Coal mine hidden disaster-causing multi-source factor superposition disaster-causing risk prediction method

Abstract

The invention discloses a coal mine hidden disaster-causing multi-source factor superposition disaster-causing risk prediction method, and relates to the technical field of coal mine construction risk prediction. The method comprises the following steps of S01, information acquisition and processing, S02, risk assessment model construction, S03, risk early warning result display, S04, accident risk assessment, S05, risk information three-dimensional visualization, S06, emergency rescue resource management and S07, monitoring and early warning technology application. According to the method, through utilization of a GIS information acquisition platform, multi-source data of a coal mine can be effectively integrated, and the risk of the coal mine can be effectively analyzed; comprising production monitoring, geographic information, safety management and control and the like; the method overcomes the defect of incomplete data acquisition in the prior art, and ensures that the risk assessment can be performed based on comprehensive and accurate data, thereby improving the monitoring capability of the hidden disaster-causing factors of the coal mine.

Description

Technical Field

[0001] The present invention relates to the technical field of coal mine construction risk prediction, and in particular to a method for predicting the risk of coal mine hidden disasters caused by the superposition of multiple source factors. Background Art

[0002] With the rapid development of the coal industry, coal mine safety issues have become increasingly prominent. Hidden disaster factors in coal mines, such as gas explosions, water damage, roof accidents, etc., have brought great challenges to coal mine safety production due to their concealment and complexity. Traditional coal mine safety monitoring and risk assessment methods often rely on manual inspections and experience judgments. These methods have the following main disadvantages:

[0003] Incomplete data collection: Traditional monitoring methods can usually only collect limited on-site data, such as gas concentration and temperature, but cannot fully collect multi-source information of coal mines, including production monitoring data, geographic information system (GIS) data, safety management data, etc. This limitation of data collection leads to incomplete risk assessment and cannot accurately reflect the real-time safety status of coal mines.

[0004] Risk assessment lacks real-time and dynamic features: Due to the limitations of data collection and insufficient processing capabilities, traditional risk assessment methods often cannot achieve real-time monitoring and dynamic updates. This means that when the coal mine environment changes, the risk assessment results may be outdated and cannot provide timely risk warnings and decision-making support for mine managers.

[0005] In order to overcome these shortcomings, the present invention proposes a method for predicting the risk of hidden disasters caused by multiple sources of superimposed disasters in coal mines. The method realizes comprehensive monitoring and real-time risk assessment of hidden disaster factors in coal mines by integrating GIS information collection platform, data mining technology, intelligent early warning model construction, multi-source information fusion, evidence theory and confidence allocation rule establishment, thereby improving the early warning capability and management efficiency of coal mine safety production. Summary of the invention

[0006] In view of the shortcomings of the prior art, the present invention provides a method for predicting the risk of coal mine hidden disasters caused by the superposition of multiple sources of disaster factors, which solves the problems of incomplete data collection in the prior art and lack of real-time and dynamic risk assessment mentioned in the background technology.

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions: a method for predicting the risk of coal mine hidden disasters caused by multiple sources of superposition, the steps of which are as follows:

[0008] S01. Information collection and processing: Use the GIS information collection platform to effectively collect and process coal mine hazard source information, including basic enterprise information, production monitoring information and safety management information;

[0009] S02. Risk assessment model construction: Construct a coal mine risk warning model, conduct data mining analysis, algorithm operation analysis, comprehensive prediction analysis on the collected information, and draw risk warning conclusions;

[0010] S03. Risk warning result display: Use GIS analysis and display platform to conduct comprehensive visual warning and display of warning results, and provide auxiliary decision-making support for mine safety production scheduling;

[0011] S04. Accident risk assessment: By cleaning, integrating and filtering the collected data, and combining it with the risk assessment grading table constructed by the coal mine risk warning model, the mean algorithm of data mining is used to realize clustering law discovery and capture the comprehensive information of coal mine accidents;

[0012] S05. 3D visualization of risk information: 3D visualization of coal mine risk warning conclusions on GIS maps, and display of information related to mine risk levels, categories, status details, and risk disposal through a variety of thematic charts;

[0013] S06. Emergency rescue resource management: GIS comprehensive analysis and map visualization of rescue resources, rescue plans, rescue routes and rescue strategies, including rescue resource management analysis, rescue plan analysis, and rescue decision analysis;

[0014] S07. Application of monitoring and early warning technology: Apply the Internet of Things, big data, cloud computing and artificial intelligence technologies to realize intelligent monitoring and early warning of coal mines, improve the reliability of precursor information collection, build a coal mine monitoring database based on blockchain technology, and develop a coal mine intelligent safety monitoring and early warning system with self-adaptation and deep learning.

