Mine disaster risk assessment method based on Bayesian network and three-dimensional model

Abstract

The invention discloses a mine disaster risk assessment method based on a Bayesian network and a three-dimensional model, and the method comprises the steps: collecting first data, including gas data, temperature data, rock stratum stress data, smoke concentration data and sound signal data, of a preset position of a mine; screening out target data from the preprocessed first data through historical disaster-causing factors of the mine, and constructing a Bayesian network through the target data and a dependency relationship between different target data; performing three-dimensional scanning in a mine to obtain point cloud data; constructing a digital twinborn three-dimensional model of the mine based on the preprocessed point cloud data; acquiring second data at other positions based on the preprocessed first data, and integrating the first data and the second data according to different moments to obtain multi-dimensional data at different moments; inputting the data into a Bayesian network to obtain a three-dimensional risk distribution map and a probability value; and carrying out space mapping and calculation on the three-dimensional risk distribution diagram, the probability value and the digital twinborn three-dimensional model to obtain a mine disaster risk value.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent coal mine safety management and disaster prevention and control, and relates to, but is not limited to, a mine disaster risk assessment method based on a Bayesian network and a three-dimensional model. Background Art

[0002] Western China holds over 80% of the nation's coal reserves and accounts for approximately 60-70% of annual production. However, many coalfields in this region are prone to spontaneous combustion or prone to spontaneous combustion, resulting in frequent spontaneous combustion disasters. According to statistics, the area of ​​coalfield fires in Region 1 reached 19.03 million square meters, while Region 2 had 37 fires covering a total area of ​​3.9456 million square meters. In Region 3, there are 40 uncontrolled coalfield fires covering a total area of ​​4.7773 million square meters. Annual coal resource losses due to spontaneous combustion reach 4.5616 million tons. Spontaneous coal combustion, as the ignition source for gas combustion, explosions, and other disasters, is crucial for preventing complex disasters in coal mine goafs. Monitoring and early warning of spontaneous combustion of gas-laden coal in goafs requires addressing challenges such as multi-physics recognition of fire information, real-time collection of massive data, and analysis and application of multi-source heterogeneous data using intelligent algorithms.

[0003] Related technologies primarily rely on long-distance transmission via gas beam tubes, which suffer from long response times and insufficient reliability. Furthermore, remote gas sampling fails to fully account for the interference of complex environmental factors such as underground temperature and humidity on the measurement of target gas concentrations. This makes it difficult to meet the requirements for long-term, accurate monitoring and early warning of coal spontaneous combustion, leading to the inability to conduct timely and accurate assessments of mine disaster risks.

[0004] Therefore, how to quickly and accurately assess the risk of mine disasters has become an urgent problem to be solved. Summary of the Invention

[0005] In view of this, an embodiment of the present invention provides a mine disaster risk assessment method based on a Bayesian network and a three-dimensional model, which at least solves the problem that related technologies cannot quickly and accurately assess the risk of mine disasters.

[0006] According to a first aspect of an embodiment of the present invention, a mine disaster risk assessment method based on a Bayesian network and a three-dimensional model is provided, comprising:

[0007] collecting first data at a preset location of the mine, the first data including gas data, temperature data, rock formation stress data, smoke concentration data, and acoustic signal data;

[0008] Based on historical disaster-causing factors of the mine, target data are screened from the preprocessed first data, and a Bayesian network is constructed based on the target data and dependencies between different target data;

[0009] Performing a three-dimensional scan in the mine using a mobile three-dimensional laser scanner to obtain point cloud data; and constructing a digital twin three-dimensional model of the mine based on the pre-processed point cloud data;

[0010] acquiring second data of remaining positions based on the preprocessed first data, and integrating the first data and the second data at different times to obtain multidimensional data at different times; the remaining positions are positions other than the preset positions;

[0011] The multidimensional data is input into the Bayesian network to obtain a three-dimensional risk distribution map and a risk probability value of the mine; and the three-dimensional risk distribution map and the risk probability value are spatially mapped and calculated with the digital twin three-dimensional model to obtain a risk value of a disaster occurring in the mine.

[0012] According to a second aspect of an embodiment of the present invention, an electronic device is provided, comprising: a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other via the communication bus; the memory is used to store at least one executable instruction, and the executable instruction enables the processor to perform an operation corresponding to the method described in the first aspect.

