Stope roof risk intelligent early warning method and system based on uncertainty fusion model

Abstract

A method and system for intelligent early warning of roof risks in mining areas based on an uncertainty fusion model is disclosed. The method includes: determining the static and dynamic indicators required for assessment and establishing risk level assessment standards; defining the data range and level classification of various indicators and their correspondence with roof safety risk levels; obtaining static and dynamic indicators using a dynamic data acquisition scheme, and performing defuzzification and standardization on the data; extracting data features using multiple subjective and objective weight determination methods to obtain comprehensive indicator weights and redistributing weights under incomplete indicator conditions; assessing the roof safety risk level using the maximum membership principle, obtaining and visualizing the risk tendency judgment result; and implementing linked early warning based on the assessment results. The system includes an acquisition module, a weight management module, an assessment standard module, a computational logic module, a risk level assessment module, an early warning module, a visualization module, and a dynamic optimization module. This invention enables accurate identification and effective early warning of roof risks.

Description

Technical Field

[0001] This invention belongs to the field of intelligent early warning technology for mine roof safety, specifically a method and system for intelligent early warning of mine roof risk based on an uncertainty fusion model. Background Technology

[0002] China boasts a complete range of mineral resources with abundant reserves, serving as a crucial strategic support for national economic development and energy security. However, with ongoing mining, resources available for shallow open-pit mining are gradually dwindling, leading to a significant increase in the proportion of underground mines. In underground mines, machinery, equipment, and personnel are exposed beneath the roof, which also provides the connection to the overlying strata. Therefore, roof safety risks not only endanger personnel and equipment below but can also induce surface subsidence and water-conducting fissures in the overlying strata, potentially triggering mine water inrushes. Roof safety risks are characterized by fluctuations and suddenness, making risk warnings for the entire lifecycle of mine roof operations extremely valuable for ensuring safe mining operations. However, current underground mine roof safety risk management relies heavily on redundant designs from the design phase and reliance on enhanced support and experience-based inspections by technical personnel during production. This results in overly passive roof management systems that fail to effectively identify and warn of roof safety risks in a timely manner. Summary of the Invention

[0003] To address the problems existing in the prior art, this invention provides a method and system for intelligent early warning of mine roof risks based on an uncertainty fusion model. This method is simple to implement and highly intelligent, enabling efficient and accurate assessment of roof risk levels. It can also trigger corresponding early warning actions based on different roof risk levels, which is of paramount value for ensuring safe production operations in mines. The system has a simple structure and high level of intelligence, enabling precise identification and dynamic early warning of mine roof risk levels, effectively improving the safety management level of mine roofs.

[0004] To achieve the above objectives, the present invention provides an intelligent early warning method for mining roof risk based on an uncertainty fusion model, comprising the following steps;

[0005] Step 1: Based on the requirements of underground mine engineering, determine the static and dynamic indicators needed for roof safety risk assessment, and establish roof safety risk level assessment standards;

[0006] Step 2: Based on numerical simulation and engineering standards, define the data range and scoring range corresponding to the indicators under different safety levels;

[0007] Step 3: Based on data requirements, adopt a dynamic data acquisition scheme to obtain static and dynamic indicator data, and perform deblurring and standardization processing on the obtained data;

[0008] Step 4: Based on data-driven approach, extract data features by combining multiple subjective and objective weighting methods, and obtain the comprehensive index weights and the weights redistributed under the condition of incomplete indicators.

[0009] Step 5: Based on the current optimal weight, assess the safety risk level of the roof using the maximum membership principle, obtain the risk tendency judgment result, and visualize it through the basic cloud model of risk assessment;

[0010] Step Six: Based on the risk threshold set by the evaluation results, automatically activate the early warning equipment to provide simultaneous sound, light, and electrical warnings and automatically report the warning information;

[0011] Step 7: By collecting information on roof risk warnings and comparing it with actual on-site information, a comparative database is formed, the weights of indicators are dynamically optimized, and the basic cloud model for risk assessment is revised.

[0012] As a preferred embodiment, in step one, the static index data includes roof rock mass grade, stress field data, mining technology, and historical accident information; the dynamic index data includes roof displacement data, stress data, and crack conditions; and the roof safety risk level includes Level I roof risk, Level II roof risk, Level III roof risk, Level IV roof risk, and Level V roof risk. Specifically, the scoring range for Level I roof risk is ≥81 and ≤100; for Level II roof risk, it is ≥61 and ≤80; for Level III roof risk, it is ≥41 and ≤60; for Level IV roof risk, it is ≥21 and ≤40; and for Level V roof risk, it is ≥0 and ≤20.

[0013] As a preferred option, the correspondence in step two is as follows:

[0014] For the roof rock mass grade, it is divided into five grades according to the rock mass stability: extremely stable, stable, moderate, unstable and extremely unstable, which correspond to Grade I roof risk, Grade II roof risk, Grade III roof risk, Grade IV roof risk and Grade V roof risk, respectively. Among them, extremely stable corresponds to 81≤rock mass stability≤100, stable corresponds to 61≤rock mass stability≤80, moderate corresponds to 41≤rock mass stability≤60, unstable corresponds to 21≤rock mass stability≤40, and extremely unstable corresponds to 0≤rock mass stability≤20.

[0015] For stress field data, it is divided into five levels according to the magnitude of ground stress: no risk, low risk, potential risk, high risk and extremely high risk, which correspond to the roof risk of Class I, Class II, Class III, Class IV and Class V respectively. Among them, no risk corresponds to ground stress < 5 MPa, low risk corresponds to 5 MPa ≤ ground stress < 10 MPa, potential risk corresponds to 10 MPa ≤ ground stress < 15 MPa, high risk corresponds to 15 MPa ≤ ground stress < 20 MPa, and extremely high risk corresponds to ground stress > 20 MPa.

[0016] Mining processes are classified into five levels based on the disturbance force of each method: extremely low disturbance, low disturbance, moderate disturbance, high disturbance, and relatively high disturbance. These levels correspond to Class I, Class II, Class III, Class IV, and Class V roof risks, respectively. Specifically, extremely low disturbance corresponds to bench mining, residual ore recovery, and shallow layered mining; low disturbance corresponds to layered mining, shallow room-and-pillar mining, and strip mining; moderate disturbance corresponds to room-and-pillar mining, longwall mining, sublevel caving, and room-and-pillar mining with ore-holding zones; high disturbance corresponds to sublevel backfilling mining, staged open area mining, staged backfilling, and staged caving; and relatively high disturbance corresponds to deep full-face mining, large-stope unsupported caving, block caving, continuous mining, and full-face mining.

[0017] Based on the average number of roof falls per year, historical accident data is categorized into five levels: safe, low risk, medium risk, high risk, and extremely high risk. These levels correspond to Level I, Level II, Level III, Level IV, and Level V roof risks, respectively. Specifically, safe corresponds to no roof falls per year; low risk corresponds to 0 to 1 roof falls per year; medium risk corresponds to 2 to 1 roof falls per year; high risk corresponds to 3 to 2 roof falls per year; and extremely high risk corresponds to more than 3 roof falls per year.

[0018] For roof displacement data, it is divided into five levels according to the roof displacement situation: low risk, relatively low risk, medium risk, high risk and extremely high risk, which correspond to Level I roof risk, Level II roof risk, Level III roof risk, Level IV roof risk and Level V roof risk, respectively. Among them, low risk corresponds to 0 ≤ daily average displacement rate ≤ 1 mm / d and 0 ≤ cumulative displacement ≤ 10 mm; relatively low risk corresponds to 1 mm / d < daily average displacement rate ≤ 3 mm / d and 10 mm < cumulative displacement ≤ 20 mm; medium risk corresponds to 3 mm / d < daily average displacement rate ≤ 5 mm / d and 20 mm < cumulative displacement ≤ 40 mm; high risk corresponds to 5 mm / d < daily average displacement rate ≤ 8 mm / d and 40 mm < cumulative displacement ≤ 60 mm; and extremely high risk corresponds to daily average displacement rate > 8 mm / d and cumulative displacement > 60 mm.