[0015] Preferably, the steps of constructing a coal mine risk warning model in S02 are as follows:

[0016] Data preprocessing: Collect multi-source data related to coal mines, and perform cleaning and normalization to improve data quality;

[0017] Data mining analysis: Use data mining techniques (such as association rules and cluster analysis) to analyze risk factors; in association rule analysis, calculate support (S) and confidence (C), determine the minimum support (S min ) and minimum confidence (C min ) threshold: Among them, T(B) represents the support of a certain rule, T represents the total number of transactions, and κ is the minimum confidence threshold;

[0018] Intelligent early warning model construction: Select a suitable algorithm, including neural network or decision tree, to build a risk early warning model and conduct training. During the model training process, use the strong association rules selected by support and confidence to determine key indicators.

[0019] Multi-source information fusion: Fusion of data from different sources to improve the accuracy and robustness of the early warning model; for quantitative indicators, according to the critical values ​​of indicators obtained by association rule analysis, the indicators are divided into three intervals and basic confidence levels are assigned: Z n ∈[Z0,+∞,Z n ∈[Z min ,Z0,Z n ∈(-∞,Z min Among them, Z0 is the critical value of the indicator, Z min is the interval partition coefficient;

[0020] Establishment of evidence theory and confidence allocation rules: Establish an evidence theory identification framework for highlighting early warning analysis, and determine the basic confidence allocation rules based on the results of association analysis; for the synthesis and decision-making of early warning results, use the following formula for evidence synthesis:

[0021]

[0022] Where K is the normalization constant;

[0023] Use a probability function-like method to make decisions:

[0024]

[0025] Where |X| is the basis of the focal element, and |U| is the basis of the identification frame;

[0026] Model verification and optimization: Verify the model through historical data and field feedback, and automatically adjust the model parameters according to the verification results to achieve dynamic tuning of the model;

[0027] Release of early warning results: The early warning results of the model are released in real time through the coal mine monitoring system so that preventive measures can be taken.

[0028] Preferably, the multi-source data related to coal mines mainly include:

[0029] Production monitoring data: real-time sensor data, including temperature, humidity, and pressure; Geographic Information System (GIS) data: including spatial information on geographic location, topography, and geological structure;

[0030] Safety control data: including safety monitoring data of gas, fire and roof;

[0031] Manufacturing Execution System (MES) data: including production management data on mining progress, equipment status, and material consumption;

[0032] Dispatch management data: including personnel, materials, and equipment dispatch information;

[0033] Enterprise Resource Planning (ERP) data: including enterprise management data on finance, human resources, and supply chain;

[0034] Environmental and meteorological data: the impact of meteorological conditions on coal mine safety and production;

[0035] Video and audio data: underground monitoring video and audio records;

[0036] Drawing and document data: including design drawings, geological reports, and safety regulations;

[0037] After processing, the data is used to build a coal mine risk warning model to achieve safety monitoring and risk assessment.

[0038] Preferably, the key indicators are variables or attributes that appear frequently in the data set and have strong correlation; the four important key indicators screened out based on support and confidence include: gas concentration, groundwater level, roof stability, and equipment operating status.

[0039] Preferably, in the model verification and optimization, the model parameters adjusted include:

[0040] Sensor parameters: basic measurements related to coal mine safety monitoring systems, such as set values ​​and thresholds for methane concentration and carbon monoxide concentration;

[0041] Model hyperparameters: including learning rate, number of hidden layer neurons, number of iterations, etc. in machine learning models. These parameters directly affect the learning ability and prediction accuracy of the model.

