Deep mine rockburst risk grading early warning and active pressure relief system and method

Abstract

The invention discloses a deep mine rockburst risk grading early warning and active pressure relief system and method, and relates to the technical field of rockburst risk early warning. The system comprises a multi-source parameter sensing module, a risk grading early warning module, an active pressure relief execution module, an intelligent control center module and a data storage and interaction module, wherein the multi-source sensing module comprises a distributed sensor and a mobile inspection robot; risk early warning is divided into four levels, and sensitive features are extracted after data are preprocessed; active pressure relief comprises fixed-point and regional pressure relief equipment; the control center uses a redundant architecture to automatically match a pressure relief scheme; and data storage retains more than or equal to 5 years of data, and multi-system docking is supported. The rock burst early warning accuracy and the pressure relief pertinence are improved, multi-module closed-loop cooperation is achieved, emergency response is efficient, false alarm, missing alarm and pressure relief blindness are reduced, roadway stability and personnel safety are guaranteed, and support is provided for safe and efficient mining of deep mines.

Description

Technical Field

[0001] This invention relates to the field of rockburst risk early warning technology, and in particular to a rockburst risk classification, early warning and active pressure relief system and method for deep mines. Background Technology

[0002] As mineral resource development extends deeper, the proportion of mines buried at depths exceeding 800 meters is increasing annually, significantly raising the frequency and severity of rockburst disasters in high-stress deep environments. Rockbursts, a sudden geological hazard, are triggered by the sudden release of elastic strain energy accumulated in rock masses. They can instantly generate strong vibrations, shock waves, and rock ejection, leading to tunnel collapses, equipment damage, and even casualties, severely restricting the safety and production efficiency of deep mines. Statistics show that production interruptions due to rockbursts in deep metal mines average 30-60 days annually, with direct economic losses accounting for 20%-30% of total mine losses. Therefore, accurate early warning and effective control of rockburst risks have become a core technical challenge in deep mine development.

[0003] Existing rockburst prevention technologies have several limitations: Monitoring methods often rely on single parameters (such as stress or vibration) for judgment, lacking comprehensive consideration of geological conditions, mining disturbances, and other factors, resulting in an early warning accuracy rate generally below 60%. Risk classification standards often use fixed thresholds, failing to consider the differences and dynamic changes in geological conditions across different mines, easily leading to false alarms or missed alarms. Pressure relief measures are primarily reactive; for example, the timing of blasting pressure relief lags behind risk warnings, and the pressure relief range and intensity are determined empirically, often resulting in insufficient pressure relief leading to continued rockburst risk or excessive pressure relief causing roadway instability. Furthermore, existing systems mostly consist of independent monitoring, early warning, and pressure relief modules, lacking real-time data interaction and intelligent decision-making linkage, making it difficult to form a closed-loop control of "perception-assessment-decision-execution," and unable to adapt to the complex and variable high-stress environment of deep mines.

[0004] In recent years, although some studies have introduced the Internet of Things and intelligent algorithms to improve rockburst early warning capabilities, key technical bottlenecks still exist: the fusion of multi-source parameters does not consider the dynamic changes in parameter sensitivity, resulting in insufficient feature extraction accuracy; the decompression scheme lacks an effect prediction model, making it difficult to balance decompression effect and engineering safety; emergency response relies on manual command, resulting in low coordination efficiency and failing to meet the needs for rapid evacuation in the event of a sudden rockburst. Therefore, developing an integrated system with multi-parameter dynamic perception, intelligent hierarchical early warning, precise active decompression, and multi-system coordinated response functions has become an urgent need to solve the rockburst problem in deep mines. Summary of the Invention

[0005] The present invention proposes a risk classification, early warning and active pressure relief system and method for rockburst in deep mines to solve the problems mentioned in the prior art.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a deep mine rockburst risk classification, early warning, and active pressure relief system, comprising:

[0007] Parameter sensing module: Includes distributed sensing units and mobile monitoring units. The distributed sensing units are deployed every 5m along the tunnel outline. Each set includes a fiber optic stress sensor, a three-component vibration sensor, an acoustic sensor, and an infrared temperature sensor. The sensors are connected to the data acquisition substation via waterproof armored cables. The mobile monitoring unit is mounted on an explosion-proof inspection robot, equipped with a portable borehole stress gauge and a high-definition industrial camera. It conducts supplementary monitoring of the area along a preset track, and the data is uploaded to the control center via a 5G explosion-proof base station.

[0008] The risk classification and early warning module includes a data preprocessing unit, a feature extraction unit, and a classification decision unit. The data preprocessing unit uses a wavelet threshold denoising algorithm to remove interference noise from mining equipment in stress and vibration signals, smooths temperature and acoustic data using Kalman filtering, and uses the 3σ criterion to determine outliers, marking parameters that exceed the range and triggering secondary acquisition. The feature extraction unit extracts rockburst-sensitive features from the preprocessed data, including stress concentration factor, vibration signal frequency, acoustic wave velocity change rate, and temperature gradient. The classification decision unit classifies rockburst risk into four levels: Level I, Level II, Level III, and Level IV. Early warning information is simultaneously released through roadway audible and visual alarms, management personnel's handheld terminals, and the mine dispatch center's large screen.

[0009] Active pressure relief execution module: includes a fixed-point pressure relief unit and a regional pressure relief unit: the fixed-point pressure relief unit deploys hydraulic fracturing equipment for stress concentration areas and is equipped with an automatic drill rod changing drill to construct pressure relief holes according to preset hole pattern parameters; the regional pressure relief unit is suitable for stress areas and includes a carbon dioxide phase change fracturing system and a deep hole pre-fracturing blasting device, which achieves micro-differential pressure relief through a programmed detonator; all pressure relief equipment is equipped with displacement sensors and pressure sensors to provide real-time feedback on the execution status;

[0010] The control center module adopts an industrial computer + PLC redundant architecture, equipped with a rockburst risk assessment algorithm and a pressure relief decision model: the risk assessment algorithm integrates parameter features and calculates real-time risk values ​​through a BP neural network; the pressure relief decision model automatically matches pressure relief schemes based on risk levels and allows manual adjustment of parameters through a human-machine interface;

[0011] Data storage and interaction module: Includes local server and cloud platform, using distributed database to store raw sensor data, risk assessment results, depressurization execution records and equipment status parameters; the interaction interface supports connection with mine safety monitoring system and production scheduling system, pushes real-time risk reports and depressurization suggestions to management platform, and receives mine production plan to dynamically adjust monitoring frequency and depressurization range; data transmission uses AES-256 encryption algorithm.

[0012] Furthermore, it also includes a multi-source parameter dynamic weighted fusion unit, which adjusts the weights in real time based on the differences in the sensitivity of each parameter to rockburst; the weight coefficient is determined by calculating the parameter contribution, and the higher the contribution, the greater the weight. The contribution of the stress parameter is calculated by the following formula: In the formula, Contribution to stress parameters To monitor the maximum stress value in the area, To monitor the average stress value of the area, The stress change time interval is denoted by k, the distance attenuation coefficient is denoted by d, and the distance from the measuring point to the working face is denoted by d. The calculation method for the contribution of vibration parameters is similar, except that the input parameters are replaced with peak ground acceleration and dominant frequency. Finally, the weights of each parameter are normalized according to their contribution.