[0019] For stress data, it is divided into five levels according to the stress condition of the support / column: stable pressure, moderate pressure, leakage pressure, heavy pressure, and overpressure. These levels correspond to the roof risk levels of Class I, Class II, Class III, Class IV, and Class V, respectively. Specifically, stable pressure corresponds to 30% of the design bearing capacity ≤ support / column stress < 50% of the design bearing capacity; moderate pressure corresponds to 50% of the design bearing capacity ≤ support / column stress < 70% of the design bearing capacity; leakage pressure corresponds to support / column stress < 30% of the design bearing capacity; heavy pressure corresponds to 70% of the design bearing capacity ≤ support / column stress < 90% of the design bearing capacity; and overpressure corresponds to 70% of the design bearing capacity ≥ 90% of the design bearing capacity.

[0020] For crack conditions, they are classified into five levels according to crack size: no crack, fine crack, stable crack, expanding crack, and penetrating crack. These correspond to the roof risk levels of Class I, Class II, Class III, Class IV, and Class V, respectively. Specifically, no crack corresponds to no obvious cracks or no cracks; fine crack corresponds to the presence of small cracks with a crack width < 1 mm; stable crack corresponds to the presence of non-penetrating cracks with a crack width of 1 mm ≤ crack width < 3 mm; expanding crack corresponds to the presence of non-penetrating cracks with a crack width of 3 mm ≤ crack width < 5 mm; and penetrating crack corresponds to the presence of penetrating cracks with a crack width ≥ 5 mm.

[0021] As a preferred approach, in step three, static index data is obtained by technical and design personnel through simulation calculations based on the mine geological report and is then manually input into the data processing terminal. Dynamic index data is collected in real time and automatically using monitoring equipment deployed at the mine site, thereby obtaining both static and dynamic index data. The monitoring equipment is deployed differently in coal mines and non-coal mines.

[0022] To ensure the acquisition of high-quality monitoring data, the specific process of deblurring and standardizing the acquired data in step three is as follows:

[0023] S31; Introducing the rule mapping defuzzification mechanism from fuzzy set theory to convert qualitative indicators into quantitative data, and obtaining the level score corresponding to the i-th indicator according to formula (1). ;

[0024] (1);

[0025] In the formula, Indicates the corresponding indicator level;

[0026] S32: Perform linear standardization on indicators with actual data; for data ranges of... The score range is Data value X in When, the corresponding score range is The score of the j-th index after nonlinear standardized mapping is obtained according to formula (2). ,

[0027] (5).

[0028] As a preferred option, the specific process of step four is as follows:

[0029] S41: At the subjective level, an expert evaluation matrix is ​​constructed based on the analytic hierarchy process to extract expert knowledge and experience; at the objective level, the importance of indicators is quantified from the perspectives of information distribution and volatility by combining the entropy method and the coefficient of variation method; finally, a unified weighting system is formed by combining the three methods through a linear weighted fusion strategy, providing accurate input feature guarantee for the subsequent risk level identification model; specifically, the comprehensive weight is obtained according to formula (3). ;

[0030] (3);

[0031] In the formula, Weights for the analytic hierarchy process (AHP); Weights are determined by entropy. Weights are calculated using the coefficient of variation method. =0.4, representing the reliability coefficient of the analytic hierarchy process; =0.3, which is the reliability coefficient of the entropy method; =0.3 is the reliability coefficient using the coefficient of variation method; ;

[0032] S42: For a weight vector containing n indicators , , When m indicators are missing, the weight of the i-th indicator after redistribution is obtained according to formula (4). According to formula (5), the following formula is obtained: Weight vector of each indicator ;

[0033] (4);

[0034] (5);

[0035] In the formula, The weight of the i-th indicator before redistribution; The weighted sum of the residual indicators before redistribution; The number of participating indicators, ; .

[0036] As a preferred option, the specific process of step five is as follows:

[0037] S51: Collect static and dynamic indicator data of the assessed area in real time, calculate the final evaluation score based on the established roof safety risk level assessment standard, the data range and level classification of various indicators and the corresponding relationship of roof safety risk level, the comprehensive indicator weight and the redistribution of weight under the condition of incomplete indicators, and calculate the membership degree of the final evaluation score under each risk level, as shown in formula (6); calculate the membership degree of the data group at the corresponding time for each level, and determine the maximum membership degree as the roof safety risk level of the area at the current time;

[0038] (6);

[0039] In the formula, c represents the membership degree of the final evaluation score under risk level i. i This represents the central value of the membership degree under risk level i. x represents the final evaluation score;

[0040] S52: Based on the established roof safety risk level assessment criteria, construct a basic cloud model for risk assessment corresponding to each risk level. Each basic cloud model for risk assessment is defined by three parameters: expected value, entropy, and hyperentropy, forming a standard risk distribution cloud corresponding to different risk levels.

[0041] S53: Based on the real-time collected static and dynamic indicator data, calculate the evaluation score and convert it into basic cloud model parameters for risk assessment to form the current risk assessment cloud;

[0042] S54: Compare the basic risk assessment cloud model with the current risk judgment cloud to obtain the roof safety risk tendency judgment result, so as to provide a basis for auxiliary decision-making.

[0043] In order to execute different levels of warning and alert actions based on different risk levels, the specific process of step six is ​​as follows:

[0044] When the roof safety risk reaches a set threshold score, the early warning devices deployed in the working face and roadways will execute audible, visual, and electrical early warning actions and send them to the control and monitoring terminal. Specifically, when a Level I roof risk occurs, the green light on the control early warning device will remain constantly lit, and normal power supply will be maintained. At the same time, an early warning message will be generated to ensure daily inspections and maintain the current monitoring and support measures. When a Level II roof risk occurs, the blue light on the control early warning device will remain constantly lit, the early warning system will be activated, and an early warning message will be generated to increase the frequency of inspections and continuously observe changes in monitoring data. When a Level III roof risk occurs, the yellow light on the control early warning device will remain constantly lit, and a prompt voice will be broadcast. Power supply to non-critical equipment will be disconnected, and only important systems will be kept powered on. The system generates early warning information for adding monitoring points, conducting special investigations, and optimizing support measures. When a Level IV roof slab risk occurs, the system controls the orange lights of the early warning equipment to flash frequently, and broadcasts continuous beeps and prompts. Power is cut off to the designated disconnected areas, and the system is linked to safety lighting and broadcasting systems. Simultaneously, it generates early warning information for initiating the early warning response process, arranging for personnel to evacuate the relevant areas, and reinforcing the support structure. When a Level V roof slab risk occurs, the system controls the red lights of the early warning equipment to flash frequently, continuously broadcast long beeps and voice prompts, and completely cuts off power to all work areas. Simultaneously, it generates early warning information for immediately evacuating personnel, suspending work, and implementing emergency reinforcement measures.

[0045] To achieve adaptive updates to the risk assessment model, step seven involves the following steps:

[0046] S71: Continuously collect early warning data generated by roof risk warnings, and simultaneously collect actual on-site monitoring data and handling feedback information.

[0047] S72: By constructing a database comparing early warning information with actual on-site data, analyze the matching degree and deviation between early warning results and actual risk events;

[0048] S73: Based on the comparative analysis results, the dynamic weight adjustment algorithm is used to repeat steps four and five to correct the weight distribution of risk assessment indicators in real time, eliminate the influence of misleading indicators, enhance the weight sensitivity of key indicators, and thus optimize the performance of the risk assessment model.

[0049] This invention first determines the multi-source data of static and dynamic indicators required for roof safety risk assessment based on the needs of underground mine engineering, and establishes roof safety risk level assessment standards. Then, it establishes the correspondence between data ranges and risk scores for various indicators under different risk levels. This facilitates the collaborative analysis of static indicators (such as geological structures) and dynamic indicators (such as real-time stress changes), forming a more comprehensive multi-dimensional safety risk assessment framework and effectively ensuring the accurate assessment of roof risk levels. A dynamic data acquisition scheme is used to obtain static and dynamic indicator data, and qualitative indicators are quantified and analyzed through defuzzification. Standardization methods are used to improve the comparability of quantitative indicator data, effectively ensuring that the obtained multi-source indicator data accurately reflects roof status information. Data features are extracted separately through a subjective and objective weight fusion mechanism, providing accurate input feature guarantees for subsequent risk level identification models. Simultaneously obtaining comprehensive indicator weights and weight redistribution under incomplete indicator conditions allows the method to support both comprehensive indicators and weight redistribution under incomplete indicators, ensuring accurate and effective assessment of roof safety risks even when some indicator data is missing, effectively enhancing the adaptability of the indicator system. The risk level is dynamically determined by the principle of maximum membership. Risk tendency assessment accurately displays the degree of risk deviation and its development trend. Visualization maps the risk level, providing a clear and intuitive representation of the risk situation in real time. Based on the assessment results, early warning equipment is automatically activated for simultaneous audible, visual, and electrical warnings, enabling multi-channel warning linkage and ensuring timely and effective emergency response in dangerous situations. By dynamically optimizing indicator weights to correct the assessment module, a closed-loop monitoring-early warning-feedback-optimization process allows for adaptive updates to the risk assessment model, ensuring high predictive accuracy and stability under different operating conditions and time points.