[0013] According to a third aspect of an embodiment of the present invention, a computer storage medium is provided, on which a computer program is stored. When the program is executed by a processor, the method according to the first aspect is implemented.

[0014] According to the solution provided by an embodiment of the present invention, first data of a preset position of a mine is collected, the first data including gas data, temperature data, rock stress data, smoke concentration data and acoustic signal data; based on the historical disaster-causing factors of the mine, target data is screened out from the preprocessed first data, and a Bayesian network is constructed based on the target data and the dependency relationship between different target data; a three-dimensional scan is performed in the mine by a mobile three-dimensional laser scanner to obtain point cloud data; and a digital twin three-dimensional model of the mine is constructed based on the preprocessed point cloud data; second data of the remaining positions are obtained based on the preprocessed first data, and the first data and the second data are integrated at different times to obtain multidimensional data at different times; the remaining positions are positions other than the preset positions; the multidimensional data are input into the Bayesian network to obtain a three-dimensional risk distribution map and a risk probability value of the mine; and the three-dimensional risk distribution map and the risk probability value are spatially mapped and calculated with the digital twin three-dimensional model to obtain a risk value of a disaster in the mine. In this process, the second data is obtained from the first data, bridging the differences in sampling frequency and spatial location between different data, providing high-quality, continuous environmental parameter input for the digital twin 3D model and Bayesian network reasoning. A high-precision digital twin 3D model is constructed, supporting extended reality visualization technology to achieve immersive presentation of the mine's internal structure, facilitating disaster monitoring and emergency drills. Real-time data streaming and interface communication are established within the constructed digital twin 3D model, and a Bayesian network is embedded to achieve data-driven collaborative analysis and interactive visualization. The combination of the Bayesian network's reasoning capabilities and the intuitive display of the digital twin 3D model significantly enhances the digital twin's intelligent decision-making support capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. Those skilled in the art can also derive other drawings based on these drawings without inventive work, among which:

[0016] Figure 1 A schematic diagram of a flow chart of a mine disaster risk assessment method based on a Bayesian network and a three-dimensional model provided by an embodiment of the present invention;

[0017] Figure 2 A schematic diagram showing the effect of sensor layout in a mine and Bayesian network node variables provided by an embodiment of the present invention;

[0018] Figure 3 A schematic diagram of the effect of the constructed Bayesian network topology structure provided in an embodiment of the present invention;

[0019] Figure 4 A schematic structural diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments. The following embodiments are used to illustrate the present invention, but are not used to limit the scope of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0021] In the following description, reference is made to “some embodiments”, which describes a subset of all possible embodiments, but it will be understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0022] It should be pointed out that the terms "first\second\third" involved in the embodiments of the present invention are only used to distinguish similar objects and do not represent a specific ordering of the objects. It can be understood that "first\second\third" can be interchanged with a specific order or sequence where permitted, so that the embodiments of the present invention described here can be implemented in an order other than that illustrated or described here.

[0023] Those skilled in the art will understand that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by those skilled in the art in the art to which the embodiments of the present invention pertain. It should also be understood that terms such as those defined in common dictionaries should be understood to have meanings consistent with those in the context of the prior art, and will not be interpreted in an idealized or overly formal sense unless specifically defined as herein.

[0024] Figure 1 A flow chart of a mine disaster risk assessment method based on a Bayesian network and a three-dimensional model provided in an embodiment of the present invention is provided. A mine disaster risk assessment method based on a Bayesian network and a three-dimensional model provided in an embodiment of the present invention can be executed by an electronic device, such as a computer, a server, etc.

[0025] like Figure 1 As shown in FIG, the mine disaster risk assessment method based on Bayesian network and three-dimensional model includes:

[0026] S101. Collect first data at a preset location of a mine, where the first data includes gas data, temperature data, rock formation stress data, smoke concentration data, and acoustic signal data.

[0027] In an embodiment of the present invention, the preset locations may be high-risk areas such as goafs, working faces, and upper corners of tunnels in a mine. These locations can be preset in advance based on expert experience or historical records. Multiple sensors can be deployed at the preset locations in the mine to monitor real-time risk data for mine disasters. This risk data is derived from multiple sources, including but not limited to gas data, temperature data, rock formation stress data, smoke concentration data, and acoustic signal data.