[0013] Furthermore, it also includes a dynamic risk level correction unit, which updates the classification standards by combining historical rockburst cases with real-time geological conditions. The correction is based on the parameter characteristics of rockbursts that occurred in the same geological unit within the past 3 months, as well as the current roadway support strength and mining disturbance intensity. When fault activation or water inrush signs are detected, the judgment thresholds for Level III and Level IV risks are automatically lowered. At the same time, the correction unit generates a classification standard validity report every 72 hours by comparing the actual risks with the early warning results.

[0014] Furthermore, the active pressure relief execution module also includes a pressure relief effect prediction unit, which evaluates the rationality of the scheme by calculating the stress release amount after pressure relief; the stress release amount calculation formula is as follows: In the formula, The stress is released after pressure relief, α is the Poisson's ratio of the rock, and E is the elastic modulus of the rock. The volume of the pressure relief chamber. To affect the volume of the region, To cause cracking working pressure, This represents the original geostress.

[0015] Furthermore, the control center module also includes an emergency linkage unit, which automatically activates the mine emergency response mechanism when a Level IV risk warning is issued. The linkage includes: sending evacuation instructions to the underground personnel positioning system and displaying the evacuation route on the LED screen in the roadway; controlling the emergency stop of the underground transportation system and sealing the entrance to the risk area; sending adjustment instructions to the ventilation system to increase the air volume in the risk area and reduce the risk of gas accumulation; and simultaneously connecting the underground refuge chamber with the ground command center through the emergency communication system to transmit on-site images and personnel status in real time.

[0016] Furthermore, the parameter sensing module also includes a microseismic positioning unit. This unit determines the location of the microseismic source by analyzing the signal time difference of three acoustic sensors. A particle swarm optimization-based positioning algorithm is used to calculate the microseismic source coordinates (x, y, z). A terrain correction coefficient is introduced to correct the wave velocity deviation caused by roadway undulations. The correction coefficient is calculated using the following formula: In the formula, β is the terrain correction coefficient, and θ is the tunnel slope angle.

[0017] Furthermore, the data storage and interaction module also includes a digital twin mapping unit, which constructs a virtual digital twin of the mine roadway and system to achieve real-time synchronization between the physical entity and the virtual model. The twin includes a geological model, an equipment model, and an environmental model. The virtual model is updated by real-time collected physical parameters to simulate the stress evolution process under different pressure relief schemes and predict the risk change trend within 30 minutes. At the same time, it supports the pre-simulation of pressure relief schemes in the virtual environment, calculates the stress release effect corresponding to different hole network parameters and fracturing pressures, and provides visual decision support for optimizing pressure relief strategies.

[0018] A method for applying the aforementioned deep mine rockburst risk classification, early warning, and active pressure relief system includes:

[0019] Parameter acquisition and preprocessing steps: The parameter sensing module is activated, and distributed sensors acquire stress, vibration, sound wave, and temperature data at a frequency of 1kHz. The mobile inspection robot conducts supplementary monitoring of key areas every 2 hours. The raw data is transmitted to the data preprocessing unit via fiber optic bus and 5G network. The stress and vibration signals are processed using wavelet threshold denoising algorithm, and the temperature and sound wave data are smoothed by Kalman filtering. Outliers are removed using the 3σ criterion. The marked abnormal parameters trigger sensor self-check and secondary acquisition.

[0020] Feature extraction and data fusion steps: The feature extraction unit extracts rockburst sensitive features from the preprocessed data, including stress concentration factor, vibration signal dominant frequency, sound wave velocity change rate and temperature gradient; the parameter dynamic weighted fusion unit calculates the contribution of each feature, determines the weight according to the contribution, and obtains the fused feature value by weighted summation;

[0021] Risk grading and early warning release steps: The risk grading decision unit compares the fusion characteristic value with the grading standard to determine the risk level, where Level I corresponds to a fusion characteristic value < 0.3, Level II corresponds to 0.3 ≤ fusion characteristic value < 0.5, Level III corresponds to 0.5 ≤ fusion characteristic value < 0.7, and Level IV corresponds to a fusion characteristic value ≥ 0.7; The risk level dynamic correction unit adjusts the grading threshold based on historical cases and real-time geological conditions. When fault activation is present, the Level IV threshold is reduced from 0.7 to 0.6; Early warning information is simultaneously released through tunnel audible and visual alarms, handheld terminals, and the dispatch center's large screen, with Level IV early warnings accompanied by emergency SMS pushes to the mine manager;

[0022] Active pressure relief scheme decision and execution steps: The control center automatically matches the pressure relief scheme according to the risk level. Level I does not require pressure relief, Level II initiates fixed-point hydraulic fracturing, Level III initiates fixed-point + local area pressure relief, and Level IV initiates full-area coordinated pressure relief. The pressure relief effect prediction unit calculates the stress release amount. If the requirements are not met, the pressure relief parameters are adjusted. The active pressure relief execution module controls the drilling rig and fracturing equipment according to the scheme and provides real-time feedback of displacement and pressure data.

[0023] Effect feedback and system optimization steps: 30 minutes after depressurization, the parameter sensing module re-collects data and calculates the risk level after depressurization. If it is still ≥ Level III, a second depressurization is triggered. The data storage and interaction module records the parameters, effects, and risk change curves of this depressurization. The digital twin mapping unit reviews the depressurization process. The system operation data is analyzed weekly to optimize feature weights, grading thresholds, and depressurization parameters, so that the system can continuously adapt to changes in mine geological conditions.

[0024] Furthermore, it also includes:

[0025] Microseismic source location and risk tracing steps: The microseismic location unit receives signals from at least three acoustic sensors, records the signal arrival time difference, and uses a particle swarm optimization algorithm to calculate the microseismic source coordinates (x, y, z). A terrain correction coefficient is introduced to correct the error caused by the roadway slope. The location results are superimposed on the three-dimensional model of the mine to analyze the correlation between the distribution of microseismic sources and the geological structure. If 70% of the microseismic sources are concentrated near the fault, the fault is identified as a risk source. An enhanced monitoring plan is formulated for the risk source, and pressure relief holes are densified within 5m on both sides of the fault.

[0026] Emergency Response and Collaborative Evacuation Procedures: When a Level IV risk warning is issued, the emergency response unit immediately activates the emergency response, sends evacuation instructions to the underground personnel positioning system, plans the optimal evacuation route based on the real-time location of personnel, and dynamically updates route guidance on the LED screen in the tunnel; controls the emergency stop of the underground transportation system, seals the entrance to the risk area to prevent personnel from accidentally entering; adjusts the ventilation system to increase the airflow in the risk area and reduce the gas concentration; connects the refuge chamber with the ground command center through the emergency communication system to transmit on-site images and personnel status in real time; once the risk level drops to Level II, the emergency status is lifted after manual confirmation of safety.

[0027] Compared with existing technologies, the beneficial effects of this invention are:

[0028] In terms of risk early warning, the use of multi-source parameter dynamic sensing and weighted fusion technology breaks through the limitations of traditional single parameter monitoring. By combining geological conditions and mining disturbances to dynamically correct risk levels, the accuracy and timeliness of early warning are improved, effectively reducing false alarms and missed alarms, and providing mines with more reliable risk prediction basis.

