is compared with the first body temperature threshold and the second body temperature threshold. If the body temperature information is lower than the first body temperature threshold, it indicates that the body temperature information is a level three warning. If the body temperature information is higher than the first body temperature threshold but lower than the second body temperature threshold, and the duration exceeds the second unit of time, it indicates that the body temperature information is a level two warning; If the body temperature information is higher than the second body temperature threshold and lasts for more than the second unit of time, it indicates that the body temperature information is a level one warning, and the second body temperature threshold is higher than the first body temperature threshold; The blood oxygen saturation information is compared with the first blood oxygen saturation threshold and the second blood oxygen saturation threshold. If the blood oxygen saturation information is lower than the first blood oxygen saturation threshold, it indicates that the blood oxygen saturation information is a level three warning. If the blood oxygen saturation information is higher than the first blood oxygen saturation threshold and lower than the second blood oxygen saturation threshold, and the duration exceeds the third unit of time, it indicates that the blood oxygen saturation information is a level two warning. If the blood oxygen saturation information is higher than the second blood oxygen saturation threshold and lasts for more than the third unit of time, it indicates that the blood oxygen saturation information is a level one warning, and the second blood oxygen saturation threshold is higher than the first blood oxygen saturation threshold. The fatigue value is compared with the first fatigue threshold and the second fatigue threshold. If the fatigue value is lower than the first fatigue threshold, it indicates that the fatigue value is a level three warning. If the fatigue value is higher than the first fatigue value threshold but lower than the second fatigue value threshold, and the duration exceeds the fourth unit of time, the fatigue value is considered a level two warning. If the fatigue value is higher than the second fatigue value threshold and lasts for more than four units of time, it indicates that the fatigue value is a level one warning, and the second fatigue value threshold is higher than the first fatigue value threshold.

6. The coal mine safety risk identification system based on edge IoT as described in claim 5, characterized in that: The device information's Level 1, Level 2, and Level 3 early warnings include: The wind speed information is compared with the first wind speed information threshold and the second wind speed information threshold. If the wind speed information is greater than the first wind speed information threshold, it indicates that the wind speed information is a level three warning. If the wind speed information is less than the first wind speed information threshold and greater than the second wind speed information threshold, and the duration exceeds the fifth unit of time, it indicates that the wind speed information is a level two warning. If the wind speed information is less than the second wind speed information threshold and the duration exceeds the fifth unit of time, it indicates that the wind speed information is a level one warning, and the second wind speed information threshold is greater than the first wind speed information threshold. The traffic information is compared with the first traffic information threshold and the second traffic information threshold. If the traffic information is greater than the first traffic information threshold, it indicates that the traffic information is a level three warning. If the traffic flow information is less than the first traffic flow information threshold and greater than the second traffic flow information threshold, and the duration exceeds the sixth unit of time, it indicates that the traffic flow information is a level two warning. If the traffic flow information is greater than the second traffic flow information threshold and the duration exceeds the sixth unit of time, it indicates that the traffic flow information is a level one warning, and the second traffic flow information threshold is greater than the first traffic flow information threshold.

7. The coal mine safety risk identification system based on edge IoT as described in claim 6, characterized in that: The first-level, second-level, and third-level early warnings for the multimodal environment data include: If ethylene and acetylene data are detected, it indicates that the ethylene and acetylene data are at the level of a Level 1 warning. The gas concentration is compared with the first gas concentration threshold and the second gas concentration threshold. If the gas concentration information is lower than the first gas concentration threshold, it indicates that the gas concentration information is a level three warning. If the gas concentration information is higher than the first gas concentration threshold but lower than the second gas concentration threshold, it indicates that the gas concentration information is at the level of a level two warning. If the gas concentration information is higher than the second gas concentration threshold, it indicates that the gas concentration information is at the first level of warning, and the second gas concentration threshold is higher than the first gas concentration threshold; The carbon monoxide concentration is compared with the first carbon monoxide concentration threshold and the second carbon monoxide concentration threshold. If the carbon monoxide concentration information is lower than the first carbon monoxide concentration threshold, it indicates that the carbon monoxide concentration information is a level three warning. If the carbon monoxide concentration information is higher than the first carbon monoxide concentration threshold but lower than the second carbon monoxide concentration threshold, it indicates that the carbon monoxide concentration information is a level two warning. If the carbon monoxide concentration information is higher than the second carbon monoxide concentration threshold, it indicates that the carbon monoxide concentration information is a level one warning, and the second carbon monoxide concentration threshold is higher than the first carbon monoxide concentration threshold. The carbon dioxide concentration is compared with a first carbon dioxide concentration threshold and a second carbon dioxide concentration threshold. If the carbon dioxide concentration information is lower than the first carbon dioxide concentration threshold, it indicates that the carbon dioxide concentration information is a level three warning. If the carbon dioxide concentration information is higher than the first carbon dioxide concentration threshold but lower than the second carbon dioxide concentration threshold, it indicates that the carbon dioxide concentration information is a level two warning. If the carbon dioxide concentration information is higher than the second carbon dioxide concentration threshold, it indicates that the carbon dioxide concentration information is a level one warning, and the second carbon dioxide concentration threshold is higher than the first carbon dioxide concentration threshold; The dust concentration is compared with the first dust concentration threshold and the second dust concentration threshold. If the dust concentration information is lower than the first dust concentration threshold, it indicates that the dust concentration information is a level three warning. If the dust concentration information is higher than the first dust concentration threshold but lower than the second dust concentration threshold, it indicates that the dust concentration information is a level two warning. If the dust concentration information is higher than the second dust concentration threshold, it indicates that the dust concentration information is a level one warning, and the second dust concentration threshold is higher than the first dust concentration threshold; The oxygen concentration is compared with the first oxygen concentration threshold and the second oxygen concentration threshold. If the oxygen concentration information is higher than the first oxygen concentration threshold, it indicates that the oxygen concentration information is a level three warning. If the oxygen concentration information is lower than the first oxygen concentration threshold but higher than the second oxygen concentration threshold, it indicates that the oxygen concentration information is a level two warning. If the oxygen concentration information is lower than the second oxygen concentration threshold, it indicates that the oxygen concentration information is in the first level warning stage, and the second oxygen concentration threshold is lower than the first oxygen concentration threshold. The temperature data is compared with a first temperature data threshold and a second temperature data threshold. If the temperature data is lower than the first temperature data threshold, it indicates that the temperature data is a level three warning. If the temperature data is higher than the first temperature data threshold and lower than the second temperature data threshold, and the duration exceeds the seventh unit of time, the temperature data is considered a level two warning. If the temperature data is higher than the second temperature data threshold and the duration exceeds the seventh unit of time, it indicates that the temperature data is a level one warning, and the second temperature data threshold is higher than the first temperature data threshold. The humidity data is compared with the humidity threshold range. If the humidity data is lower or higher than the humidity threshold range, it indicates that the humidity data is a level one warning.