[0028] Specifically, if Figure 2 As shown, Figure 2 A schematic diagram of the sensor layout in a mine and the effects of Bayesian network node variables provided by an embodiment of the present invention. Figure 2 The internal structure of the mine was first demonstrated, including the main shaft, auxiliary shaft, ventilation shaft, and mining face lights. The sensor layout and data collected include: ① Gas sensors are used to detect key gases in the mine, such as methane (CH4), carbon monoxide (CO), carbon dioxide (CO2), and oxygen (O2). They are located in the roof and upper corners of the goaf to monitor gas accumulation areas. ② Temperature sensors: Monitor temperature changes to identify early signs of coal spontaneous combustion. They are installed near hydraulic supports, shearers, coal conveyors, and other equipment to monitor equipment temperature. ③ Stress sensors: Monitor rock stress changes to prevent gas outbursts and coal seam collapse accidents. They are deployed on roadway support structures to monitor support pressure changes. ④ Microseismic sensors: Capture microseismic activity and provide early warning of collapses and gas dynamics. They are deployed in the rock strata surrounding the goaf to capture microseismic signals. ⑤ Photoelectric smoke detectors: Detect smoke concentration to quickly identify fires. They are deployed around the goaf to form a continuous temperature monitoring network. ⑥ Acoustic Emission Sensor: Monitors the acoustic signals generated by the development of cracks within the coal body; installed in the coal wall and roof areas, it captures the acoustic emission signals of the rock and coal seams. ⑦ Distributed Fiber Optic Temperature Sensor: Continuously monitors temperature changes along the fiber optic line.

[0029] In addition, personnel working within the working face can also wear portable gas detectors to monitor the gas concentration around them in real time, further generating gas data. Methane sensors are installed at the corners, nose, and tail of the working face. Temperature sensors are installed near hydraulic supports, shearers, coal conveyors, and other equipment to monitor equipment temperature. Microseismic sensors are placed near the coal walls of the working face to monitor microseismic activity in the coal body. Acoustic sensors are installed in the coal walls and roof areas to capture acoustic emission signals from the rock and coal seams. Y, D, and H in the diagram are described below.

[0030] S102. Based on historical disaster-causing factors of the mine, target data are screened out from the pre-processed first data, and a Bayesian network is constructed based on the target data and dependencies between different target data.

[0031] In an embodiment of the present invention, after the first data is collected, it can be transmitted back to the ground server in real time via the network and preprocessed to obtain the preprocessed first data. Based on the historical disaster-causing factors of the mine, target data is screened from the preprocessed first data. The target data can be 13 target data such as geological conditions, probability of fire source existence, and oxygen concentration level. The target data is used as key nodes, and the dependencies between the target data are used as edges to establish a Bayesian network topology structure for fire and gas coupled disasters, thereby forming a Bayesian network.

[0032] Preprocessing the first data includes data cleaning: removing outliers and noise. Data calibration: correcting measurement errors based on the sensor's calibration curve. Data synchronization: synchronizing the time of data from different sensors to ensure data consistency. Zero and span calibration is performed based on the sensor's calibration curve. Sliding average or median filtering is used to reduce the impact of random noise. Finally, data from different sensors is aligned to a common time interval Δt to establish a synchronized database.

[0033] Specifically, based on the mechanism of fire and gas coupling disasters and the causal relationship between influencing factors, when constructing the Bayesian topological structure through 13 target data (as shown in Table 1 below), geological conditions, equipment status, human factors, and coal spontaneous combustion tendency are used as root nodes; gas accumulation and fire source existence are used as core intermediate nodes; oxygen concentration, monitoring system, and early warning response system are used as intermediate nodes; gas explosion probability, fire probability, and coupled disaster probability are used as leaf nodes; these target data can be obtained through sensors, monitoring systems, or manual inspections, reflecting the direct or indirect causal relationship in the disaster chain. The Bayesian network topological structure (such as Figure 3 As shown, Figure 3 Schematic diagram of the effect of the constructed Bayesian network topology structure provided in an embodiment of the present invention).

[0034] Table 1 Target data of fire and gas disasters

[0035]

[0036]

[0037] Depend on Figure 3 It can be seen that the key causal chain of fire and gas coupling disasters is as follows:

[0038] Z→W→K→M

[0039] B→H→L→M

[0040] Q→H→L→M

[0041] R→H→L→M

[0042] W→T→Y→L→M

[0043] I→J→M

[0044] S103. Perform three-dimensional scanning in the mine using a mobile three-dimensional laser scanner to obtain point cloud data; and construct a digital twin three-dimensional model of the mine based on the pre-processed point cloud data.