[0029] In the active depressurization phase, by introducing a depressurization effect prediction model and precise execution control, the depressurization scheme was scientifically matched and dynamically optimized. This not only avoided the continued risk of rockburst due to insufficient depressurization, but also prevented roadway instability caused by excessive depressurization. This improved the pertinence and safety of depressurization measures, while reducing the interference of depressurization operations on normal production.

[0030] The system's intelligent collaboration has been significantly enhanced. Multiple modules form a closed-loop control through real-time data interaction, achieving seamless integration from parameter acquisition and risk assessment to pressure relief execution and emergency response, resulting in a significant improvement in efficiency compared to the traditional independent operation mode. The application of digital twin and microseismic location technologies provides visual support for risk tracing and scheme optimization, making prevention and control strategies more aligned with the actual geological conditions of the mine.

[0031] Furthermore, the establishment of an emergency response mechanism allows for rapid activation of multi-system collaborative risk avoidance during high-risk warnings, significantly enhancing the safety of underground personnel and reducing disaster losses. Overall, this invention achieves a shift from passive response to proactive prevention in rockburst control, providing comprehensive technical support for safe and efficient mining in deep mines, and yielding significant safety and economic benefits. Attached Figure Description

[0032] Figure 1 This is a schematic block diagram of the deep mine rockburst risk classification, early warning, and active pressure relief system proposed in this invention;

[0033] Figure 2 This is a schematic block diagram of the deep mine rockburst risk classification, early warning, and active pressure relief method proposed in this invention;

[0034] Figure 3 This is a schematic diagram comparing the accuracy of rockburst early warning under different monitoring methods of the rockburst risk classification and early warning and active pressure relief method for deep mines proposed in this invention.

[0035] Figure 4 This is a schematic diagram comparing the stress release effects of different stress relief schemes under the deep mine rockburst risk classification and early warning and active stress relief method proposed in this invention. Detailed Implementation

[0036] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0037] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0038] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified. Furthermore, the terms "installed," "connected," and "linked" should be interpreted broadly; for example, they may refer to a fixed connection, a detachable connection, or an integral connection; they may refer to a mechanical connection or an electrical connection; they may refer to a direct connection or an indirect connection through an intermediate medium; and they may refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances. The invention will now be described in further detail with reference to the accompanying drawings.

[0039] Reference Figures 1 to 4 A rockburst risk classification, early warning, and active pressure relief system for deep mines, comprising:

[0040] The parameter sensing module comprises a distributed sensing unit and a motion monitoring unit: The distributed sensing units are deployed every 5 meters along the tunnel outline. Each group includes a fiber optic stress sensor (measurement range 0-600MPa, accuracy ±1MPa, sampling frequency 1kHz), a three-component vibration sensor (range ±500g, frequency response 0.1-10kHz), an acoustic sensor (receiving frequency 20Hz-2MHz, sensitivity ≥90dB), and an infrared temperature sensor (measurement range -20-80℃, accuracy ±0.5℃). The sensors are protected by waterproof armored batteries. The cable (temperature resistant -40-120℃, pressure resistant 10MPa) is connected to the data acquisition substation; the mobile monitoring unit is mounted on the explosion-proof inspection robot (endurance ≥8h, walking speed 0.5-1.5m / s), including a portable borehole stress gauge (measuring depth 0-5m, resolution 0.1MPa) and a high-definition industrial camera (5 million pixels, night vision distance ≥30m), which can perform supplementary monitoring of key areas (such as fault fracture zones, goaf edges) along a preset track, and the data is uploaded to the control center via a 5G explosion-proof base station (transmission rate ≥100Mbps).

[0041] The risk grading and early warning module includes a data preprocessing unit, a feature extraction unit, and a grading decision-making unit. The data preprocessing unit uses a wavelet threshold denoising algorithm (threshold coefficient 0.65) to remove interference noise from mining equipment in stress and vibration signals. It smooths temperature and acoustic data using Kalman filtering (process noise covariance Q=0.02, measurement noise covariance R=0.1). Outlier detection uses the 3σ criterion (σ is the data standard deviation), marking parameters outside the range and triggering secondary acquisition. The feature extraction unit extracts rockburst-sensitive features from the preprocessed data, including stress concentration factor (maximum stress / average stress), vibration signal dominant frequency (10-500Hz), and acoustic wave velocity change rate (wave velocity difference per unit time / initial wave velocity). Temperature gradient (temperature difference / distance between adjacent measuring points); the graded decision-making unit divides rockburst risk into four levels: Level I (no risk), Level II (low risk), Level III (medium risk), and Level IV (high risk). Level I corresponds to stress concentration coefficient <1.2 and dominant frequency <50Hz; Level II corresponds to 1.2≤stress concentration coefficient <1.5 and 50Hz≤dominant frequency <100Hz; Level III corresponds to 1.5≤stress concentration coefficient <2.0 and 100Hz≤dominant frequency <200Hz; and Level IV corresponds to stress concentration coefficient ≥2.0 and dominant frequency ≥200Hz. Early warning information is simultaneously released through roadway audible and visual alarms (sound pressure level ≥110dB, flash frequency 2Hz), management personnel handheld terminals (response time ≤1s), and the mine dispatch center's large screen.

[0042] The active pressure relief module includes a fixed-point pressure relief unit and a regional pressure relief unit: the fixed-point pressure relief unit deploys hydraulic fracturing equipment (working pressure 30-60MPa, fracturing radius 1-3m) for stress concentration areas (such as roadway sides and roof), equipped with an automatic drill rod changing drilling rig (drilling diameter 50-100mm, depth 5-15m), and can construct pressure relief holes according to preset hole pattern parameters (hole spacing 2-5m, row spacing 1-3m); the regional pressure relief unit is suitable for large-scale high-stress areas and includes a carbon dioxide phase change fracturing system (filled with...) The pressure is 20-30MPa, fracturing energy is 50-300kJ) and the deep hole pre-splitting blasting device (bore diameter 100-200mm, charge coefficient 0.6-0.8) achieves multi-stage micro-differential pressure relief through a programmable detonator (delay accuracy ±1ms) to avoid secondary rock bursts induced during the pressure relief process; the pressure relief equipment is equipped with displacement sensors (measuring range 0-500mm, accuracy ±0.1mm) and pressure sensors (range 0-100MPa, accuracy ±0.5MPa) to provide real-time feedback on the execution status.

[0043] The intelligent control hub module adopts a redundant architecture of "industrial computer + PLC" (CPU main frequency ≥3.0GHz, memory ≥16GB, PLC input / output points ≥128 points), and is equipped with a rockburst risk assessment algorithm and a pressure relief decision model: The risk assessment algorithm integrates multi-source parameter features and calculates real-time risk values ​​through an improved BP neural network (15 hidden layer nodes, learning rate 0.01, 1000 iterations); The pressure relief decision model automatically matches the pressure relief scheme based on the risk level: Level I requires no pressure relief, Level II initiates fixed-point hydraulic fracturing (single-hole fracturing pressure 30MPa), Level III initiates fixed-point + local area pressure relief (carbon dioxide fracturing energy 100kJ), and Level IV initiates full-domain coordinated pressure relief (deep-hole blasting + carbon dioxide fracturing combination), and parameters can be manually adjusted through the human-machine interface; The control hub has a self-diagnostic function, and when the communication of a certain module is interrupted for ≥10s, it automatically switches to the backup channel (such as fiber optic to wireless) to ensure continuous system operation.