[0042] Warning threshold: including alarm threshold, power-off threshold, and power-restore threshold. This parameter determines when to trigger warning measures, and is related to risk management and emergency response;

[0043] Time series parameters: involve data sampling frequency and time window size, which affect the timeliness of data and the model's ability to respond to dynamic changes;

[0044] Spatial parameters: For GIS data, spatial resolution and coordinate system parameters affect the ability to accurately locate risk areas and perform spatial analysis.

[0045] Preferably, the formula for constructing the coal mine accident risk assessment grading table in S04 is as follows:

[0046] For underground coal mines, the calculation formula for the safety production standardization assessment score M is:

[0047]

[0048] Among them, M i represents the safety production standardization assessment score of the i-th part, a i Represents the weight value of the i-th part;

[0049] For open-pit coal mines, the calculation formula for the safety production standardization assessment score N is:

[0050]

[0051] Among them, N i represents the safety production standardization assessment score of the i-th part, b i Represents the weight value of the i-th part. Through calculation, the safety production standardization level of the coal mine can be quantitatively evaluated, and then a risk assessment grading table can be constructed to provide a scientific basis for coal mine safety management.

[0052] Preferably, the method of using the mean algorithm of data mining to realize clustering rules in S04 is:

[0053] Step 1: For each data point x in the dataset j and cluster center c i , calculate the Euclidean distance between them:

[0054] Where D is the dimension of the data, x j,d and c i,d The data points x are j and cluster center c i The value in the dth dimension;

[0055] Step 2: Assign data points to the nearest cluster center: Each data point is assigned to the nearest cluster center, that is:

[0056]

[0057] Among them, S i is assigned to the cluster center c i A collection of data points;

[0058] Step 3: Update the cluster center: For each cluster S i , update the cluster center c i is the average value of all data points in the cluster:

[0059]

[0060] Among them, |S i | is cluster S i The number of data points in ;

[0061] Step 4: Iterate until convergence: Repeat steps 2 and 3 until the cluster center no longer changes significantly or the preset number of iterations is reached.

[0062] Preferably, in S05, risks are divided into different levels according to risk assessment values ​​and represented by different colors. For example, a risk assessment value in the range of [0,1.5] is a low risk, represented by blue; 1.5,2.6] is a general risk, represented by yellow; 2.6,3.2] is a greater risk, represented by orange; 3.2,4.0] is a major risk, represented by red.

[0063] The present invention provides a method for predicting the risk of coal mine hidden disasters caused by multiple sources of superposition of disasters. It has the following beneficial effects:

[0064] (1) Comprehensive data collection and processing: This method can effectively integrate multi-source data of coal mines, including production monitoring, geographic information, safety control, etc., by utilizing the GIS information collection platform. This method overcomes the defect of incomplete data collection in traditional technologies and ensures that risk assessment can be based on comprehensive and accurate data, thereby improving the monitoring ability of hidden disaster-causing factors in coal mines.

[0065] (2) Real-time dynamic risk assessment: The coal mine risk warning model constructed by this method can process the collected data in real time and update it dynamically. This technological innovation solves the problem that traditional risk assessment lacks real-time and dynamic characteristics, allowing mine managers to obtain the latest risk assessment results in a timely manner, thereby making quick and effective decisions and enhancing the safety management capabilities of coal mines.

[0066] (3) 3D visualization and emergency management support: By visualizing the risk warning results in 3D, this method not only improves the readability and intuitiveness of risk information, but also provides effective auxiliary decision support for mine safety production scheduling. This visualization can help managers better understand the risk distribution and status, thereby optimizing the allocation and management of emergency rescue resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0068] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0069] See also Figure 1 The embodiment of the present invention provides a technical solution: a method for predicting the risk of coal mine hidden disasters caused by multiple sources of superposition, the steps of which are as follows:

[0070] S01. Information collection and processing: Use the GIS information collection platform to effectively collect and process coal mine hazard source information, including basic enterprise information, production monitoring information and safety management information;

[0071] S02. Risk assessment model construction: Construct a coal mine risk warning model, conduct data mining analysis, algorithm operation analysis, comprehensive prediction analysis on the collected information, and draw risk warning conclusions;