[0045] In an embodiment of the present invention, a large amount of point cloud data can be obtained by performing three-dimensional scanning in a mine using a mobile three-dimensional laser scanner. The point cloud data may include coordinates (X, Y, Z) in three-dimensional space, and sometimes also include additional attributes such as color information (RGB values) or reflection intensity. The digital twin three-dimensional model includes not only the geometry of the mine, but also information about the properties of the ore body (such as mineral type and grade), ventilation system, and transportation channels. After obtaining the point cloud data, it is preprocessed, and then a visual digital twin three-dimensional model of the mine is constructed using the preprocessed point cloud data.

[0046] S104 , obtaining second data of remaining positions based on the preprocessed first data, and integrating the first data and the second data at different times to obtain multidimensional data at different times; the remaining positions are positions other than the preset positions.

[0047] In some embodiments of the present invention, the remaining locations are all locations in the high-risk area except the preset locations. The first data are data collected by different sensors at different times. For the uncovered spatial areas (the remaining locations), a weighted spatial interpolation method is used to calculate the estimated values ​​of the environmental parameters, i.e., the second data. The second data are then arranged according to time and integrated into the gaps in the existing first data. All the data are integrated to obtain multi-dimensional data at different times, so that the obtained spatial data sequence is more reasonable.

[0048] S105. Input the multidimensional data into the Bayesian network to obtain the three-dimensional risk distribution map and risk probability value of the mine; and spatially map and calculate the three-dimensional risk distribution map and risk probability value with the digital twin three-dimensional model to obtain the risk value of the mine disaster.

[0049] In an embodiment of the present invention, multidimensional data is input into a Bayesian network, and based on the Bayesian network's analysis results, a three-dimensional risk map can be generated that reflects the risk level of each area within the mine. This map not only shows the spatial distribution of risk but also provides a risk probability value for each location. The risk value of a mine disaster can include the risk value of a disaster at each location in the mine, as well as the risk value of a disaster for the entire mine. The generated three-dimensional risk map and the corresponding risk probability values ​​are spatially mapped to the three-dimensional geometric model of the mine, accurately locating the risk information at the corresponding location on the mine model, and further calculating the risk value of a mine disaster.

[0050] It is understood that in an embodiment of the present invention, first data is collected at a preset location in the mine, the first data including gas data, temperature data, rock formation stress data, smoke concentration data, and acoustic signal data; based on the historical disaster-causing factors of the mine, target data is filtered from the preprocessed first data, and a Bayesian network is constructed based on the dependency relationship between the target data and the target data; a mobile 3D laser scanner is used to perform a 3D scan of the mine to obtain point cloud data; a digital twin 3D model of the mine is constructed based on the preprocessed point cloud data; second data at other locations is obtained based on the preprocessed first data, and the first and second data are integrated at different times to obtain multidimensional data at different times; the multidimensional data is input into the Bayesian network to obtain a 3D risk distribution map and risk probability value for the mine; the 3D risk distribution map and probability value are spatially mapped and calculated with the digital twin 3D model to obtain a risk value for a disaster in the mine. In this process, obtaining the second data from the first data bridges the sampling frequency and spatial location differences of the different data, providing high-quality, continuous environmental parameter input for the digital twin 3D model and Bayesian network reasoning. Build a high-precision digital twin 3D model, support extended reality visualization technology, realize immersive presentation of the internal structure of the mine, facilitate disaster monitoring and emergency drills, establish real-time data flow and interface communication in the constructed digital twin 3D model, embed Bayesian network, realize data-driven collaborative analysis and interactive visualization, and combine the reasoning ability of the Bayesian network with the intuitive display of the digital twin 3D model to significantly enhance the intelligent decision-making support capability of the digital twin.

[0051] In some embodiments of the present invention, constructing a digital twin three-dimensional model of a mine based on pre-processed point cloud data in S103 can be implemented through S1031, which is explained through the following steps.

[0052] S1031. Construct an initial digital twin three-dimensional model through the preprocessed point cloud data, and optimize the initial digital twin three-dimensional model through the acquired geological radar data to obtain a digital twin three-dimensional model.