[0044] The data storage and interaction module includes a local server (storage capacity ≥10TB, read / write speed ≥500MB / s) and a cloud platform (supporting millions of concurrent data transactions). It uses a distributed database (MySQL+MongoDB) to store raw sensor data, risk assessment results, depressurization execution records, and equipment status parameters, with a data retention period of ≥5 years. The interaction interface supports integration with the mine safety monitoring system and production scheduling system (compatible with OPCUA and ModbusTCP protocols), and can push real-time risk reports and depressurization suggestions to the management platform. It can also receive mine production plans (such as mining progress) to dynamically adjust the monitoring frequency and depressurization range. Data transmission uses the AES-256 encryption algorithm to prevent information leakage and tampering.

[0045] This invention also includes a multi-source parameter dynamic weighted fusion unit. This unit adjusts the weights in real time based on the differences in the sensitivity of each parameter to rockburst, improving the accuracy of risk assessment. The weight coefficient is determined by calculating the parameter contribution; the higher the contribution, the greater the weight. The contribution of the stress parameter is calculated using the following formula: In the formula, Contribution to stress parameters (dimensionless). To monitor the maximum stress value (MPa) in the area. The average stress value (MPa) of the monitoring area. The stress change time interval (h) is given by k, and the distance attenuation coefficient (m) is given by k. -1The value is 0.1-0.3 (with a larger value near the fault), and d is the distance (m) from the measuring point to the tunnel face. The calculation method for the contribution of vibration parameters is similar, except that the input parameters are replaced with peak ground acceleration and dominant frequency. Finally, the weights of each parameter are normalized according to their contribution (the sum of the weights is 1), so that highly sensitive parameters have a higher proportion in risk assessment. For example, the weight of stress parameters in stress concentration areas can be increased to 0.4-0.5, which is better than the fusion effect of traditional fixed weights.

[0046] This invention also includes a dynamic risk level correction unit. This unit updates the classification standards by combining historical rockburst cases with real-time geological conditions, avoiding misjudgments caused by fixed thresholds. The correction is based on the parameter characteristics of rockbursts occurring in the same geological unit (similar lithology, burial depth, and structure) within the past three months, as well as the current roadway support strength (support resistance ≥ 200 kN / m) and mining disturbance intensity (daily advance ≤ 3 m). When fault activation (displacement rate ≥ 0.5 mm / d) or water inrush signs (water inflow ≥ 5 m³ / d) are detected, the correction is applied. 3 When the system is in operation ( / h), the threshold for judging Level III and Level IV risks is automatically lowered. For example, the stress concentration factor threshold is reduced from 2.0 to 1.8, making the system more adaptable to complex geological conditions. At the same time, the correction unit generates a report on the effectiveness of the classification standard every 72 hours. By comparing the actual risks with the warning results, if the false alarm rate is ≥15%, a manual review process is triggered to ensure that the classification standard continues to conform to the actual situation of the mine.

[0047] In this invention, the active pressure relief execution module also includes a pressure relief effect prediction unit. This unit evaluates the rationality of the scheme by calculating the stress release amount after pressure relief, ensuring sufficient pressure relief without inducing new risks. The stress release amount calculation formula is as follows: In the formula, The stress released after pressure relief (MPa) is given, α is the Poisson's ratio of the rock (0.2-0.35, 0.25 for sandstone), and E is the elastic modulus of the rock (GPa, 80-100GPa for granite). Volume of the pressure relief chamber (m³) 3 Drilling is calculated as a cylinder: πr 2 h, r is the radius, h is the depth). To influence the volume of the region (m) 3 (Take 5-10 times the volume of the pressure relief chamber). The cracking working pressure (MPa). The original ground stress (MPa) is used. When the calculated Δσ is less than 30% of the current stress concentration value, the number of pressure relief holes or the fracturing pressure is automatically increased until Δσ meets the requirements; if Δσ exceeds 50% of the original ground stress, the pressure relief intensity is reduced to prevent excessive pressure relief from causing roadway instability.

[0048] In this invention, the intelligent control central module also includes an emergency linkage unit. This unit automatically activates the mine emergency response mechanism during a Level IV risk warning, enabling multi-system coordinated risk avoidance. The linkage includes: sending evacuation commands to the underground personnel positioning system (positioning accuracy ≤3m), displaying the optimal evacuation route on the roadway LED screen (calculation time ≤5s); controlling the emergency stop of the underground transportation system (response time ≤2s), and sealing the entrance to the risk area (automatic door closer action time ≤1s); sending adjustment commands to the ventilation system to increase the airflow in the risk area (wind speed increased to 1.5-2m / s) to reduce the risk of gas accumulation; simultaneously, connecting the underground refuge chamber with the ground command center through the emergency communication system (call delay ≤0.5s), transmitting real-time images and personnel status, ensuring efficient and accurate emergency command, and improving the response speed by 3-5 times compared to traditional manual linkage.

[0049] In this invention, the multi-source parameter sensing module also includes a microseismic location unit. This unit determines the location of the microseismic source by analyzing the signal time difference of at least three acoustic wave sensors, providing accurate spatial information for risk assessment. The location accuracy is ≤5m. A particle swarm optimization-based location algorithm (50 iterations, 30 population size) is used. When calculating the microseismic source coordinates (x, y, z), a terrain correction coefficient is introduced to correct the wave velocity deviation caused by roadway undulations. The correction coefficient is calculated using the following formula: In the formula, β is the terrain correction coefficient (dimensionless, 1.0-1.2), and θ is the roadway slope angle (°, positive for uphill and negative for downhill). When the microseismic source is located above the goaf (z≤-5m), an additional correction value of 0.1 is added to reduce the positioning error under complex terrain to within 3m; the positioning results are superimposed on the mine's 3D model in real time (updated every 1 minute), intuitively showing the spatial distribution of high-risk areas and providing a basis for targeted stress relief.

[0050] In this invention, the data storage and interaction module also includes a digital twin mapping unit. This unit constructs a virtual digital twin of the mine roadway and system, achieving real-time synchronization between the physical entity and the virtual model (delay ≤100ms). The twin includes a geological model (accuracy 0.5m), an equipment model (including the three-dimensional parameters of sensors and pressure relief equipment), and an environmental model (stress, temperature, and vibration field distribution). By driving the virtual model update through real-time collected physical parameters, it can simulate the stress evolution process under different pressure relief schemes (simulation step size 1s) and predict the risk change trend within 30 minutes. At the same time, it supports the pre-simulation of pressure relief schemes in the virtual environment, calculating the stress release effect corresponding to different hole network parameters and fracturing pressures, providing visual decision support for optimizing pressure relief strategies, and improving efficiency by more than 10 times compared to traditional numerical simulation.