[0050] This method is simple to implement and highly intelligent. It can efficiently and accurately assess the risk level of the roof and take corresponding early warning actions according to different roof risk levels. It is of great value for ensuring safe production operations in mines.

[0051] This invention also provides an intelligent early warning system for mining roof risk based on an uncertainty fusion model, which is used to implement an intelligent early warning method for mining roof risk based on an uncertainty fusion model, including a data acquisition module, a weight management module, a judgment standard module, a calculation logic module, a risk level judgment module, an early warning module, a visualization module, and a dynamic optimization module;

[0052] The acquisition module is used to collect multi-source monitoring data in the area to be evaluated, record the spatial layout information of the monitoring equipment and the data acquisition frequency, form an original dataset, and send the original dataset to the weight management module and the risk level assessment module.

[0053] The weight management module is used to extract dataset features based on the volatility of the obtained data, the correlation between indicators and historical evaluation performance, initialize the weight vector of each evaluation indicator, support the allocation of incomplete indicator weights and subsequent model feedback for dynamic weight adjustment, and output the corrected indicator weight set to the risk level assessment module.

[0054] The evaluation criteria module is used to construct a multi-level risk level system. Based on expert experience, policies and regulations, historical cases and numerical analysis results, it defines the boundary thresholds and judgment logic for each level, forms a set of evaluation criteria for the risk level of underground mine roof, and sends it to the risk level evaluation module.

[0055] The computational logic module is used to perform normalization and defuzzification processing on the collected data information, and to calculate relevant parameters and results on the processed data through fuzzy comprehensive evaluation, cloud model or multi-factor fusion algorithm, and send them to the risk level assessment module.

[0056] The risk level assessment module is used to receive the indicator data collected by the acquisition module, the indicator weights output by the weight management module, the quantitative analysis results output by the calculation logic module, and the risk level assessment standard set of the underground mine roof output by the assessment standard module. It then classifies the risk levels according to the maximum membership principle and the assessment standard module, and sends the risk level classification results to the early warning module and the visualization module.

[0057] The early warning module is used to execute corresponding sound, light, and electrical early warning measures based on the risk level output by the risk level assessment module.

[0058] The visualization module is used to display real-time assessment results, historical trend evolution, risk area distribution and early warning status in a graphical way, and to provide a human-computer interaction interface, supporting two-dimensional and three-dimensional risk maps, time series diagrams, risk cloud maps and console operation interfaces.

[0059] The dynamic optimization module is used to collect deviation information between risk level assessment results and actual on-site feedback, construct an assessment-measurement comparison database, and dynamically correct the indicator weights based on the comparison database. At the same time, the corrected indicator weights are sent to the weight management module to achieve adaptive optimization of the assessment model.

[0060] The system has a simple structure and a high degree of intelligence. It can intelligently and accurately identify and dynamically warn of the risk level of the mine roof, effectively improving the safety management level of the mine roof. Attached Figure Description

[0061] Figure 1 This is a flowchart illustrating the workflow of the intelligent early warning method in this invention.

[0062] Figure 2 This is a schematic diagram of the layout of monitoring equipment in the coal mine working face according to the present invention;

[0063] Figure 3 This is a schematic diagram showing the arrangement of monitoring equipment in a non-coal working face according to the present invention;

[0064] Figure 4 This is a schematic diagram showing the layout of monitoring equipment in the main roadways of this invention;

[0065] Figure 5 This is a visual standard risk distribution cloud map in this invention;

[0066] Figure 6 This is a cloud map for visually assessing the risk distribution in this invention;

[0067] Figure 7 This is a risk propensity comparison cloud map in this invention;

[0068] Figure 8 This is a schematic diagram of the intelligent early warning system in this invention.

[0069] In the diagram: 201, working face roof 1; 202, support stress sensor; 203, hydraulic support; 204, working face; 205, scraper conveyor; 206, working face roadway communication and amplification device 1; 207, hydraulic support telescopic column; 208, goaf.

[0070] 301. Ore blasting pile; 302. Working face roof II; 303. Working face support stress sensor; 304. Working face hydraulic support; 305. Working face roadway communication and amplification device II;

[0071] 401. Tunnel roof; 402. Support anchor bolt; 403. Support stress sensor; 404. Anchor bolt stress sensor; 405. Anchor bolt tray; 406. Displacement sensor; 407. Roof monitoring equipment; 408. Tunnel communication and amplification device; 409. Tunnel hydraulic support; 410. Tunnel sidewall; 411. Tunnel floor;

[0072] 501, Level V Assessment Cloud; 502, Level IV Assessment Cloud; 503, Level III Risk Assessment Cloud; 502, Level II Assessment Cloud; 503, Level I Risk Assessment Cloud;

[0073] 701, Level V Assessment Cloud; 702, Level IV Assessment Cloud; 703, Level III Risk Assessment Cloud; 704, Assessment Risk Cloud; 706, Level II Assessment Cloud; 707, Level I Risk Assessment Cloud. Detailed Implementation

[0074] The invention will now be further described with reference to the accompanying drawings.

[0075] This invention provides a method and system for intelligent early warning of risks to the roof of a mining area based on an uncertainty fusion model. The method and system can be implemented by an electronic device, which can be a control terminal or a server. Figure 1 A flowchart of an intelligent early warning method for mining roof risk based on an uncertainty fusion model is provided. The processing flow of this method may include the following steps:

[0076] Step 1: Based on the requirements of underground mine engineering, determine the static and dynamic indicators needed for roof safety risk assessment, and establish roof safety risk level assessment standards;

[0077] Through extensive research on current mine roof accidents, recommended data indicators were determined. Based on data changes, the data were categorized into static and dynamic indicators. Static indicators include roof rock mass grade, stress field data, mining technology, and historical accident data. Dynamic indicators include roof displacement data, stress data, and crack conditions. Based on the degree of danger to personnel and equipment according to these indicators, roof safety risk levels were divided into five levels. The meaning and scoring of each level are shown in Table 1. Specifically, the roof safety risk levels include Level I roof risk, Level II roof risk, Level III roof risk, and Level IV roof risk. The roof risk is classified into Level IV and Level V. The scoring range for Level I roof risk is ≥81 and ≤100; for Level II roof risk, it is ≥61 and ≤80; for Level III roof risk, it is ≥41 and ≤60; for Level IV roof risk, it is ≥21 and ≤40; and for Level V roof risk, it is ≥0 and ≤20.

[0078] Table 1: Classification of Safety Risk Levels for Roof Slab

[0079] Roof safety risk intensity Level I Level II Level III Level IV Level V Meaning of indicator levels Security level Controllable level Attention Level Danger Level Hazard level Indicator Score Range 81-100 61-80 41-60 21-40 0-20

[0080] The explanations and recommended measures for the five levels are as follows:

[0081] Level I roof risk indicates extremely low risk (safety level), and routine inspections are recommended to maintain current measures.

[0082] Level II roof risk indicates a low risk (controllable level). It is recommended to strengthen inspections and pay attention to dynamic changes in monitoring data.

[0083] Level III roof risk indicates medium risk (attention level). It is recommended to add monitoring points, launch a special investigation, and optimize support measures if necessary.

[0084] Level IV roof risk indicates a high risk (critical level). It is recommended to activate the risk warning procedure, arrange for personnel to evacuate the risk area, and strengthen support.

[0085] A Class V roof slab risk indicates an extremely high risk (hazard level), and emergency measures are recommended, including immediate evacuation of personnel, suspension of operations, and implementation of emergency reinforcement measures.

[0086] Step 2: Based on numerical simulation and engineering standards, classify the data range and scoring range of the indicators under different safety levels; and match the data range, level classification and scoring standards of the evaluation indicators such as roof rock mass grade, stress field data, mining technology, historical accident situation, roof displacement, stress, and crack condition.

[0087] The roof rock mass grade index and rock mass quality rating index score can be directly correlated with the roof rock mass safety risk score and used directly. Table 2 shows the correspondence between the safety risk assessment score of the underground mine roof rock mass based on the rock mass quality rating score.