[0053] In some embodiments of the present invention, a mobile 3D laser scanner is used to perform 3D scanning in a mine, combining laser radar technology and simultaneous positioning and mapping technology. The mobile 3D laser scanner is used to collect point cloud data in the goaf, main tunnel, and mining face. The point cloud data collected at intervals of d are recorded as V1, V2, ..., V n , the collected point cloud data is recorded as V1(x,y,z),V2(x+d,y,z),…,V n (x+(n-1)d,y,z), where (x,y,z) are the spatial coordinates of the point cloud data and n is the total number of collected point cloud data. An initial digital twin 3D model is constructed using preprocessed point cloud data. Combined with mine geological exploration data or geological radar data, the initial digital twin 3D model is calibrated and supplemented with information such as coal seam thickness, roof lithology, and fault locations to create a digital twin 3D model.

[0054] In some embodiments of the present invention, S103 also includes S1031 to S1032 before S103, which is explained through the following steps.

[0055] S1031 . Perform filtering processing on the point cloud data to obtain first point cloud data, and perform registration processing on the first point cloud data to obtain second point cloud data.

[0056] S1032: Perform gridding processing on the second point cloud data to obtain pre-processed point cloud data.

[0057] For example, filtering, as the first step in point cloud data processing, can filter out and correct a large amount of noise generated by the influence of surrounding objects and the environment when performing 3D laser scanning of the on-site space, while preserving the geometric characteristics of the original point cloud data. The point cloud bilateral filtering method considers the influence of neighboring points and moves the point cloud data along the normal vector direction to achieve filtering. By setting a reasonable weight function, the edge information can be maintained while smoothing the data. The bilateral filtering calculation formula is:

[0058] V n '=V n +μ*λ n

[0059] In the above formula, V n is the point cloud data; V′ n is the first point cloud data; n It is V n The normal vector of ; μ is the filter factor, specifically:

[0060]

[0061] In the above formula, y n It is V nThe domain point set of g c 、g s is the Gaussian function filter coefficient; N c (x) and N s (x) is the weight function, x is ||V n -y n ||, as shown in the formula:

[0062]

[0063] The processed first point cloud data has rotation and translation transformations in different coordinate systems. The registration algorithm based on geometric feature description solves the corresponding rotation matrix and translation matrix to calculate the coordinates of the first point cloud data, making the point cloud registration high in accuracy and robustness, and the calculation speed faster than the global search method. Therefore, point cloud registration can integrate the first point cloud data in different coordinate systems through the transformation matrix. In three-dimensional space, the rotation matrix R and translation matrix T can be expressed as:

[0064]

[0065] T=[t x ,t y ,t z ] T

[0066] In the above formula, α, β, and γ represent the rotation angles of the first point cloud data along the X, Y, and Z axes; t x , t y , t z Indicates the displacement from the first point cloud data to the second point cloud data.

[0067] Furthermore, after obtaining the second point cloud data through the rotation matrix R and the translation matrix T, the extended reality technology is used to convert the file of the second point cloud data into OBJ and STL formats that can be recognized by Paraview, and then the point cloud data is meshed using an irregular triangular mesh to obtain the preprocessed point cloud data.

[0068] In some embodiments of the present invention, obtaining the second data at the remaining positions based on the preprocessed first data in S104 can be implemented through S1041 to S1042, which is explained in the following steps.

[0069] S1041. Obtain the spatial positions and weights of all sensors corresponding to the first data collected, and calculate the spatial distances between the current position and all sensors based on the spatial positions; the current position is any position among the remaining positions.

[0070] In one embodiment of the present invention, the first data includes multiple different data, each of which is acquired by different sensors. The spatial positions and corresponding weights of all sensors are acquired, and the spatial distance between each of the remaining positions and the sensor is calculated. Taking the current position as an example, the calculation is as follows:

[0071]

[0072] In the above formula, (x, y, z) is the current position, i is the subscript of the sensor, and d is the current position. i is the spatial distance between the current position and the i-th sensor.

[0073] S1042: Calculate weight functions corresponding to all sensors according to the spatial distance and the weight, and calculate second data corresponding to the current position based on the weight function and the first data.

[0074] In some embodiments of the present invention, the spatial distance and weight are substituted into the weight function formula to obtain the weight function corresponding to all sensors, and the weight function and the first data are substituted into the second data calculation formula to obtain the second data corresponding to the current position.