[0051] This invention also discloses a method for a rockburst risk classification, early warning, and active pressure relief system in deep mines, comprising the following steps:

[0052] Parameter acquisition and preprocessing steps: The multi-source parameter sensing module is activated, and distributed sensors acquire stress, vibration, sound wave, and temperature data at a frequency of 1kHz. The mobile inspection robot conducts supplementary monitoring of key areas every 2 hours. The raw data is transmitted to the data preprocessing unit via fiber optic bus and 5G network. The stress and vibration signals are processed using a wavelet threshold denoising algorithm, retaining the effective frequency band of 10-500Hz. The temperature and sound wave data are smoothed by Kalman filtering. Outliers are removed using the 3σ criterion (such as a sudden increase in stress exceeding 50MPa is considered abnormal). The marked abnormal parameters trigger sensor self-checks and secondary acquisition to ensure data validity. The preprocessed data is updated every 100ms.

[0053] Feature extraction and data fusion steps: The feature extraction unit extracts rockburst-sensitive features from the preprocessed data, including stress concentration factor (the ratio of maximum stress to average stress), vibration signal dominant frequency (obtained through Fourier transform), acoustic wave velocity change rate (the ratio of wave velocity difference per unit time to initial wave velocity), and temperature gradient (the ratio of temperature difference between adjacent measuring points to distance). The multi-source parameter dynamic weighted fusion unit calculates the contribution of each feature, determines the weights according to the contribution normalization (stress feature weight 0.3-0.5, vibration feature weight 0.2-0.3, acoustic wave and temperature feature weights 0.1-0.2 each), and obtains the fused feature value through weighted summation, thereby improving the data's recognizability.

[0054] Risk grading and early warning release steps: The risk grading decision unit compares the fusion characteristic value with the grading standard to determine the risk level (Level I to Level IV), where Level I corresponds to a fusion characteristic value < 0.3, Level II corresponds to 0.3 ≤ fusion characteristic value < 0.5, Level III corresponds to 0.5 ≤ fusion characteristic value < 0.7, and Level IV corresponds to a fusion characteristic value ≥ 0.7; The risk level dynamic correction unit adjusts the grading threshold based on historical cases and real-time geological conditions (such as faults and support strength). When fault activation is present, the Level IV threshold is reduced from 0.7 to 0.6; Early warning information is simultaneously released through tunnel audible and visual alarms, handheld terminals, and the dispatch center's large screen. Level IV early warnings are accompanied by emergency SMS messages sent to the mine's main responsible persons to ensure no information is missed.

[0055] Active pressure relief scheme decision-making and execution steps: The intelligent control center automatically matches the pressure relief scheme according to the risk level. Level I: no pressure relief is required. Level II: fixed-point hydraulic fracturing is initiated (drill hole diameter 50mm, depth 5m, fracturing pressure 30MPa). Level III: fixed-point + local area pressure relief is initiated (addition of carbon dioxide fracturing, energy 100kJ). Level IV: global coordinated pressure relief is initiated (deep hole blasting + carbon dioxide fracturing combination, borehole diameter 150mm, charge coefficient 0.7). The pressure relief effect prediction unit calculates the stress release amount. If the requirements are not met, the pressure relief parameters are adjusted (such as increasing the number of boreholes). The active pressure relief execution module controls the drilling rig and fracturing equipment according to the scheme, and provides real-time feedback of displacement and pressure data to achieve controllable pressure relief process.

[0056] Effect feedback and system optimization steps: 30 minutes after depressurization, the multi-source parameter sensing module re-collects data and calculates the risk level after depressurization. If it is still ≥ Level III, a second depressurization is triggered (the parameters are increased by 20% compared to the first time). The data storage and interaction module records the parameters, effects, and risk change curves of this depressurization. The digital twin mapping unit reviews the depressurization process. The system operation data is analyzed weekly to optimize feature weights, classification thresholds, and depressurization parameters, so that the system can continuously adapt to changes in mine geological conditions and improve long-term early warning and depressurization effects.

[0057] This invention also includes:

[0058] Microseismic source location and risk tracing steps: The microseismic location unit receives signals from at least three acoustic sensors, records the signal arrival time difference, and uses a particle swarm optimization algorithm to calculate the microseismic source coordinates (x, y, z). A terrain correction coefficient is introduced to correct the error caused by the roadway slope. The location results are superimposed on the mine's three-dimensional model to analyze the correlation between the distribution of microseismic sources and geological structures (faults, joints). If more than 70% of the microseismic sources are concentrated near a certain fault, the fault is determined to be a risk source. An enhanced monitoring plan is developed for the risk source (the sensor sampling frequency is increased to 2kHz, and the inspection interval is shortened to 1 hour). The pressure relief holes are densified within 5m on both sides of the fault (the hole spacing is reduced from 3m to 2m) to reduce the risk of rockburst from the source. The efficiency of pressure relief is improved by more than 40% compared with the traditional blind pressure relief without location.

[0059] Emergency Response and Collaborative Evacuation Procedures: When a Level IV risk warning is issued, the emergency response unit immediately activates the emergency response, sends evacuation instructions to the underground personnel positioning system, plans the optimal evacuation route based on the real-time location of personnel (avoiding high-stress areas and areas with dense micro-seismic activity), and dynamically updates route guidance on the LED screens in the tunnels; controls the emergency stop of the underground transportation system, and seals the entrance to the risk area (automatic door closers are closed) to prevent personnel from accidentally entering; adjusts the ventilation system to increase the airflow in the risk area and reduce the methane concentration to below 0.5%; connects the refuge chamber with the ground command center through the emergency communication system, transmits on-site images and personnel status in real time, and directs personnel to evacuate in an orderly manner; once the risk level drops to below Level II and safety is confirmed manually, the emergency status can be lifted and normal operations can resume. This procedure shortens the evacuation time by more than 50% compared to traditional emergency procedures, significantly improving the safety of underground personnel.

[0060] Specific implementation methods of the deep mine rockburst risk classification, early warning and active pressure relief system and method:

[0061] Example 1: Rockburst prevention and control in a copper mine roadway at a depth of 1200m (metal mine scenario)

[0062] This embodiment focuses on a 1200m horizontal transport roadway (cross-section size 4m×3.5m, lithology granite, including F1 fault fracture zone) in a large copper mine. It applies a deep mine rockburst risk classification early warning and active pressure relief system, and focuses on verifying the effects of multi-source parameter fusion, microseismic location and full-area coordinated pressure relief. It covers all the technical contents of claims 1-10 and solves the problems of low accuracy and blind pressure relief in traditional monitoring and early warning systems.

[0063] I. Hardware Deployment and Parameter Configuration Multi-source parameter sensing module: Distributed sensing units are deployed every 5m along the tunnel roof and both sides, totaling 20 sets. The roof sensor set includes a fiber optic stress sensor (model FBG-S100, measurement range 0-600MPa, accuracy ±1MPa, sampling frequency 1kHz) and a three-component vibration sensor (model VS-300, range ±500g, frequency response 0.1-10kHz). The side sensor set additionally includes an acoustic sensor (model AC-200, receiving frequency 20Hz-2MHz, sensitivity 92dB) and an infrared temperature sensor (model IR-50, measurement range -20-80℃, accuracy ±0.5℃). The sensors are connected to three data acquisition substations (deployed at the tunnel entrance, middle section, and exit) via high-pressure resistant waterproof armored cables (model KC-YJV22, temperature resistance -40-120℃, pressure resistance 10MPa). The substations upload data via industrial Ethernet (transmission rate 1000Mbps). The mobile monitoring unit is mounted on an explosion-proof inspection robot (model KB-100, 10-hour battery life, 0.8m / s walking speed), equipped with a portable borehole stress gauge (model ZKY-2, 0-5m measurement depth, 0.1MPa resolution) and a 5-megapixel industrial camera (35m night vision distance). The preset inspection route covers a 100m range on both sides of the F1 fault, and the robot performs an inspection every 2 hours. The data is uploaded via a 5G explosion-proof base station (model 5G-EX100, 150Mbps transmission rate).