[0088] Table 2: Scoring Criteria for Roof Rock Mass

[0089] RMR score 81-100 61-80 41-60 21-40 0-20 Top plate score 81-100 61-80 41-60 21-40 0-20 Rock mass stability Extremely stable Stablize medium Unstable Extremely unstable Roof safety risk intensity Level I Level II Level III Level IV Level V

[0090] For the roof rock mass grades, based on rock mass stability, they are divided into five levels: extremely stable (Level I), stable (Level II), moderate (Level III), unstable (Level IV), and extremely unstable (Level V), which correspond to Level I roof risk, Level II roof risk, Level III roof risk, Level IV roof risk, and Level V roof risk, respectively. Among them, extremely stable corresponds to 81≤rock mass stability≤100, stable corresponds to 61≤rock mass stability≤80, moderate corresponds to 41≤rock mass stability≤60, unstable corresponds to 21≤rock mass stability≤40, and extremely unstable corresponds to 0≤rock mass stability≤20.

[0091] Stress field data indicators, based on the magnitude of mine ground stress and numerical simulation methods, classify and score the stress field range. The classification results are as follows:

[0092] For stress field data, it is divided into five levels according to the magnitude of the in-situ stress: no risk (Level I), low risk (Level II), potential risk (Level III), high risk (Level IV), and extremely high risk (Level V), corresponding to Level I, Level II, Level III, Level IV, and Level V roof risks, respectively. Specifically, no risk corresponds to in-situ stress < 5 MPa, indicating no significant safety risk to the roof; low risk corresponds to 5 MPa ≤ in-situ stress < 10 MPa, indicating a low safety risk to the roof; potential risk corresponds to 10 MPa ≤ in-situ stress < 15 MPa, indicating a potential safety risk to the roof; high risk corresponds to 15 MPa ≤ in-situ stress < 20 MPa, indicating a high safety risk to the roof; and extremely high risk corresponds to in-situ stress > 20 MPa, indicating an extremely high safety risk to the roof. Quantitative values ​​are assigned to the five stress ranges, as shown in Table 3.

[0093] Table 3: Risk Assessment Table for Roof Slab Safety Caused by Geostress

[0094] Roof safety risk intensity Level I Level II Level III Level IV Level V Indicator Meaning Risk-free Lower risk Potential risks Higher risk Extremely high risk Top plate score 81-100 61-80 41-60 21-40 0-20

[0095] Based on relevant standards such as the "Safety Regulations for Underground Metal and Non-metal Mines" and the "Safety Regulations for Coal Mines," mining technology indicators classify the mining methods currently used in underground mines into five levels. The classification results are as follows:

[0096] For mining techniques, based on the disturbance force of the mining method, they are classified into five levels: extremely low disturbance (Level I), low disturbance (Level II), moderate disturbance (Level III), high disturbance (Level IV), and relatively high disturbance (Level V), corresponding to Level I, Level II, Level III, Level IV, and Level V roof risks, respectively. Extremely low disturbance involves weak ground pressure activity and minimal roof disturbance; the corresponding mining methods are bench mining, residual ore recovery, and shallow layering. Low disturbance involves slight stress concentration and relatively stable roof; the corresponding mining method is layered mining. Shallow room-and-pillar and strip mining methods result in moderate disturbance of ground stress and moderate disturbance of the roof. The corresponding mining methods are room-and-pillar, longwall, sublevel caving, and room-and-pillar with ore-holding, respectively. Highly disturbed rock masses are fractured, with large goafs causing strong disturbance of the roof. The corresponding mining methods are sublevel backfilling, staged open-stope mining, staged backfilling, and staged caving, respectively. Highly disturbed rock masses cause extremely strong disturbance of the roof. The corresponding mining methods are deep full-run mining, large-stope unsupported caving, block caving, continuous mining, and full-face mining. Quantitative scores are assigned to the five levels, as shown in Table 4.

[0097] Table 4: Risk Scoring Table for Roof Safety Caused by Mining Methods

[0098] Roof safety risk intensity Level I Level II Level III Level IV Level V Indicator Meaning Extremely low disturbance low disturbance Medium disturbance High disturbance higher disturbance Top plate score 81-100 61-80 41-60 21-40 0-20

[0099] Historical accident indicators, based on the "Classification Standard for Mine Safety Accidents," classify the occurrence of historical roof fall safety accidents in mines into 5 levels:

[0100] Historical accident data is categorized into five levels based on the average annual number of roof fall accidents: Safe (Level I), Low Risk (Level II), Medium Risk (Level III), High Risk (Level IV), and Extremely High Risk (Level V). These levels correspond to Level I, Level II, Level III, Level IV, and Level V roof risk, respectively. Safe corresponds to no roof fall accidents and no recorded roof fall accidents in the mine. Low risk corresponds to 0 to 1 roof fall accidents per year, meaning the mine has a history of roof fall accidents. Roof accidents are classified as follows: Medium risk corresponds to 1 to 2 roof accidents per year (on average annually in the mine), indicating minor injuries or near-misses due to roof accidents; High risk corresponds to 2 to 3 roof accidents per year (on average annually in the mine), indicating injuries or equipment damage due to roof accidents; Very high risk corresponds to more than 3 roof accidents per year (on average annually in the mine), indicating multiple injuries or major equipment damage due to roof accidents. The five risk levels are explained and quantitatively scored as shown in Table 5.

[0101] Table 5: Safety Risk Score Assignment Table for Historical Roof Fall Accidents in Mines

[0102] Roof Accident Risk Intensity Level I Level II Level III Level IV Level V Indicator Meaning Safety Low risk Medium risk High risk Extremely high risk Top plate score 81-100 61-80 41-60 21-40 0-20

[0103] The roof displacement index, based on the deformation response of the roof strata under mining disturbance, classifies the impact of roof displacement on roof safety risk into 5 levels, as shown in Table 6:

[0104] Table 6: Correspondence between Roof Displacement and Roof Safety Risk

[0105] Roof safety risk intensity Level I Level II Level III Level IV Level V Indicator Meaning Low-risk displacement Lower risk displacement Medium risk displacement High-risk displacement Extremely high risk displacement Daily average displacement rate (mm / d) ≤1 1-3 3-5 5-8 >8 Cumulative displacement (mm) ≤10 10-20 20-40 40-60 >60 Top plate score 81-100 61-80 41-60 21-40 0-20

[0106] For roof displacement data, it is divided into five levels according to the roof displacement situation: low risk (Level I), relatively low risk (Level II), medium risk (Level III), high risk (Level IV), and extremely high risk (Level V), which correspond to Level I roof risk, Level II roof risk, Level III roof risk, Level IV roof risk, and Level V roof risk, respectively. Among them, low risk corresponds to 0 ≤ daily average displacement rate ≤ 1 mm / d and 0 ≤ cumulative displacement ≤ 10 mm; relatively low risk corresponds to 1 mm / d < daily average displacement rate ≤ 3 mm / d and 10 mm < cumulative displacement ≤ 20 mm; medium risk corresponds to 3 mm / d < daily average displacement rate ≤ 5 mm / d and 20 mm < cumulative displacement ≤ 40 mm; high risk corresponds to 5 mm / d < daily average displacement rate ≤ 8 mm / d and 40 mm < cumulative displacement ≤ 60 mm; and extremely high risk corresponds to daily average displacement rate > 8 mm / d and cumulative displacement > 60 mm.

[0107] Stress indices, based on monitored stress and the design bearing capacity of the support structure, classify the safety risk level of the roof slab as follows:

[0108] For stress data, it is divided into five levels according to the stress condition of the support / column: stable pressure (Level I), moderate pressure (Level II), leakage pressure (Level III), heavy pressure (Level IV), and overpressure (Level V), which correspond to the roof risk levels of Level I, Level II, Level III, Level IV, and Level V, respectively. Stable pressure refers to the state where the stress on the support / column is within the initial support force range, corresponding to 30% of the design bearing capacity ≤ the stress on the support / column (as monitored by the equipment) < 50% of the design bearing capacity. Moderate pressure refers to the state where the support / column is subjected to... The impact of stress concentration is defined as follows: 50% of the design bearing capacity ≤ support / column stress < 70% of the design bearing capacity; leakage pressure is an abnormal situation where the support / column stress is far below the design bearing capacity, corresponding to support / column stress < 30% of the design bearing capacity; heavy pressure is a state where the support / column stress is close to its limit, corresponding to 70% of the design bearing capacity ≤ support / column stress < 90% of the design bearing capacity; overpressure is a state where the support stress reaches the critical point, corresponding to 70% of the design bearing capacity ≥ 90% of the design bearing capacity. The explanations and scores for these levels are shown in Table 7.