[0072] S03. Risk warning result display: Use GIS analysis and display platform to conduct comprehensive visual warning and display of warning results, and provide auxiliary decision-making support for mine safety production scheduling;

[0073] S04. Accident risk assessment: By cleaning, integrating and filtering the collected data, and combining it with the risk assessment grading table constructed by the coal mine risk warning model, the mean algorithm of data mining is used to realize clustering law discovery and capture the comprehensive information of coal mine accidents;

[0074] S05. 3D visualization of risk information: 3D visualization of coal mine risk warning conclusions on GIS maps, and display of information related to mine risk levels, categories, status details, and risk disposal through a variety of thematic charts;

[0075] S06. Emergency rescue resource management: GIS comprehensive analysis and map visualization of rescue resources, rescue plans, rescue routes and rescue strategies, including rescue resource management analysis, rescue plan analysis, and rescue decision analysis;

[0076] S07. Application of monitoring and early warning technology: Apply the Internet of Things, big data, cloud computing and artificial intelligence technologies to realize intelligent monitoring and early warning of coal mines, improve the reliability of precursor information collection, build a coal mine monitoring database based on blockchain technology, and develop a coal mine intelligent safety monitoring and early warning system with self-adaptation and deep learning.

[0077] The steps for constructing the coal mine risk warning model in S02 are as follows:

[0078] Data preprocessing: Collect multi-source data related to coal mines, and perform cleaning and normalization to improve data quality;

[0079] Data mining analysis: Use data mining techniques (such as association rules and cluster analysis) to analyze risk factors; in association rule analysis, calculate support (S) and confidence (C), determine the minimum support (S min) and minimum confidence (C min ) threshold: Where T(B represents the support of a certain rule, T represents the total number of transactions, and κ is the minimum confidence threshold;

[0080] Intelligent early warning model construction: Select a suitable algorithm, including neural network or decision tree, to build a risk early warning model and conduct training. During the model training process, use the strong association rules selected by support and confidence to determine key indicators.

[0081] Multi-source information fusion: Fusion of data from different sources to improve the accuracy and robustness of the early warning model; for quantitative indicators, according to the critical values ​​of indicators obtained by association rule analysis, the indicators are divided into three intervals and basic confidence levels are assigned: Z n ∈[Z0,+∞,Z n ∈[Z min ,Z0,Z n ∈(-∞,Z min Among them, Z0 is the critical value of the indicator, Z min is the interval partition coefficient;

[0082] Establishment of evidence theory and confidence allocation rules: Establish an evidence theory identification framework for highlighting early warning analysis, and determine the basic confidence allocation rules based on the results of association analysis; for the synthesis and decision-making of early warning results, use the following formula for evidence synthesis:

[0083]

[0084] Where K is the normalization constant;

[0085] Use a probability function-like method to make decisions:

[0086]

[0087] Where |X| is the basis of the focal element, and |U| is the basis of the identification frame;

[0088] Model verification and optimization: Verify the model through historical data and field feedback, and automatically adjust the model parameters according to the verification results to achieve dynamic tuning of the model;

[0089] Release of early warning results: The early warning results of the model are released in real time through the coal mine monitoring system so that preventive measures can be taken.

[0090] The multi-source data related to coal mines mainly include:

[0091] Production monitoring data: real-time sensor data, including temperature, humidity, and pressure; Geographic Information System (GIS) data: including spatial information on geographic location, topography, and geological structure;

[0092] Safety control data: including safety monitoring data of gas, fire and roof;

[0093] Manufacturing Execution System (MES) data: including production management data on mining progress, equipment status, and material consumption;

[0094] Dispatch management data: including personnel, materials, and equipment dispatch information;

[0095] Enterprise Resource Planning (ERP) data: including enterprise management data on finance, human resources, and supply chain;

[0096] Environmental and meteorological data: the impact of meteorological conditions on coal mine safety and production;

[0097] Video and audio data: underground monitoring video and audio records;

[0098] Drawing and document data: including design drawings, geological reports, and safety regulations;

[0099] After processing, the data is used to build a coal mine risk warning model to achieve safety monitoring and risk assessment.

[0100] Key indicators are variables or attributes that appear frequently and have strong correlation in the data set; the four important key indicators selected based on support and confidence include: gas concentration, groundwater level, roof stability, and equipment operating status.