8. The coal mine safety risk identification system based on edge IoT as described in claim 7, characterized in that: The first set of reference points is marked in advance in the underground coal mine and numbered once. The second set of reference points is marked in the same position on the original 3D model. The second set of reference points uses the same number as the first set of reference points. The overall analysis module assigns secondary numbering to the acquisition modules in sequence, receives the original 3D models sent by each acquisition module, converts each original 3D model to a second 3D coordinate system, and fuses the original 3D models sent by adjacent secondary-numbered acquisition modules.

9. The coal mine safety risk identification system based on edge IoT as described in claim 8, characterized in that: The fusion of the original 3D models sent by adjacent acquisition modules includes: The original 3D models sent by adjacent secondary numbering acquisition modules are matched with the same primary numbering information. The same primary number is used as the registration point to stitch the two original 3D models into one 3D model. This process continues until all original 3D models are stitched together to form a real-time 3D model. After the real-time 3D model is formed, a new original 3D model is received from the acquisition module. The same first number on the new original 3D model and the real-time 3D model is used as the registration point. The new original 3D model is then stitched onto the real-time 3D model to update the real-time 3D model.

10. The coal mine safety risk identification system based on edge IoT as described in claim 9, characterized in that: The first early warning information is annotated on the real-time 3D model, and the second early warning information is output according to the regional early warning mechanism, including: The same set of original 3D models, secondary numbers and first warning information are treated as a group of regions. Based on the correspondence between the original 3D models, secondary numbers and first warning information, the first warning information is marked on the real-time 3D model, and the second warning information is output. The second warning information includes Level 1, Level 2, and Level 3 warnings for personnel vital signs data, Level 1, Level 2, and Level 3 warnings for equipment information, and Level 1, Level 2, and Level 3 warnings for multimodal environment data. The first-level, second-level, and third-level warnings of equipment information will be directly output as the first-level, second-level, and third-level warnings of the second warning information of equipment information; The Level 1, Level 2, and Level 3 early warnings of personnel vital signs data are integrated and output as the Level 2 early warning information, which includes: If any two or more of the heart rate, body temperature, blood oxygen saturation, and fatigue values ​​show a Level II warning, the second warning information output for the personnel's vital signs data will be a Level II warning. Once the personnel's vital signs data reach a Level II warning, the staff member needs to rest and be observed. If any one of the heart rate, body temperature, blood oxygen saturation, or fatigue level reaches a Level 3 warning, the second warning information output for the personnel's vital signs data will be a Level 3 warning. After the personnel's vital signs data reaches a Level 3 warning, the staff member needs to undergo further physical examination. The first, second, and third level early warnings from multimodal environmental data are integrated and output as the second early warning information, which includes: If any two or more of the following data—gas concentration, carbon monoxide concentration, oxygen concentration, carbon dioxide concentration, dust concentration, temperature, and humidity—show a Level II warning, then the multimodal environmental data output for that area will be a Level II warning, indicating that work needs to be suspended for observation. If two or more adjacent groups of areas with secondary numbering subsequently issue a Level II warning, the multimodal environmental data output for the affected areas will be a Level III warning, requiring personnel evacuation and work stoppage. If any of the following data points triggers a Level 3 warning: methane concentration, carbon monoxide concentration, oxygen concentration, carbon dioxide concentration, ethylene concentration, acetylene concentration, or dust concentration, the multimodal environmental data output for that area will be a Level 3 warning, requiring personnel evacuation and work stoppage.

Images