[0109] Table 7: Correspondence between monitored stress and safety risk of roof slab

[0110] Roof safety risk intensity Level I Level II Level III Level IV Level V Indicator Meaning Voltage Regulator Appropriate pressure Leakage heavy pressure Overpressure Rock mass scoring 81-100 61-80 41-60 21-40 0-20

[0111] Crack condition indicators, based on image recognition (using high-definition image acquisition equipment for image acquisition and recognition) and crack measurement, are classified into 5 levels according to their impact on roof safety risks:

[0112] For crack conditions, cracks are classified into five levels based on their size: no cracks (Level I), fine cracks (Level II), stable cracks (Level III), expanding cracks (Level IV), and penetrating cracks (Level V). These levels correspond to Level I, Level II, Level III, Level IV, and Level V roof risks, respectively. Specifically, no cracks correspond to no obvious cracks or no cracks at all; fine cracks correspond to the presence of small, stable cracks with slow propagation and a crack width < 1 mm; stable cracks correspond to the presence of non-penetrating cracks, meaning some cracks exist, which expand but do not penetrate, and the crack width is 1 mm ≤ crack width < 3 mm; expanding cracks correspond to the presence of non-penetrating cracks, meaning obvious cracks exist, which propagate rapidly, and the crack width is 3 mm ≤ crack width < 5 mm; and penetrating cracks correspond to penetrating cracks, where the cracks are clearly penetrating and the crack width is ≥ 5 mm. The explanation and scoring of the roof risk levels corresponding to cracks are shown in Table 8.

[0113] Table 8: Correspondence between Roof Crack Monitoring and Roof Safety Risks

[0114] Roof safety risk intensity Level I Level II Level III Level IV Level V Indicator Meaning No cracks Fine cracks Stable crack Crack Through crack Rock mass scoring 81-100 61-80 41-60 21-40 0-20

[0115] Step 3: Based on data requirements, adopt a dynamic data acquisition scheme to obtain static and dynamic indicator data, and perform deblurring and standardization processing on the obtained data;

[0116] As a preferred approach, static index data (roof rock mass grade, stress field data, mining technology, and historical accident information) are obtained by technical and design personnel through simulation calculations based on the mine geological report and are then manually input into the data processing terminal. Dynamic index data (roof displacement, stress, and crack conditions) are collected in real time and automatically using monitoring equipment deployed at the mine site to obtain both static and dynamic index data. The monitoring equipment is deployed differently in coal mines and non-coal mines.

[0117] Based on the supplementary renovation of existing underground mine equipment, the equipment layout is divided into two main categories: underground mine working face layout and main roadway layout.

[0118] In the coal mine working face, the support pressure sensor 202 is installed above the hydraulic support 203 to monitor the stress of the working face roof in real time, and the working face roadway communication and amplification device 206 is installed on the support telescopic column 207 for roof early warning. Figure 2 The diagram shows the layout of equipment in an underground coal mine working face, including but not limited to: a pressure sensor 202 on the support above the hydraulic support 203 between working face 204 and goaf 208 to monitor the stress on the roof 201 of the working face above; and a roadway communication and amplification device 206 installed on the hydraulic support telescopic column 207 to provide early warning of roof safety risks and to protect key working face equipment such as the scraper conveyor 205.

[0119] In non-coal mines, the working face support stress sensor 303 is installed between the hydraulic support 304 and the roof to monitor the impact of ore explosion on the roof. The roadway communication and amplification device 305 is also used for early warning. Figure 3 The diagram shows the equipment layout of some underground non-coal mine working faces, including but not limited to: a working face support stress sensor 303, located between the working face roof 302 and the working face hydraulic support 304, used to monitor the impact of the ore bursting 301 formation process on the roof. A working face roadway communication and amplification device 305, fixed to the working face hydraulic support, is used to provide early warning of roof safety risks.

[0120] In the main roadway, the support stress sensor 403 and the anchor stress sensor 404 jointly monitor the stress on the roof, the displacement sensor 406 is used to monitor the roof displacement, the crack monitoring device (which can be a high-definition camera) is used to track the development of roof cracks, and the loudspeaker emits audible and visual warning signals. Figure 4 The diagram shows the layout of monitoring equipment in the main roadways of the underground mine, including but not limited to: hydraulic props 409 supported between the roadway floor 411 and the roadway roof 401; a prop stress sensor 403 located above the hydraulic props 409 and an anchor stress sensor 404 placed between the anchor tray 405 of the support anchor 402 and the roadway roof 401, jointly monitoring roof stress; a displacement sensor 406 located in the roadway sidewall 410 to monitor roof displacement; a roof monitoring device 407 to monitor changes in roof cracks; and a roadway communication and amplification device 408 using audible and visual alarms to alert users of roof safety risks.

[0121] To ensure the acquisition of high-quality monitoring data, the specific process of deblurring and standardizing the acquired data is as follows:

[0122] S31; To achieve the quantitative expression of qualitative indicators, a rule mapping defuzzification mechanism from fuzzy set theory is introduced to convert qualitative indicators (mining technology, historical accident situation) into quantitative data. The level score corresponding to the i-th indicator is obtained according to formula (1). ;

[0123] (1);

[0124] In the formula, These represent the corresponding indicator levels, namely Level I, Level II, Level III, Level IV, and Level V.

[0125] The quantitative relationship between the five risk levels of the roof is shown in Table 9:

[0126] Table 9: Corresponding Quantitative Relationship Table

[0127] Roof safety risk intensity Level I Level II Level III Level IV Level V Quantitative score Q 90 70 50 30 10

[0128] S32: Standardization is performed based on the historical data range, corresponding to four levels and a score of 0-100. For indicators with available data (roof rock mass grade, stress field data, roof displacement data, stress data, crack conditions), linear standardization is applied; for data ranges... The score range is Data value X in When, the corresponding score range is The score of the j-th index after nonlinear standardized mapping is obtained according to formula (2). ;

[0129] (5).

[0130] Step 4: Based on data-driven approach, extract data features by combining multiple subjective and objective weighting methods, and obtain the comprehensive index weights and the weights redistributed under the condition of incomplete indicators.

[0131] As a preferred option, the specific process is as follows:

[0132] S41: Based on the established evaluation criteria and data defuzzification and standardization principles. At the subjective level, an expert evaluation matrix is ​​constructed based on the Analytic Hierarchy Process (AHP) to extract expert knowledge and experience; at the objective level, the importance of indicators is quantified from the perspectives of information distribution and volatility by combining the Entropy Method (EWM) and the Coefficient of Variation Method (VCM); finally, a unified weighting system is formed by combining the three methods through a linear weighted fusion strategy, providing accurate input feature guarantees for the subsequent risk level identification model; specifically, the comprehensive weight is obtained according to formula (3). ;

[0133] (3);

[0134] In the formula, Weights for the analytic hierarchy process (AHP); Weights are determined by entropy. Weights are calculated using the coefficient of variation method. =0.4, representing the reliability coefficient of the analytic hierarchy process; =0.3, which is the reliability coefficient of the entropy method; =0.3 is the reliability coefficient using the coefficient of variation method; ;

[0135] As a preferred option, the reliability coefficient values ​​are shown in Table 10:

[0136] Table 10: Recommended Reliability Coefficient Table

[0137] Weight extraction method Analytic Hierarchy Process Entropy method Coefficient of variation method Reliability coefficient 0.4 0.3 0.3

[0138] S42: Considering that some evaluation indicators may be unavailable or have monitoring failures, missing data, or other issues in practical applications, this invention proposes an adaptive weight redistribution mechanism to ensure the model's universality. For a weight vector containing n indicators... , , When m indicators are missing, the weight of the i-th indicator after redistribution is obtained according to formula (4). According to formula (5), the following formula is obtained: Weight vector of each indicator ;

[0139] (4);

[0140] (5);

[0141] In the formula, The weight of the i-th indicator before redistribution; The weighted sum of the residual indicators before redistribution; The number of participating indicators, ; .