[0075] The weight function formula is as follows:

[0076]

[0077] In the above formula, w i (x, y, z) is the weight function corresponding to the i-th sensor, γ is the distance attenuation coefficient, which controls the attenuation speed of the weight with distance, and w si is the weight of the i-th sensor, It refers to the spatial distance between the current position and the i-th sensor under the γ attenuation index.

[0078] The second data calculation formula is as follows:

[0079]

[0080] In the above formula, U(x,y,z) is the second data corresponding to the current position, v i is the first data corresponding to the i-th sensor.

[0081] In some embodiments of the present invention, spatial mapping and calculation of the three-dimensional risk distribution map and the risk probability value with the digital twin three-dimensional model in S105 to obtain the risk value of the mine disaster can be achieved through S1051 to S1052, which is explained in the following steps.

[0082] S1051. Spatially map the three-dimensional risk distribution map and the risk probability value with the digital twin three-dimensional model to obtain the initial risk value of each position on the digital twin three-dimensional model.

[0083] S1052. Obtain a critical value for disaster occurrence at each location, and obtain a risk value for mine disaster occurrence based on the initial risk value and the critical value.

[0084] In some embodiments of the present invention, the critical value is the maximum of the corresponding values ​​of the first and second collected data. The three-dimensional risk distribution map and risk probability values ​​are spatially mapped to the digital twin three-dimensional model to obtain an initial risk value for each location on the digital twin three-dimensional model. The initial risk value and the critical value are then calculated to obtain the final disaster risk value for each location and the comprehensive risk value for a mine disaster.

[0085] For example, the gas explosion risk R 瓦 =(x,y,z) can be expressed as the ratio of gas concentration to critical concentration:

[0086]

[0087] Among them, C CH4 (x, y, z) is the methane concentration at the position (x, y, z); C 临界 is the critical concentration of gas explosion (usually 5%). 瓦斯 (x,y,z)>R 瓦斯临界 (x,y,z), it means there is a risk of gas explosion.

[0088] Coal spontaneous combustion fire risk R 火 (x,y,z) can be calculated based on temperature and carbon monoxide concentration:

[0089]

[0090] Where, T(x,y,z) is the temperature at the position (x,y,z); T 基准 is the ambient reference temperature; T 临界 C is the critical temperature of coal spontaneous combustion; co (x, y, z) is the carbon monoxide concentration; C 临界 is the critical value of carbon monoxide concentration; α T and α co is the weight coefficient, satisfying α T +α co = 1. When R 火 (x,y,z)>R 火临界 (x,y,z), there is a risk of coal spontaneous combustion disaster.

[0091] Roof collapse risk R 顶(x,y,z) is calculated based on stress and microseismic signal intensity:

[0092]

[0093] Among them, σ(x,y,z) is the stress value; σ 临界 is the critical stress of the roof rock layer; E 微震 (x, y, z) is the microseismic energy; E max is the maximum value of microseismic energy; β σ and β E is the weight coefficient, satisfying β σ +β E = 1. When R 顶 (x,y,z)>R 顶临界 (x,y,z), there is a risk of roof collapse.

[0094] In some embodiments of the present invention, S1052 may be implemented through S301, which is explained through the following steps.

[0095] S301. Obtain a risk value for each location in the mine based on an initial risk value and a critical value, and obtain a risk value for a disaster occurring in the mine based on the risk value of each location.

[0096] For example, the risk value of each location is integrated to obtain the comprehensive risk R 综合 (x,y,z):

[0097] R 综合 (x,y,z)=γ 瓦 ×R 瓦 +γ 火 ×R 火 +γ 顶 ×R 顶

[0098] In the above formula, γ 瓦 , γ 火 , γ 顶 is the weight coefficient of disaster risk, satisfying γ 顶 +γ 火 +γ 瓦 =1.

[0099] In some embodiments of the present invention, S1052 further includes S401 to S402, which is explained through the following steps.

[0100] S401: When the risk value of any position is not less than the corresponding first threshold, trigger an early warning.

[0101] S402: When the risk value of a mine disaster is not less than a second threshold, trigger an early warning.