[0101] During model verification and optimization, the model parameters adjusted include:

[0102] Sensor parameters: basic measurements related to coal mine safety monitoring systems, such as set values ​​and thresholds for methane concentration and carbon monoxide concentration;

[0103] Model hyperparameters: including learning rate, number of hidden layer neurons, number of iterations, etc. in machine learning models. These parameters directly affect the learning ability and prediction accuracy of the model.

[0104] Warning threshold: including alarm threshold, power-off threshold, and power-restore threshold. This parameter determines when to trigger warning measures, and is related to risk management and emergency response;

[0105] Time series parameters: involve data sampling frequency and time window size, which affect the timeliness of data and the model's ability to respond to dynamic changes;

[0106] Spatial parameters: For GIS data, spatial resolution and coordinate system parameters affect the ability to accurately locate risk areas and perform spatial analysis.

[0107] The formula for constructing the coal mine accident risk assessment grading table in S04 is as follows:

[0108] For underground coal mines, the calculation formula for the safety production standardization assessment score M is:

[0109]

[0110] Among them, M i represents the safety production standardization assessment score of the i-th part, a i Represents the weight value of the i-th part;

[0111] For open-pit coal mines, the calculation formula for the safety production standardization assessment score N is:

[0112]

[0113] Among them, N i represents the safety production standardization assessment score of the i-th part, b i Represents the weight value of the i-th part. Through calculation, the safety production standardization level of the coal mine can be quantitatively evaluated, and then a risk assessment grading table can be constructed to provide a scientific basis for coal mine safety management.

[0114] The mean algorithm of data mining used in S04 to implement clustering law is:

[0115] Step 1: For each data point x in the dataset j and cluster center c i , calculate the Euclidean distance between them:

[0116] Where D is the dimension of the data, x j,d and c i,d The data points x are j and cluster center c i The value in the dth dimension;

[0117] Step 2: Assign data points to the nearest cluster center: Each data point is assigned to the nearest cluster center, that is:

[0118]

[0119] Among them, S i is assigned to the cluster center c i A collection of data points;

[0120] Step 3: Update the cluster center: For each cluster S i , update the cluster center c i is the average value of all data points in the cluster:

[0121]

[0122] Among them, |S i | is cluster S i The number of data points in ;

[0123] Step 4: Iterate until convergence: Repeat steps 2 and 3 until the cluster center no longer changes significantly or the preset number of iterations is reached.

[0124] In S05, risks are divided into different levels according to the risk assessment value and represented by different colors. For example, the risk assessment value in the range of [0,1.5] is low risk, represented by blue; (1.5,2.6] is general risk, represented by yellow; (2.6,3.2] is relatively high risk, represented by orange; (3.2,4.0] is significant risk, represented by red.

[0125] In order to more specifically demonstrate the risk assessment method of the present invention, the following two tables provide specific data and results of the coal mine risk assessment. These data tables not only quantify the risk factors of the coal mine, but also propose corresponding management measures based on the assessment results.

[0126] Table 1: Quantitative table of coal mine risk factors:

[0127]

[0128] Table 2: Coal mine risk assessment results and classification table

[0129]

[0130] Based on the data provided in the two tables, the following conclusions can be drawn:

[0131] Coal Mine A has the highest risk assessment value (85 points), indicating that it faces a high risk level. It needs to pay special attention to the problems of high gas concentration and unstable roof. It is recommended to take emergency measures such as strengthening ventilation and reinforcing the roof to reduce risks.

[0132] The risk assessment value of Coal Mine B is 60 points, which is at a medium risk level. The main risk factors include aging equipment and rising groundwater levels. It is recommended to update equipment and monitor water level changes to control risks.

[0133] The risk assessment value of Coal Mine C is 40 points, which is a low risk level. The gas concentration is low and the equipment is in good condition, but regular monitoring and vigilance are still required.

[0134] Coal Mine D has the lowest risk assessment value (25 points), which is the lowest risk level. Due to its stable environment and new equipment, it is recommended to maintain the status quo and conduct regular inspections.