[0064] Risk grading, early warning, and intelligent control module: The data preprocessing unit uses a wavelet threshold denoising algorithm (threshold coefficient 0.65) to denoise the 450MPa stress signal (including drilling rig vibration interference), retaining the effective frequency band of 10-500Hz; the Kalman filter parameters are set to Q=0.02 and R=0.1 to smooth the acoustic wave velocity data (original wave velocity 5200m / s, smoothed to 5180-5220m / s). The feature extraction unit calculates the stress concentration factor = 450MPa (maximum stress) / 300MPa (average stress) = 1.5, the vibration dominant frequency 180Hz, the acoustic wave velocity change rate = (5180-5200) / 5200 / 2h = -0.0019 / h, and the temperature gradient = (25.2-24.8)℃ / 5m = 0.08℃ / m. The multi-source parameter dynamic weighted fusion unit calculates the stress contribution according to the formula in claim 2: The vibration contribution (peak acceleration 0.8g, dominant frequency 180Hz) was calculated to be 0.021. The final weights were normalized to stress 0.44, vibration 0.28, sound wave 0.16, and temperature 0.12. The fused feature value = 1.5×0.44+(180 / 500)×0.28+|-0.0019|×0.16+0.08×0.12≈0.66+0.1008+0.0003+0.0096≈0.77, which was determined to be a Level IV risk.

[0065] Active pressure relief and emergency linkage unit: The active pressure relief execution module initiates full-area coordinated pressure relief. The fixed-point pressure relief unit uses an automatic drill rod changing rig (model ZD-200, borehole diameter 80mm, depth 12m) to construct pressure relief holes (hole spacing 2.5m, row spacing 1.5m) on the roadway sidewall. The hydraulic fracturing equipment (model ZY-60, working pressure 40MPa) operates at a single-hole fracturing time of 15min. The regional pressure relief unit uses a carbon dioxide phase change fracturing system (model CO2-300, filling pressure 25MPa, fracturing energy 200kJ) to densely deploy fracturing tubes (spaced 3m apart) 5m apart on both sides of the F1 fault. The programmable detonator (model BQ-100, delay accuracy ±1ms) is set with 3 micro-delays (interval 50ms). The pressure relief effect prediction unit calculates the stress release amount according to the formula in claim 4: The current stress concentration value is 150MPa. 30% × 150 = 45MPa, Δσ = 1.71MPa is insufficient. The fracturing pressure is increased to 50MPa. Δσ is recalculated as 0.25 × 9 × ln(1 + 50 / 35) = 2.25 × ln(2.428) ≈ 2.25 × 0.883 ≈ 1.99MPa, still insufficient. Finally, the number of pressure relief holes is increased by 50%, raising Δσ to 48MPa, meeting the requirements. After the emergency linkage unit is activated, the underground personnel positioning system sends an evacuation command, the LED screen displays the evacuation route, the transportation system stops within 3 seconds, the automatic door closer seals the risk area within 0.8 seconds, and the ventilation system increases the air volume from 1200m³ / h. 3 / min increased to 1800m 3 / min, the emergency communication system transmits on-site images.

[0066] II. Performance Verification Form

[0067] Performance indicators Traditional rockburst prevention and control solutions System solution of the present invention Improvement effect Rockburst early warning accuracy rate (%) 55-60 92-95 Significant improvement Reduction rate of stress concentration factor after pressure relief (%) 20-25 45-50 Significantly improved Level IV risk emergency response time (s) 60-90 8-12 Significantly shorten Microseismic source location error (m) 10-15 2-3 Significantly reduced Normal operation resumption time of the tunnel (h) 48-72 12-15 Significantly shorten

[0068] The data in the table is based on statistics from three months of continuous operation (a total of 8 Level IV risks and 12 Level III risks were monitored). Traditional methods, relying on single stress monitoring, have an early warning accuracy rate of less than 60%, with 4 Level IV risks missed, leading to 2 small-scale rockbursts. This invention, through multi-source fusion and dynamic correction, achieves an early warning accuracy rate exceeding 92%, providing advance warnings for all 8 Level IV risks. After decompression, the stress concentration coefficient of traditional methods decreases by only 25%, while this invention reduces it by 50%, and the microseismic positioning error is reduced from 15m to 3m, allowing for precise decompression of fault areas. Emergency response time is reduced from 90s to 12s, allowing for more timely evacuation of underground personnel and a 75% reduction in roadway recovery time, perfectly meeting the high-efficiency production needs of metal mines. It also avoids equipment damage and personnel casualties caused by rockbursts, resulting in significant safety and economic benefits.

[0069] Example 2: Rockburst prevention and control in a fully mechanized mining face at a depth of 1000m (coal mine scenario)

[0070] This embodiment focuses on the 3102 fully mechanized longwall face at a 1000m level in a coal mine (strike length 1200m, dip length 200m, coal seam thickness 3.2m, gas content 8m³ / t). It applies this system to control the risk of roof rockburst during mining, and focuses on verifying the effects of gas-rockburst synergistic monitoring, carbon dioxide fracturing and pressure relief, and digital twin simulation. It covers the technical content of claims 1-10 and solves the problem of coal mine rockburst and gas coupled disaster prevention and control.

[0071] I. Hardware Deployment and Parameter Configuration Multi-source parameter sensing module: Distributed sensing units are deployed every 10m along the working face support, totaling 20 sets. Each set includes a fiber optic stress sensor (model FBG-M200, measuring roof pressure 0-800MPa, accuracy ±1MPa), a three-component vibration sensor (model VS-500, resistant to gas corrosion), an acoustic sensor (model AC-300, moisture resistant), an infrared temperature sensor, and a gas sensor (model WG-EX, measuring range 0-10%CH4, accuracy ±0.1%). The sensors are connected to two underground explosion-proof substations via coal mine-specific flame-retardant cables (model MYJV22, temperature resistant -30-80℃), and data is uploaded via fiber optic bus (transmission rate 500Mbps). The mobile monitoring unit is mounted on a tracked explosion-proof inspection robot (model MC-200, with an 8-hour battery life and adaptability to coal seam undulations). It is equipped with a portable borehole stress gauge (measuring coal seam stress 0-50MPa) and a high-definition camera (dustproof and waterproof). It inspects the working face once every 1.5 hours, and the data is uploaded via a 5G explosion-proof base station.