[0142] Step 5: Based on the current optimal weight, assess the safety risk level of the roof using the maximum membership principle, obtain the risk tendency judgment result, and visualize it through the basic cloud model of risk assessment;

[0143] As a preferred option, the specific process is as follows:

[0144] S51: Assessing the risk level of the roof using the maximum membership principle:

[0145] Real-time collection of static and dynamic indicator data of the assessed area; calculation of the final evaluation score based on the established roof safety risk level assessment criteria, the data range and level classification of various indicators and the corresponding relationship of roof safety risk levels, the comprehensive indicator weights and the redistribution of weights under the condition of incomplete indicators; and calculation of the membership degree of the final evaluation score under each risk level, as shown in formula (6); calculation of the membership degree of the data group at the corresponding time for each level; and determination of the maximum membership degree as the roof safety risk level of the area at the current time.

[0146] (6);

[0147] In the formula, c represents the membership degree of the final evaluation score under risk level i. i This represents the central value of the membership degree under risk level i. x represents the final evaluation score;

[0148] S52: Roof risk visualization based on cloud models; based on the established roof safety risk level assessment criteria, construct basic cloud models for risk assessment corresponding to each risk level. Each basic cloud model is defined by three parameters: expected value (Ex), entropy (En), and hyperentropy (He), forming standard risk distribution clouds corresponding to different risk levels (Level I to Level V), such as... Figure 5 As shown;

[0149] S53: Assess the risk cloud; based on real-time collected static and dynamic indicator data, calculate the assessment score and convert it into basic cloud model parameters for risk assessment (Ex, En, He), forming the current risk assessment cloud, such as... Figure 6 As shown;

[0150] S54: Comparison and Risk Propensity Assessment; Compare the basic risk assessment cloud model with the current risk assessment cloud to obtain the roof safety risk tendency assessment result, providing a basis for auxiliary decision-making, such as... Figure 7 As shown.

[0151] Step Six: Based on the risk threshold set by the evaluation results, automatically activate the early warning equipment to provide simultaneous audible, visual, and electrical warnings and automatically report the warning information; the specific process is as follows:

[0152] When the roof safety risk reaches a set threshold score, the control system activates the audible, visual, and electrical warning devices deployed in the working face and roadways, sending alerts to the control and monitoring terminals. The corresponding audible, visual, and electrical warning measures are shown in Table 11. Specifically, for Level I roof risk, the control system keeps the green light on and maintains normal power supply, while simultaneously generating warnings for routine inspections and maintaining current monitoring and support measures. For Level II roof risk, the control system keeps the blue light on and activates the warning system, while simultaneously generating warnings to increase inspection frequency and continuously monitor changes in data. For Level III roof risk, the control system keeps the yellow light on and broadcasts a prompt voice, disconnecting non-critical equipment. Power supply is maintained only for critical systems. Simultaneously, early warning information is generated for adding monitoring points, conducting special inspections, and optimizing support measures. When a Level IV roof slab risk occurs, the orange lights on the control equipment flash frequently, and a continuous buzzer and voice prompt are broadcast. Power is cut off to the designated disconnected area, and the safety lighting and broadcasting systems are activated. At the same time, early warning information is generated to initiate the early warning response process, arrange for personnel evacuation from relevant areas, and strengthen the support structure. When a Level V roof slab risk occurs, the red lights on the control equipment flash frequently, continuously sound a long beep and broadcast a voice prompt, and a complete power outage is implemented, stopping power supply to all work areas. Simultaneously, early warning information is generated to immediately implement personnel evacuation, suspend operations, and carry out emergency reinforcement measures.

[0153] Table 11: Sound, light, and electrical early warning measures at corresponding levels

[0154] grade Voice Light electricity Ⅰ silent Green light always on Power supply is normal, no power outage has occurred, and the system remains in standby mode. Ⅱ silent Blue light constant The early warning system is activated, and the electronic control module remains in a ready state without power interruption. Ⅲ Voice prompts Yellow light constant Disconnect power to non-critical equipment, retain power to critical systems, and issue only a soft audible alert. Ⅳ Continuous beeping + voice broadcast Orange strobe Power outage in some areas triggers emergency lighting and public address systems. Ⅴ Long beep + continuous voice broadcast Red high-frequency flashing Initiate a complete power outage, forcibly shutting off power to all work areas to ensure personnel evacuation.

[0155] Step 7: By collecting information on roof risk warnings and comparing it with actual on-site information to form a comparative database, the weights of indicators are dynamically optimized, and the basic cloud model for risk assessment is revised. The specific process is as follows:

[0156] S71: Continuously collect early warning data generated by roof risk warnings, and simultaneously collect actual on-site monitoring data and handling feedback information.

[0157] S72: By constructing a database comparing early warning information with actual on-site data, analyze the matching degree and deviation between early warning results and actual risk events;

[0158] S73: Based on the comparative analysis results, the dynamic weight adjustment algorithm is used to repeat steps four and five to correct the weight distribution of risk assessment indicators in real time, eliminate the influence of misleading indicators, enhance the weight sensitivity of key indicators, and thus optimize the performance of the risk assessment model.

[0159] This process involves iterative refinement and calibration of model parameters to achieve adaptive updates of the risk assessment model, ensuring high predictive accuracy and stability under different operating conditions and at different time points. Simultaneously, the dynamic optimization process balances the model's generalization ability with local adaptability, enhancing its responsiveness to risk changes in complex mine rooftop environments.

[0160] This invention first determines the multi-source data of static and dynamic indicators required for roof safety risk assessment based on the needs of underground mine engineering, and establishes roof safety risk level assessment standards. Then, it establishes the correspondence between data ranges and risk scores for various indicators under different risk levels. This facilitates the collaborative analysis of static indicators (such as geological structures) and dynamic indicators (such as real-time stress changes), forming a more comprehensive multi-dimensional safety risk assessment framework and effectively ensuring the accurate assessment of roof risk levels. A dynamic data acquisition scheme is used to obtain static and dynamic indicator data, and qualitative indicators are quantified and analyzed through defuzzification. Standardization methods are used to improve the comparability of quantitative indicator data, effectively ensuring that the obtained multi-source indicator data accurately reflects roof status information. Data features are extracted separately through a subjective and objective weight fusion mechanism, providing accurate input feature guarantees for subsequent risk level identification models. Simultaneously obtaining comprehensive indicator weights and weight redistribution under incomplete indicator conditions allows the method to support both comprehensive indicators and weight redistribution under incomplete indicators, ensuring accurate and effective assessment of roof safety risks even when some indicator data is missing, effectively enhancing the adaptability of the indicator system. The risk level is dynamically determined by the principle of maximum membership. Risk tendency assessment accurately displays the degree of risk deviation and its development trend. Visualization maps the risk level, providing a clear and intuitive representation of the risk situation in real time. Based on the assessment results, early warning equipment is automatically activated for simultaneous audible, visual, and electrical warnings, enabling multi-channel warning linkage and ensuring timely and effective emergency response in dangerous situations. By dynamically optimizing indicator weights to correct the assessment module, a closed-loop monitoring-early warning-feedback-optimization process allows for adaptive updates to the risk assessment model, ensuring high predictive accuracy and stability under different operating conditions and time points.

[0161] This method is simple to implement and highly intelligent. It can efficiently and accurately assess the risk level of the roof and take corresponding early warning actions according to different roof risk levels. It is of great value for ensuring safe production operations in mines.

[0162] like Figure 8 As shown, the present invention also provides an intelligent early warning system for the risk of the roof of the mining area based on an uncertainty fusion model, which is used to implement an intelligent early warning method for the risk of the roof of the mining area based on an uncertainty fusion model, including a data acquisition module, a weight management module, a judgment standard module, a calculation logic module, a risk level judgment module, an early warning module, a visualization module and a dynamic optimization module;

[0163] The acquisition module is used to collect multi-source monitoring data in the area to be evaluated, record the spatial layout information of the monitoring equipment and the data acquisition frequency, form an original dataset, and send the original dataset to the weight management module and the risk level assessment module.

[0164] The weight management module is used to extract dataset features based on the volatility of the obtained data, the correlation between indicators and historical evaluation performance, initialize the weight vector of each evaluation indicator, support the allocation of incomplete indicator weights and subsequent model feedback for dynamic weight adjustment, and output the corrected indicator weight set to the risk level assessment module.

[0165] The evaluation criteria module is used to construct a multi-level risk level system. Based on expert experience, policies and regulations, historical cases and numerical analysis results, it defines the boundary thresholds and judgment logic for each level, forms a set of evaluation criteria for the risk level of underground mine roof, and sends it to the risk level evaluation module.