[0135] In summary, the risk conditions of different coal mines vary significantly, and corresponding risk management measures need to be taken according to their respective risk assessment values ​​and influencing factors. These measures will help reduce the incidence of coal mine accidents, ensure the safety of miners and the stable operation of coal mines.

[0136] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or basic features of the present invention. Therefore, no matter from which point of view, the embodiments should be regarded as exemplary and non-restrictive. The scope of the present invention is defined by the attached claims rather than the above description, and it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention. Any figure mark in the claims should not be regarded as limiting the claims involved.

[0137] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. A method for predicting the risk of coal mine hidden disasters caused by superposition of multiple source factors, characterized in that: The steps are as follows: S01. Information collection and processing: Use the GIS information collection platform to effectively collect and process coal mine hazard source information, including basic enterprise information, production monitoring information and safety management information; S02. Risk assessment model construction: Construct a coal mine risk warning model, conduct data mining analysis, algorithm operation analysis, comprehensive prediction analysis on the collected information, and draw risk warning conclusions; S03. Risk warning result display: Use GIS analysis and display platform to conduct comprehensive visual warning and display of warning results, and provide auxiliary decision-making support for mine safety production scheduling; S04. Accident risk assessment: By cleaning, integrating and filtering the collected data, and combining it with the risk assessment grading table constructed by the coal mine risk warning model, the mean algorithm of data mining is used to realize clustering law discovery and capture the comprehensive information of coal mine accidents; S05. 3D visualization of risk information: 3D visualization of coal mine risk warning conclusions on GIS maps, and display of information related to mine risk levels, categories, status details, and risk disposal through a variety of thematic charts; S06. Emergency rescue resource management: GIS comprehensive analysis and map visualization of rescue resources, rescue plans, rescue routes and rescue strategies, including rescue resource management analysis, rescue plan analysis, and rescue decision analysis; S07. Application of monitoring and early warning technology: Apply the Internet of Things, big data, cloud computing and artificial intelligence technologies to realize intelligent monitoring and early warning of coal mines, improve the reliability of precursor information collection, build a coal mine monitoring database based on blockchain technology, and develop a coal mine intelligent safety monitoring and early warning system with self-adaptation and deep learning.

2. According to claim 1, a method for predicting the risk of coal mine hidden disasters caused by superposition of multiple sources of factors, characterized in that: The steps of constructing the coal mine risk early warning model in S02 are as follows: Data preprocessing: Collect multi-source data related to coal mines, and perform cleaning and normalization to improve data quality; Data mining analysis: Use data mining techniques (such as association rules and cluster analysis) to analyze risk factors; in association rule analysis, calculate support (S) and confidence (C), determine the minimum support (S min ) and minimum confidence (C min ) threshold: Where T(B represents the support of a certain rule, T represents the total number of transactions, and κ is the minimum confidence threshold; Intelligent early warning model construction: Select a suitable algorithm, including neural network or decision tree, to build a risk early warning model and conduct training. During the model training process, use the strong association rules selected by support and confidence to determine key indicators. Multi-source information fusion: Fusion of data from different sources to improve the accuracy and robustness of the early warning model; for quantitative indicators, according to the critical values ​​of indicators obtained by association rule analysis, the indicators are divided into three intervals and basic confidence levels are assigned: Z n ∈[Z0,+∞,Z n ∈[Z min ,Z0,Z n ∈(-∞,Z min Among them, Z0 is the critical value of the indicator, Z min is the interval partition coefficient; Establishment of evidence theory and confidence allocation rules: Establish an evidence theory identification framework for highlighting early warning analysis, and determine the basic confidence allocation rules based on the results of association analysis; for the synthesis and decision-making of early warning results, use the following formula for evidence synthesis: Where K is the normalization constant; Use a probability function-like method to make decisions: Where |X| is the basis of the focal element, and |U| is the basis of the identification frame; Model verification and optimization: Verify the model through historical data and field feedback, and automatically adjust the model parameters according to the verification results to achieve dynamic tuning of the model; Release of early warning results: The early warning results of the model are released in real time through the coal mine monitoring system so that preventive measures can be taken.