[0072] Risk grading and digital twin module: The data preprocessing unit uses wavelet threshold denoising (threshold coefficient 0.65) to process the roof stress signal (the original signal contains coal mining machine vibration interference), and Kalman filtering to smooth the gas concentration data (original concentration 0.6-0.8%, smoothed 0.65-0.75%). The feature extraction unit calculates the stress concentration factor = 550MPa / 360MPa≈1.53, the dominant vibration frequency 160Hz, the rate of change of sound wave velocity = (4800-4900) / 4900 / 1.5h≈-0.0136 / h, the temperature gradient = (23.5-22.9)℃ / 10m=0.06℃ / m, and the gas concentration 0.7%. The risk level dynamic correction unit, due to the daily advance of the working face being 2.5m (≤3m) and the support resistance being 220kN / m (≥200kN / m), but with a risk of excessive gas (0.7% close to the 1% limit), lowered the stress concentration factor threshold for Level IV risk from 2.0 to 1.8. The current value of 1.53 is classified as Level III risk. The microseismic positioning unit calculates the terrain correction coefficient according to the formula in claim 6: working face slope θ = 12° (uphill). The microseismic source was located in the middle of the working face (x=600m, y=100m, z=-8m), with a positioning error of 2.5m. After being superimposed onto the three-dimensional model of the mine (accuracy 0.5m), it was determined that the microseismic source coincided with the coal seam joint development zone.

[0073] Active pressure relief and digital twin simulation: The active pressure relief execution module starts fixed-point and local area pressure relief. The fixed-point pressure relief unit uses a drilling rig (model ZM-150, drilling diameter 70mm, depth 8m) to construct pressure relief holes (hole spacing 3m) between supports, with a hydraulic fracturing pressure of 35MPa; the area pressure relief unit uses a carbon dioxide phase change fracturing system (model CO2-200, coal mine explosion-proof type, filling pressure 22MPa, fracturing energy 150kJ) to avoid gas explosion caused by blasting. Before decompression, a digital twin mapping unit was used to construct a virtual model of the working face (including 200 sets of support, coal seam, and joint models) to simulate the stress evolution under different fracturing pressures (30MPa, 35MPa, and 40MPa): at 30MPa, the stress release was 32MPa (less than 30% of the current concentration value of 190MPa); at 35MPa, the release was 58MPa (satisfactory); and at 40MPa, the release was 85MPa (50% above the original ground stress). Ultimately, 35MPa was selected. Thirty minutes after decompression, sensor data showed that the stress concentration factor decreased to 1.1, the dominant vibration frequency decreased to 80Hz, and the risk level decreased to Level II.

[0074] II. Performance Verification Form

[0075] Performance indicators Traditional coal mine rockburst prevention and control solutions System solution of the present invention Improvement effect Accuracy rate of rockburst-gas coordinated early warning (%) 50-55 90-93 Significant improvement <![CDATA[Gas concentration control effect after pressure relief (%CH4)]]> 0.8-1.2 0.5-0.7 Significantly reduced Efficiency optimization (times) of digital twin decompression solution 1 (No rehearsal) 10-12 Significantly improved Resumption of operations after depressurization for Level III risks (hours) 24-36 6-8 Significantly shorten Average number of rockbursts affecting the working face per month (times) 4-5 0-1 Significantly reduced

[0076] The data in the table is based on statistics from a 2-month mining cycle (advance distance 150m). Traditional methods lacked coordinated monitoring of gas and rockbursts, resulting in three instances where rockburst pressure relief caused gas concentrations to rise to 1.2%, triggering production shutdowns. This invention, through coordinated gas-rockburst early warning and carbon dioxide-induced fracturing, stabilized gas concentrations below 0.7%, preventing production shutdowns. Digital twin simulation improved the efficiency of pressure relief scheme optimization by 10 times, avoiding support deformation caused by three instances of blind pressure relief. Traditional methods experienced an average of 5 rockburst impacts per month, while this invention only experienced one small-scale micro-tremor, reducing operation recovery time from 36 hours to 8 hours, meeting the high-yield and high-efficiency requirements of coal mines. Simultaneously, it eliminated the risk of coupled rockburst-gas disasters, ensuring the safety of underground personnel and complying with coal mine safety regulations.

[0077] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A rockburst risk classification, early warning, and active pressure relief system for deep mines, comprising: Parameter sensing module: Includes distributed sensing units and mobile monitoring units. The distributed sensing units are deployed every 5m along the tunnel outline. Each set contains a fiber optic stress sensor, a three-component vibration sensor, an acoustic sensor, and an infrared temperature sensor. The sensors are connected to the data acquisition substation via waterproof armored cables. The mobile monitoring unit is mounted on an explosion-proof inspection robot, equipped with a portable borehole stress gauge and an industrial camera. It conducts supplementary monitoring of the area along a preset track, and the data is uploaded to the control center via a 5G explosion-proof base station. The risk classification and early warning module includes a data preprocessing unit, a feature extraction unit, and a classification decision unit. The data preprocessing unit uses a wavelet threshold denoising algorithm to remove interference noise from mining equipment in stress and vibration signals, smooths temperature and acoustic data using Kalman filtering, and uses the 3σ criterion to determine outliers, marking parameters that exceed the range and triggering secondary acquisition. The feature extraction unit extracts rockburst-sensitive features from the preprocessed data, including stress concentration factor, vibration signal frequency, acoustic wave velocity change rate, and temperature gradient. The classification decision unit classifies rockburst risk into four levels: Level I, Level II, Level III, and Level IV. Early warning information is simultaneously released through roadway audible and visual alarms, management personnel's handheld terminals, and the mine dispatch center's large screen. Active pressure relief execution module: includes a fixed-point pressure relief unit and a regional pressure relief unit: the fixed-point pressure relief unit deploys hydraulic fracturing equipment for stress concentration areas and is equipped with an automatic drill rod changing drill to construct pressure relief holes according to preset hole pattern parameters; the regional pressure relief unit is suitable for stress areas and includes a carbon dioxide phase change fracturing system and a deep hole pre-fracturing blasting device, which achieves micro-differential pressure relief through a programmed detonator; all pressure relief equipment is equipped with displacement sensors and pressure sensors to provide real-time feedback on the execution status; Control center module: adopts an industrial computer + PLC redundant architecture, equipped with rockburst risk assessment algorithm and pressure relief decision model: the risk assessment algorithm integrates parameter features and calculates real-time risk value through BP neural network; The depressurization decision model automatically matches depressurization schemes based on risk levels, and parameters can be manually adjusted through a human-computer interaction interface. Data storage and interaction module: Includes local server and cloud platform, uses distributed database to store raw sensor data, risk assessment results, pressure relief execution records and equipment status parameters; the interaction interface supports connection with mine safety monitoring system and production scheduling system, pushes real-time risk reports and pressure relief suggestions to management platform, and receives mine production plan to dynamically adjust monitoring frequency and pressure relief range. Data transmission uses the AES-256 encryption algorithm.

2. The deep mine rockburst risk classification, early warning, and active pressure relief system according to claim 1, characterized in that, It also includes a multi-source parameter dynamic weighted fusion unit, which adjusts the weights in real time based on the differences in the sensitivity of each parameter to rockburst. The weighting coefficients are determined by calculating the contribution of the parameters; the higher the contribution, the greater the weight. The contribution of the stress parameters is calculated using the following formula: In the formula, Contribution to stress parameters To monitor the maximum stress value in the area, To monitor the average stress value of the area, The stress change time interval is denoted by k, the distance attenuation coefficient is denoted by d, and the distance from the measuring point to the working face is denoted by d. The calculation method for the contribution of vibration parameters is similar, except that the input parameters are replaced with peak ground acceleration and dominant frequency. Finally, the weights of each parameter are normalized according to their contribution.