[0166] The computational logic module is used to perform normalization and defuzzification processing on the collected data information, and to calculate relevant parameters and results on the processed data through fuzzy comprehensive evaluation, cloud model or multi-factor fusion algorithm, and send them to the risk level assessment module to further support the efficient operation of the assessment module.

[0167] The risk level assessment module is used to receive the indicator data collected by the acquisition module, the indicator weights output by the weight management module, the quantitative analysis results output by the calculation logic module, and the risk level assessment standard set of the underground mine roof output by the assessment standard module. It then classifies the risk levels according to the maximum membership principle and the assessment standard module, and sends the risk level classification results to the early warning module and the visualization module.

[0168] The early warning module is used to execute corresponding sound, light, and electrical early warning measures based on the risk level output by the risk level assessment module.

[0169] The visualization module is used to display real-time assessment results, historical trend evolution, risk area distribution and early warning status in a graphical way, and to provide a human-computer interaction interface, supporting two-dimensional and three-dimensional risk maps, time series diagrams, risk cloud maps and console operation interfaces, thereby improving information understanding and interaction efficiency.

[0170] The dynamic optimization module is used to collect deviation information between risk level assessment results and actual on-site feedback, construct an assessment-measurement comparison database, and dynamically correct the indicator weights based on the comparison database. At the same time, the corrected indicator weights are sent to the weight management module to achieve adaptive optimization of the assessment model.

[0171] The system has a simple structure and a high degree of intelligence. It can intelligently and accurately identify and dynamically warn of the risk level of the mine roof, effectively improving the safety management level of the mine roof.

Claims

1. A method for intelligent early warning of roof risk in mining areas based on an uncertainty fusion model, characterized in that, Includes the following steps; Step 1: Based on the requirements of underground mine engineering, determine the static and dynamic indicators needed for roof safety risk assessment, and establish roof safety risk level assessment standards; Step 2: Based on numerical simulation and engineering standards, define the data range and scoring range corresponding to the indicators under different safety levels; Step 3: Based on data requirements, adopt a dynamic data acquisition scheme to obtain static and dynamic indicator data, and perform deblurring and standardization processing on the obtained data; Step 4: Based on data-driven approach, extract data features by combining multiple subjective and objective weighting methods, and obtain the comprehensive index weights and the weights redistributed under the condition of incomplete indicators. Step 5: Based on the current optimal weight, assess the safety risk level of the roof using the maximum membership principle, obtain the risk tendency judgment result, and visualize it through the basic cloud model of risk assessment; Step Six: Based on the risk threshold set by the evaluation results, automatically activate the early warning equipment to provide simultaneous sound, light, and electrical warnings and automatically report the warning information; Step 7: By collecting information on roof risk warnings and comparing it with actual on-site information, a comparative database is formed, the weights of indicators are dynamically optimized, and the basic cloud model for risk assessment is revised.

2. The intelligent early warning method for mining roof risk based on an uncertainty fusion model according to claim 1, characterized in that, In step one, static index data includes roof rock mass grade, stress field data, mining technology, and historical accident information; dynamic index data includes roof displacement data, stress data, and crack conditions; and roof safety risk levels include Level I roof risk, Level II roof risk, Level III roof risk, Level IV roof risk, and Level V roof risk. Specifically, the scoring range for Level I roof risk is ≥81 and ≤100; for Level II roof risk, it is ≥61 and ≤80; for Level III roof risk, it is ≥41 and ≤60; for Level IV roof risk, it is ≥21 and ≤40; and for Level V roof risk, it is ≥0 and ≤20.

3. The intelligent early warning method for mining roof risk based on an uncertainty fusion model according to claim 2, characterized in that, The correspondence in step two is as follows: For the roof rock mass grade, it is divided into five grades according to the rock mass stability: extremely stable, stable, moderate, unstable and extremely unstable, which correspond to Grade I roof risk, Grade II roof risk, Grade III roof risk, Grade IV roof risk and Grade V roof risk, respectively. Among them, extremely stable corresponds to 81≤rock mass stability≤100, stable corresponds to 61≤rock mass stability≤80, moderate corresponds to 41≤rock mass stability≤60, unstable corresponds to 21≤rock mass stability≤40, and extremely unstable corresponds to 0≤rock mass stability≤20. For stress field data, it is divided into five levels according to the magnitude of ground stress: no risk, low risk, potential risk, high risk and extremely high risk, which correspond to the roof risk of Class I, Class II, Class III, Class IV and Class V respectively. Among them, no risk corresponds to ground stress < 5 MPa, low risk corresponds to 5 MPa ≤ ground stress < 10 MPa, potential risk corresponds to 10 MPa ≤ ground stress < 15 MPa, high risk corresponds to 15 MPa ≤ ground stress < 20 MPa, and extremely high risk corresponds to ground stress > 20 MPa. Mining processes are classified into five levels based on the disturbance force of each method: extremely low disturbance, low disturbance, moderate disturbance, high disturbance, and relatively high disturbance. These levels correspond to Class I, Class II, Class III, Class IV, and Class V roof risks, respectively. Specifically, extremely low disturbance corresponds to bench mining, residual ore recovery, and shallow layered mining; low disturbance corresponds to layered mining, shallow room-and-pillar mining, and strip mining; moderate disturbance corresponds to room-and-pillar mining, longwall mining, sublevel caving, and room-and-pillar mining with ore-holding zones; high disturbance corresponds to sublevel backfilling mining, staged open area mining, staged backfilling, and staged caving; and relatively high disturbance corresponds to deep full-face mining, large-stope unsupported caving, block caving, continuous mining, and full-face mining. Based on the average number of roof falls per year, historical accident data is categorized into five levels: safe, low risk, medium risk, high risk, and extremely high risk. These levels correspond to Level I, Level II, Level III, Level IV, and Level V roof risks, respectively. Specifically, safe corresponds to no roof falls per year; low risk corresponds to 0 to 1 roof falls per year; medium risk corresponds to 2 to 1 roof falls per year; high risk corresponds to 3 to 2 roof falls per year; and extremely high risk corresponds to more than 3 roof falls per year. For roof displacement data, it is divided into five levels according to the roof displacement situation: low risk, relatively low risk, medium risk, high risk and extremely high risk, which correspond to Level I roof risk, Level II roof risk, Level III roof risk, Level IV roof risk and Level V roof risk, respectively. Among them, low risk corresponds to 0 ≤ daily average displacement rate ≤ 1 mm / d and 0 ≤ cumulative displacement ≤ 10 mm; relatively low risk corresponds to 1 mm / d < daily average displacement rate ≤ 3 mm / d and 10 mm < cumulative displacement ≤ 20 mm; medium risk corresponds to 3 mm / d < daily average displacement rate ≤ 5 mm / d and 20 mm < cumulative displacement ≤ 40 mm; high risk corresponds to 5 mm / d < daily average displacement rate ≤ 8 mm / d and 40 mm < cumulative displacement ≤ 60 mm; and extremely high risk corresponds to daily average displacement rate > 8 mm / d and cumulative displacement > 60 mm. For stress data, it is divided into five levels according to the stress condition of the support / column: stable pressure, moderate pressure, leakage pressure, heavy pressure, and overpressure. These levels correspond to the roof risk levels of Class I, Class II, Class III, Class IV, and Class V, respectively. Specifically, stable pressure corresponds to 30% of the design bearing capacity ≤ support / column stress < 50% of the design bearing capacity; moderate pressure corresponds to 50% of the design bearing capacity ≤ support / column stress < 70% of the design bearing capacity; leakage pressure corresponds to support / column stress < 30% of the design bearing capacity; heavy pressure corresponds to 70% of the design bearing capacity ≤ support / column stress < 90% of the design bearing capacity; and overpressure corresponds to 70% of the design bearing capacity ≥ 90% of the design bearing capacity. For crack conditions, they are classified into five levels according to crack size: no crack, fine crack, stable crack, expanding crack, and penetrating crack. These correspond to the roof risk levels of Class I, Class II, Class III, Class IV, and Class V, respectively. Specifically, no crack corresponds to no obvious cracks or no cracks; fine crack corresponds to the presence of small cracks with a crack width < 1 mm; stable crack corresponds to the presence of non-penetrating cracks with a crack width of 1 mm ≤ crack width < 3 mm; expanding crack corresponds to the presence of non-penetrating cracks with a crack width of 3 mm ≤ crack width < 5 mm; and penetrating crack corresponds to the presence of penetrating cracks with a crack width ≥ 5 mm.