3. According to claim 2, a method for predicting the risk of coal mine hidden disasters caused by superposition of multiple sources of factors, characterized in that: The multi-source data related to coal mines mainly include: Production monitoring data: real-time sensor data, including temperature, humidity, and pressure; Geographic Information System (GIS) data: including spatial information on geographic location, topography, and geological structure; Safety control data: including safety monitoring data of gas, fire and roof; Manufacturing Execution System (MES) data: including production management data on mining progress, equipment status, and material consumption; Dispatch management data: including personnel, materials, and equipment dispatch information; Enterprise Resource Planning (ERP) data: including enterprise management data on finance, human resources, and supply chain; Environmental and meteorological data: the impact of meteorological conditions on coal mine safety and production; Video and audio data: underground monitoring video and audio records; Drawing and document data: including design drawings, geological reports, and safety regulations; After processing, the data is used to build a coal mine risk warning model to achieve safety monitoring and risk assessment.

4. According to claim 2, a method for predicting the risk of coal mine hidden disasters caused by superposition of multiple sources of factors, characterized in that: The key indicators are variables or attributes that appear frequently and have strong correlation in the data set; The four important key indicators selected based on support and confidence include: gas concentration, groundwater level, roof stability, and equipment operating status.

5. According to claim 2, a method for predicting the risk of coal mine hidden disasters caused by superposition of multiple sources of factors, characterized in that: In the model verification and optimization, the model parameters adjusted include: Sensor parameters: basic measurements related to coal mine safety monitoring systems, such as set values ​​and thresholds for methane concentration and carbon monoxide concentration; Model hyperparameters: including learning rate, number of hidden layer neurons, number of iterations, etc. in machine learning models. These parameters directly affect the learning ability and prediction accuracy of the model. Warning threshold: including alarm threshold, power-off threshold, and power-restore threshold. This parameter determines when to trigger warning measures, and is related to risk management and emergency response; Time series parameters: involve data sampling frequency and time window size, which affect the timeliness of data and the model's ability to respond to dynamic changes; Spatial parameters: For GIS data, spatial resolution and coordinate system parameters affect the ability to accurately locate risk areas and perform spatial analysis.

6. According to claim 1, a method for predicting the risk of coal mine hidden disasters caused by superposition of multiple sources of factors, characterized in that: The formula for constructing the coal mine accident risk assessment grading table in S04 is as follows: For underground coal mines, the calculation formula for the safety production standardization assessment score M is: Among them, M i represents the safety production standardization assessment score of the i-th part, a i Represents the weight value of the i-th part; For open-pit coal mines, the calculation formula for the safety production standardization assessment score N is: Among them, N i represents the safety production standardization assessment score of the i-th part, b i Represents the weight value of the i-th part. Through calculation, the safety production standardization level of the coal mine can be quantitatively evaluated, and then a risk assessment grading table can be constructed to provide a scientific basis for coal mine safety management.

7. According to claim 1, a method for predicting the risk of coal mine hidden disasters caused by superposition of multiple sources of factors, characterized in that: The method of implementing clustering rules by using the mean value algorithm of data mining in S04 is: Step 1: For each data point x in the dataset j and cluster center c i , calculate the Euclidean distance between them: Where D is the dimension of the data, x j,d and c i,d The data points x are j and cluster center c i The value in the dth dimension; Step 2: Assign data points to the nearest cluster center: Each data point is assigned to the nearest cluster center, that is: Among them, S i is assigned to the cluster center c i A collection of data points; Step 3: Update the cluster center: For each cluster S i , update the cluster center c i is the average value of all data points in the cluster: Among them, |S i | is cluster S i The number of data points in ; Step 4: Iterate until convergence: Repeat steps 2 and 3 until the cluster center no longer changes significantly or the preset number of iterations is reached.

8. According to claim 1, a method for predicting the risk of coal mine hidden disasters caused by superposition of multiple sources of factors, characterized in that: In S05, the risk is divided into different levels according to the risk assessment value and represented by different colors. For example, the risk assessment value in the range of [0,1.5] is low risk, represented by blue; (1.5,2.6] is general risk, represented by yellow; (2.6,3.2] is relatively high risk, represented by orange; (3.2,4.0] is significant risk, represented by red.

Images