3. The deep mine rockburst risk classification, early warning, and active pressure relief system according to claim 1, characterized in that, It also includes a risk level dynamic correction unit, which updates the classification standards by combining historical rockburst cases with real-time geological conditions. The correction is based on the parameter characteristics of rockbursts that occurred in the same geological unit within the past 3 months, as well as the current roadway support strength and mining disturbance intensity. When fault activation or water inrush signs are detected, the judgment threshold of level III and level IV risks is automatically lowered. At the same time, the correction unit generates a classification standard validity report every 72 hours, comparing the actual risks with the early warning results.

4. The deep mine rockburst risk classification, early warning, and active pressure relief system according to claim 1, characterized in that, The active pressure relief execution module also includes a pressure relief effect prediction unit, which evaluates the rationality of the scheme by calculating the stress release amount after pressure relief; the stress release amount calculation formula is as follows: In the formula, The stress is released after pressure relief, α is the Poisson's ratio of the rock, and E is the elastic modulus of the rock. The volume of the pressure relief chamber. To affect the volume of the region, To cause cracking working pressure, This represents the original geostress.

5. The deep mine rockburst risk classification, early warning, and active pressure relief system according to claim 1, characterized in that, The control center module also includes an emergency linkage unit, which automatically activates the mine emergency response mechanism when a Level IV risk warning is issued. The linkage includes: sending evacuation instructions to the underground personnel positioning system and displaying the evacuation route on the LED screen in the roadway; controlling the emergency stop of the underground transportation system and sealing the entrance to the risk area; sending adjustment instructions to the ventilation system to increase the air volume in the risk area and reduce the risk of gas accumulation; and simultaneously connecting the underground refuge chamber with the ground command center through the emergency communication system to transmit on-site images and personnel status in real time.

6. The deep mine rockburst risk classification, early warning, and active pressure relief system according to claim 1, characterized in that, The parameter sensing module also includes a microseismic positioning unit, which determines the location of the microseismic source by analyzing the signal time difference of three acoustic sensors. A particle swarm optimization-based positioning algorithm is used to calculate the microseismic source coordinates (x, y, z). A terrain correction coefficient is introduced to correct the wave velocity deviation caused by roadway undulations. The correction coefficient is calculated using the following formula: In the formula, β is the terrain correction coefficient, and θ is the tunnel slope angle.

7. The deep mine rockburst risk classification, early warning, and active pressure relief system according to claim 1, characterized in that, The data storage and interaction module also includes a digital twin mapping unit, which constructs a virtual digital twin of the mine roadway and system to achieve real-time synchronization between the physical entity and the virtual model. The twin includes a geological model, an equipment model, and an environmental model. The virtual model is updated by real-time collected physical parameters to simulate the stress evolution process under different pressure relief schemes and predict the risk change trend within 30 minutes. At the same time, it supports the pre-simulation of pressure relief schemes in the virtual environment, calculates the stress release effect corresponding to different hole network parameters and fracturing pressures, and provides visual decision support for optimizing pressure relief strategies.

8. A method for applying the deep mine rockburst risk classification, early warning, and active pressure relief system according to any one of claims 1-7, characterized in that, include: Parameter acquisition and preprocessing steps: The parameter sensing module is activated, and distributed sensors acquire stress, vibration, sound wave, and temperature data at a frequency of 1kHz. The mobile inspection robot conducts supplementary monitoring of key areas every 2 hours. The raw data is transmitted to the data preprocessing unit via fiber optic bus and 5G network. The stress and vibration signals are processed using wavelet threshold denoising algorithm, and the temperature and sound wave data are smoothed by Kalman filtering. Outliers are removed using the 3σ criterion. The marked abnormal parameters trigger sensor self-check and secondary acquisition. Feature extraction and data fusion steps: The feature extraction unit extracts rockburst sensitive features from the preprocessed data, including stress concentration factor, vibration signal dominant frequency, sound wave velocity change rate and temperature gradient; the parameter dynamic weighted fusion unit calculates the contribution of each feature, determines the weight according to the contribution, and obtains the fused feature value by weighted summation; Risk grading and early warning release steps: The risk grading decision unit compares the fusion feature value with the grading standard to determine the risk level, where Level I corresponds to a fusion feature value < 0.3, Level II corresponds to a fusion feature value ≤ 0.3 < 0.5, Level III corresponds to a fusion feature value ≤ 0.5 < 0.7, and Level IV corresponds to a fusion feature value ≥ 0.

7. The risk level dynamic correction unit adjusts the classification threshold based on historical cases and real-time geological conditions. When fault activation occurs, the threshold for level IV drops from 0.7 to 0.

6. Early warning information is simultaneously released through roadway audible and visual alarms, handheld terminals, and the dispatch center's large screen. Level IV early warnings are accompanied by emergency SMS messages sent to the mine manager. Active pressure relief scheme decision and execution steps: The control center automatically matches the pressure relief scheme according to the risk level. Level I does not require pressure relief, Level II initiates fixed-point hydraulic fracturing, Level III initiates fixed-point + local area pressure relief, and Level IV initiates full-area coordinated pressure relief. The pressure relief effect prediction unit calculates the stress release amount. If the requirements are not met, the pressure relief parameters are adjusted. The active pressure relief execution module controls the drilling rig and fracturing equipment according to the scheme and provides real-time feedback of displacement and pressure data. Effect feedback and system optimization steps: 30 minutes after depressurization, the parameter sensing module re-collects data and calculates the risk level after depressurization. If it is still ≥ Level III, a second depressurization is triggered. The data storage and interaction module records the parameters, effects, and risk change curves of this depressurization, and the digital twin mapping unit reviews the depressurization process; the system operation data is analyzed weekly to optimize feature weights, grading thresholds, and depressurization parameters, so that the system can continuously adapt to changes in mine geological conditions.

9. The method for risk classification, early warning, and active pressure relief of rockburst in deep mines according to claim 8, characterized in that, Also includes: Steps for microseismic source location and risk tracing: The microseismic location unit receives signals from the acoustic sensor, records the signal arrival time difference, and uses a particle swarm optimization algorithm to calculate the microseismic source coordinates (x, y, z). A terrain correction coefficient is introduced to correct the error caused by the roadway slope. The location results are superimposed on the three-dimensional model of the mine to analyze the correlation between the distribution of microseismic sources and the geological structure. If 70% of the microseismic sources are concentrated near the fault, the fault is identified as a risk source. An enhanced monitoring plan is formulated for the risk source, and pressure relief holes are densified within a 5m range on both sides of the fault.

10. The method for risk classification, early warning, and active pressure relief of rockburst in deep mines according to claim 8, characterized in that, Also includes: Emergency Response and Collaborative Risk Avoidance Procedures: When a Level IV risk warning is issued, the emergency response unit immediately activates the emergency response, sends evacuation instructions to the underground personnel positioning system, plans the optimal evacuation route based on the real-time location of personnel, and dynamically updates route guidance on the LED screen in the tunnel; controls the emergency shutdown of the underground transportation system and seals off the entrance to the risk area; adjusts the ventilation system to increase the airflow in the risk area and reduce the gas concentration; connects the refuge chamber with the ground command center through the emergency communication system to transmit on-site images and personnel status in real time; once the risk level drops to Level II, the emergency status is lifted after manual confirmation of safety.

Images