[0102] In some embodiments of the present invention, the first threshold can be the critical value described above, with thresholds set separately for gas explosion risk, coal spontaneous combustion fire risk, and roof collapse risk. When the risk value at any location or the risk value of a mine disaster exceeds the predetermined threshold, a Level 3 or higher alarm is automatically issued, quickly notifying on-site management personnel through audible and visual alarms, pop-up prompts on the ground monitoring platform, and mobile phone app notifications. This early warning may include measures such as ventilation volume control, localized cooling, gas extraction, or inert gas injection.

[0103] Specifically, the operating status of the local ventilation fan or the main ventilation fan is automatically adjusted according to the gas concentration and temperature distribution; if local gas accumulation accompanied by high temperature is detected, fire extinguishing is carried out through drilling extraction or nitrogen / carbon dioxide injection; in high-risk conditions, the underground power distribution system can be linked to cut off the power to non-essential equipment, and direct personnel to evacuate through the nearest safe passage.

[0104] For example, the risk value of each location is divided into five levels: safe P1 (0-0.75), relatively safe P2 (0.75-0.85), generally safe P3 (0.85-0.90), relatively unsafe P4 (0.9-0.95), and unsafe P5 (0.95-1). The risk value of each location is compared with P1 to P5. A value in the P4 range indicates relatively unsafe conditions, requiring increased attention and preventive measures to prevent disasters. A value in the P5 range indicates unsafe conditions, requiring immediate emergency rescue measures.

[0105] The effectiveness of preventive control measures is continuously monitored through sensors. If the risk value begins to decrease, the ventilation or extraction volume is reduced in a timely manner to save resources. New monitoring data and field disposal information are collected during operation, and the prior and conditional probabilities of the Bayesian network are continuously corrected online to achieve robust response to abnormal and faulty sensors.

[0106] In an embodiment of the present invention, a risk assessment system for a mine based on a Bayesian network and a three-dimensional model is also proposed, comprising a data perception layer, a transmission and data integration layer, a Bayesian network inference layer, and a risk estimation and control layer. The data perception layer is used to collect first data through the deployed data perception layer, the transmission and data integration layer is used to transmit the first data and integrate the first data and the second data, the Bayesian network inference layer is used to infer the three-dimensional risk distribution map and risk probability value of the mine based on the integrated multi-dimensional data, and the risk estimation and control layer user calculates the final risk value of the mine and provides specific control measures. Bidirectional interaction of data and instructions is formed between the layers through wired or wireless networks. When the data perception layer detects an abnormal or faulty sensor, it compensates and interpolates based on the remaining sensors and historical data to ensure the continuity of the model. The modeling platform supports XR technology, providing managers or rescuers with immersive visual browsing of the mine structure and intuitively displaying high-risk areas or disaster evolution simulations in virtual scenes.

[0107] like Figure 4 As shown, the electronic device may include: a processor (processor) 502, a communications interface (Communications Interface 504), a memory (memory) 506, and a communication bus 508.

[0108] in:

[0109] The processor 502 , the communication interface 504 , and the memory 506 communicate with each other via a communication bus 508 .

[0110] The communication interface 504 is used to communicate with other electronic devices or servers.

[0111] The processor 502 is configured to execute the program 510 , and specifically may execute the relevant steps in the above method embodiment.

[0112] Specifically, the program 510 may include program codes, which include computer operation instructions.

[0113] Processor 502 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present invention. The one or more processors included in a smart device may be processors of the same type, such as one or more CPUs, or processors of different types, such as one or more CPUs and one or more ASICs.

[0114] The memory 506 is used to store the program 510. The memory 506 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0115] The program 510 may be specifically configured to enable the processor 502 to execute operations corresponding to the methods described in the above method embodiments.

[0116] The specific implementation of each step in program 510 can be found in the corresponding descriptions of the corresponding steps and units in the above-mentioned method embodiments, and will not be repeated here. Those skilled in the art will clearly understand that for the convenience and brevity of description, the specific working processes of the above-mentioned devices and modules can refer to the corresponding process descriptions in the above-mentioned method embodiments, and will not be repeated here.

[0117] It should be pointed out that, according to the needs of implementation, the various components / steps described in the embodiments of the present invention can be split into more components / steps, or two or more components / steps or partial operations of components / steps can be combined into new components / steps to achieve the purpose of the embodiments of the present invention.