4. The intelligent early warning method for mining roof risk based on an uncertainty fusion model according to claim 3, characterized in that, In step three, static index data are obtained by technical and design personnel through simulation calculation based on the mine geological report and are manually input into the data processing terminal. Dynamic index data are collected in real time and automatically using monitoring equipment deployed at the mine site to obtain static and dynamic index data. The monitoring equipment is deployed differently in coal mines and non-coal mines.

5. The intelligent early warning method for mining roof risk based on an uncertainty fusion model according to claim 4, characterized in that, In step three, the specific process of deblurring and standardizing the obtained data is as follows: S31; Introducing the rule mapping defuzzification mechanism from fuzzy set theory to convert qualitative indicators into quantitative data, and obtaining the level score corresponding to the i-th indicator according to formula (1). ; （1）； In the formula, Indicates the corresponding indicator level; S32: Perform linear standardization on indicators with actual data; for data ranges of... The score range is Data value X in When, the corresponding score range is The score of the j-th index after nonlinear standardized mapping is obtained according to formula (2). , （2）。 6. The intelligent early warning method for mining roof risk based on an uncertainty fusion model according to claim 5, characterized in that, The specific process of step four is as follows: S41: At the subjective level, an expert evaluation matrix is ​​constructed based on the analytic hierarchy process to extract expert knowledge and experience; at the objective level, the importance of indicators is quantified from the perspectives of information distribution and volatility by combining the entropy method and the coefficient of variation method; finally, a unified weighting system is formed by integrating the three methods through a linear weighted fusion strategy, which provides accurate input feature guarantee for the subsequent risk level identification model. Specifically, the comprehensive weight is obtained according to formula (3). ; （3）； In the formula, Weights for the analytic hierarchy process (AHP); Weights are determined by entropy. Weights are calculated using the coefficient of variation method. , representing the reliability coefficient of the analytic hierarchy process; , where is the reliability coefficient of the entropy method; The reliability coefficient is determined by the coefficient of variation method. ; S42: For a weight vector containing n indicators , , When m indicators are missing, the weight of the i-th indicator after redistribution is obtained according to formula (4). According to formula (5), the following formula is obtained: Weight vector of each indicator ; （4）； （5）； In the formula, The weight of the i-th indicator before redistribution; The weighted sum of the residual indicators before redistribution; The number of participating indicators, ; .

7. The intelligent early warning method for mining roof risk based on an uncertainty fusion model according to claim 6, characterized in that, The specific process of step five is as follows: S51: Collect static and dynamic indicator data of the evaluated area in real time, calculate the final evaluation score based on the established roof safety risk level evaluation standard, the data range and level classification of various indicators and the corresponding relationship of roof safety risk level, the comprehensive indicator weight and the redistribution of weight under the condition of incomplete indicators, and calculate the membership degree of the final evaluation score under each risk level, as shown in formula (6). Calculate the membership degree of each level for the data group at the corresponding time, and determine the maximum membership degree as the roof safety risk level of the corresponding area at the current time; （6）； In the formula, This indicates the membership degree of the final evaluation score under risk level i. This represents the central value of the membership degree under risk level i. x represents the final evaluation score; S52: Based on the established roof safety risk level assessment criteria, construct a basic cloud model for risk assessment corresponding to each risk level. Each basic cloud model for risk assessment is defined by three parameters: expected value, entropy, and hyperentropy, forming a standard risk distribution cloud corresponding to different risk levels. S53: Based on the real-time collected static and dynamic indicator data, calculate the evaluation score and convert it into basic cloud model parameters for risk assessment to form the current risk assessment cloud; S54: Compare the basic risk assessment cloud model with the current risk judgment cloud to obtain the roof safety risk tendency judgment result, so as to provide a basis for auxiliary decision-making.

8. The intelligent early warning method for mining roof risk based on an uncertainty fusion model according to claim 7, characterized in that, The specific process of step six is ​​as follows: When the roof safety risk reaches a set threshold score, the early warning devices deployed in the working face and roadways will execute audible, visual, and electrical early warning actions and send them to the control and monitoring terminal. Specifically, when a Level I roof risk occurs, the green light on the control early warning device will remain constantly lit, and normal power supply will be maintained. At the same time, an early warning message will be generated to ensure daily inspections and maintain the current monitoring and support measures. When a Level II roof risk occurs, the blue light on the control early warning device will remain constantly lit, the early warning system will be activated, and an early warning message will be generated to increase the frequency of inspections and continuously observe changes in monitoring data. When a Level III roof risk occurs, the yellow light on the control early warning device will remain constantly lit, and a prompt voice will be broadcast. Power supply to non-critical equipment will be disconnected, and only important systems will be kept powered on. The system generates early warning information for adding monitoring points, conducting special investigations, and optimizing support measures. When a Level IV roof slab risk occurs, the system controls the orange lights of the early warning equipment to flash frequently, and broadcasts continuous beeps and prompts. Power is cut off to the designated disconnected areas, and the system is linked to safety lighting and broadcasting systems. Simultaneously, it generates early warning information for initiating the early warning response process, arranging for personnel to evacuate the relevant areas, and reinforcing the support structure. When a Level V roof slab risk occurs, the system controls the red lights of the early warning equipment to flash frequently, continuously broadcast long beeps and voice prompts, and completely cuts off power to all work areas. Simultaneously, it generates early warning information for immediately evacuating personnel, suspending work, and implementing emergency reinforcement measures.

9. The intelligent early warning method for mining roof risk based on an uncertainty fusion model according to claim 8, characterized in that, The specific process of step seven is as follows: S71: Continuously collect early warning data generated by roof risk warnings, and simultaneously collect actual on-site monitoring data and handling feedback information. S72: By constructing a database comparing early warning information with actual on-site data, analyze the matching degree and deviation between early warning results and actual risk events; S73: Based on the comparative analysis results, the dynamic weight adjustment algorithm is used to repeat steps four and five to correct the weight distribution of risk assessment indicators in real time, eliminate the influence of misleading indicators, enhance the weight sensitivity of key indicators, and thus optimize the performance of the risk assessment model.

10. A smart early warning system for mining roof risk based on an uncertainty fusion model, used to implement the smart early warning method for mining roof risk based on an uncertainty fusion model as described in any one of claims 1 to 9, characterized in that, It includes a data acquisition module, a weight management module, a judgment standard module, a calculation logic module, a risk level judgment module, an early warning module, a visualization module, and a dynamic optimization module; The acquisition module is used to collect multi-source monitoring data in the area to be evaluated, record the spatial layout information of the monitoring equipment and the data acquisition frequency, form an original dataset, and send the original dataset to the weight management module and the risk level assessment module. The weight management module is used to extract dataset features based on the volatility of the obtained data, the correlation between indicators and historical evaluation performance, initialize the weight vector of each evaluation indicator, support the allocation of incomplete indicator weights and subsequent model feedback for dynamic weight adjustment, and output the corrected indicator weight set to the risk level assessment module. The evaluation criteria module is used to construct a multi-level risk level system. Based on expert experience, policies and regulations, historical cases and numerical analysis results, it defines the boundary thresholds and judgment logic for each level, forms a set of evaluation criteria for the risk level of underground mine roof, and sends it to the risk level evaluation module. The computational logic module is used to perform normalization and defuzzification processing on the collected data information, and to calculate relevant parameters and results on the processed data through fuzzy comprehensive evaluation, cloud model or multi-factor fusion algorithm, and send them to the risk level assessment module. The risk level assessment module is used to receive the indicator data collected by the acquisition module, the indicator weights output by the weight management module, the quantitative analysis results output by the calculation logic module, and the risk level assessment standard set of the underground mine roof output by the assessment standard module. It then classifies the risk levels according to the maximum membership principle and the assessment standard module, and sends the risk level classification results to the early warning module and the visualization module. The early warning module is used to execute corresponding sound, light, and electrical early warning measures based on the risk level output by the risk level assessment module. The visualization module is used to display real-time assessment results, historical trend evolution, risk area distribution and early warning status in a graphical way, and to provide a human-computer interaction interface, supporting two-dimensional and three-dimensional risk maps, time series diagrams, risk cloud maps and console operation interfaces. The dynamic optimization module is used to collect deviation information between risk level assessment results and actual on-site feedback, construct an assessment-measurement comparison database, and dynamically correct the indicator weights based on the comparison database. At the same time, the corrected indicator weights are sent to the weight management module to achieve adaptive optimization of the assessment model.

Images