[0118] The method according to the embodiment of the present invention described above can be implemented in hardware, firmware, or as software or computer code that can be stored in a recording medium (such as a CD ROM, RAM, floppy disk, hard disk or magneto-optical disk), or as computer code that is originally stored in a remote recording medium or a non-temporary machine-readable medium downloaded via a network and will be stored in a local recording medium, so that the method described herein can be stored in such software processing on a recording medium using a general-purpose computer, a dedicated processor or programmable or dedicated hardware (such as an ASIC or FPGA). It can be understood that a computer, a processor, a microprocessor controller or programmable hardware includes a storage component (e.g., RAM, ROM, flash memory, etc.) that can store or receive software or computer code, and when the software or computer code is accessed and executed by a computer, a processor or hardware, the method described herein is implemented. In addition, when a general-purpose computer accesses the code for implementing the method shown here, the execution of the code converts the general-purpose computer into a dedicated computer for executing the method shown here.

[0119] Those skilled in the art will appreciate that the units and method steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the embodiments of the present invention.

[0120] The above implementation methods are only used to illustrate the embodiments of the present invention, and are not intended to limit the embodiments of the present invention. Ordinary technicians in the relevant technical field may make various changes and modifications without departing from the spirit and scope of the embodiments of the present invention. Therefore, all equivalent technical solutions also fall within the scope of the embodiments of the present invention, and the scope of patent protection of the embodiments of the present invention should be defined by the claims.

Claims

1. A mine disaster risk assessment method based on Bayesian network and three-dimensional model, characterized in that: include: collecting first data at a preset location of the mine, the first data including gas data, temperature data, rock formation stress data, smoke concentration data, and acoustic signal data; Based on historical disaster-causing factors of the mine, target data are screened from the preprocessed first data, and a Bayesian network is constructed based on the target data and dependencies between different target data; Performing three-dimensional scanning in the mine using a mobile three-dimensional laser scanner to obtain point cloud data; A digital twin 3D model of the mine is constructed based on the pre-processed point cloud data; Acquire second data at other locations based on the preprocessed first data, and integrate the first data and the second data at different times to obtain multidimensional data at different times; The remaining positions are positions other than the preset positions; Inputting the multidimensional data into the Bayesian network to obtain a three-dimensional risk distribution map and a risk probability value of the mine; The three-dimensional risk distribution map and the risk probability value are spatially mapped and calculated with the digital twin three-dimensional model to obtain the risk value of the mine disaster.

2. The method according to claim 1, characterized in that The method of constructing a digital twin 3D model of a mine based on pre-processed point cloud data includes: An initial digital twin three-dimensional model is constructed using the preprocessed point cloud data, and the initial digital twin three-dimensional model is optimized using the acquired geological radar data to obtain the digital twin three-dimensional model.

3. The method according to claim 1, characterized in that Before constructing the digital twin 3D model of the mine based on the pre-processed point cloud data, the method further includes: Performing filtering processing on the point cloud data to obtain first point cloud data, and performing registration processing on the first point cloud data to obtain second point cloud data; The second point cloud data is meshed to obtain the pre-processed point cloud data.

4. The method according to claim 1, wherein The acquiring second data at other locations based on the preprocessed first data includes: Obtaining the spatial positions and weights of all sensors corresponding to the first data collected, and calculating the spatial distances between the current position and all the sensors based on the spatial positions; the current position is any position among the remaining positions; A weight function corresponding to each of the sensors is calculated according to the spatial distance and the weight, and second data corresponding to the current position is calculated based on the weight function and the first data.

5. The method according to claim 1, wherein The three-dimensional risk distribution map and the risk probability value are spatially mapped and calculated with the digital twin three-dimensional model to obtain the risk value of the mine disaster, including: Spatially mapping the three-dimensional risk distribution map and the risk probability value with the digital twin three-dimensional model to obtain an initial risk value for each position on the digital twin three-dimensional model; A critical value of disaster occurrence at each location is obtained, and a risk value of disaster occurrence at the mine is obtained based on the initial risk value and the critical value.

6. The method according to claim 5, characterized in that The step of obtaining the risk value of the mine disaster based on the initial risk value and the critical value includes: Obtaining a risk value for each location of the mine based on the initial risk value and the critical value, and obtaining a risk value for a disaster occurring in the mine based on the risk value of each location; After obtaining the risk value of the mine disaster based on the initial risk value and the critical value, the method further includes: When the risk value of any of the locations is not less than the corresponding first threshold, triggering an early warning; When the risk value of a disaster occurring in the mine is not less than a second threshold, an early warning is triggered.

Images