Intelligent mine pressure monitoring and early warning device based on AI algorithm

Abstract

The invention discloses an intelligent mine pressure monitoring and early warning device based on an AI algorithm, and belongs to the technical field of mine pressure monitoring. A storage groove is formed in the top of a bearing plate, piezoelectric film sensors which are uniformly arranged are detachably mounted in the storage groove, an annular shape memory alloy honeycomb plate is mounted at the top of the bearing plate, and the piezoelectric film sensors are detachably mounted in the storage groove; the ends, close to each other, of every two sets of anchor rods are sleeved with stress meters, MEMS multi-mode sensors which are evenly arranged are embedded in the outer surfaces of one set of anchor rods, and the MEMS multi-mode sensors are used for monitoring the influence of external loads on prestress. The pressure is accurately sensed through the piezoelectric film sensor, the axial force of the anchor rod is monitored through the stress meter, the mine pressure condition can be accurately mastered, meanwhile, the influence of external loads on the prestress is detected through the MEMS multi-mode sensor, and therefore the probability that accurate mine pressure data cannot be obtained due to the fact that the prestress is reduced is reduced, the mine pressure development trend is accurately judged, and the mine pressure monitoring accuracy is improved. And the accuracy and reliability of mine pressure early warning and the safety of mining are improved.

Description

Technical Field

[0001] This invention belongs to the field of mine pressure monitoring technology, specifically relating to an intelligent mine pressure monitoring and early warning device based on AI algorithms. Background Technology

[0002] In mining operations, mine pressure monitoring is a core component for ensuring safe production. Traditional mine pressure monitoring devices often rely on manual inspections or single sensors to collect data, resulting in poor real-time performance, delayed early warnings, and insufficient data analysis capabilities. Furthermore, existing anchor stress gauges typically only provide static data and cannot dynamically sense the impact of external load changes on prestress. This causes some anchor stress gauges to gradually drop to zero after the prestress is set, making them inaccurate in monitoring mine pressure data. This can lead to frequent false alarms or failure to sound an alarm when the mine pressure hazard increases, thereby increasing the probability of accidents such as roof collapse and roadway collapse, and impacting safe mine production. Summary of the Invention

[0003] The purpose of this invention is to provide an intelligent monitoring and early warning device for mine pressure based on AI algorithms, so as to solve the problems mentioned in the background art.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a mine pressure intelligent monitoring and early warning device based on AI algorithm, comprising a pressure plate and anchor bolts. A slot is provided through the top of the pressure plate, and the anchor bolts are located inside the slot. Multiple sets of anchor bolts are provided, each set being a hollow structure. A storage slot is provided on the top of the pressure plate, and piezoelectric thin-film sensors are detachably installed inside the storage slot. A uniformly arranged set of opening slots is provided on the outer surface of the pressure plate, and each set of opening slots communicates with the storage slot. A circular shape memory alloy honeycomb plate is installed on the top of the pressure plate. A stress gauge is sleeved at the end of each pair of anchor bolts that are close to each other. A uniformly arranged MEMS multimodal sensor is embedded in the outer surface of one set of anchor bolts, and the MEMS multimodal sensor is used to monitor the influence of external loads on prestress.

[0005] In the above implementation process, the storage slot is used to install piezoelectric thin film sensors. Signal lines are led out through the open slots. The sensors are distributed in a matrix, which can accurately sense the contact pressure distribution on the surface of the pressure plate. The shape memory alloy honeycomb plate adopts a nickel-titanium alloy honeycomb structure, which can adapt to the curved deformation of the tunnel roof. It maintains a tight fit with the surrounding rock through elastic deformation, improving the pressure transmission efficiency. At the same time, it absorbs the deformation energy of the surrounding rock through the elastic deformation of the honeycomb grid, maintaining continuous contact between the sensor and the surrounding rock, and avoiding the situation where the sensor falls off due to rock strata displacement. The stress gauge is installed at the joint of adjacent anchor bolts, and monitors in real time. The system monitors changes in anchor bolt axial force, with the stress gauge cable passing through the hollow part of the anchor bolt. A piezoelectric thin-film sensor monitors the roof contact pressure in real time, while the stress gauge measures the anchor bolt axial force to generate static stress field data. A MEMS multimodal sensor captures external loads, such as vibration acceleration and strain abrupt changes caused by mining vibrations or rock collapses, generating dynamic response feature vectors. Through multidimensional data correlation analysis, it distinguishes between mining vibrations and rock collapse vibrations, reducing false alarm rates. It also reduces the occurrence of gradual prestress decrease due to external loads or weak connections, thus improving the accuracy of mine pressure monitoring.

[0006] In one specific implementation, the top of the pressure plate is provided with equidistant limiting grooves, and each set of limiting grooves is connected to the groove opening. The limiting grooves are located inside the storage groove and have an L-shaped structure. The bottom of the pressure plate is provided with equidistant slots, and the slots are connected to the limiting grooves. The width of the slots is half that of the limiting grooves. Each set of limiting grooves contains a limiting block. Multiple sets of limiting blocks have collars installed on their surfaces that are close to each other. The outer surface of the collars abuts against the inner surface of the groove opening. The collars are fitted onto the outer surface of another set of anchor rods and welded to them.

[0007] In the above implementation process, during assembly, the anchor rod is first passed through the slot, the limiting block is aligned with the slot, the anchor rod is pushed upward, and then the anchor rod is rotated clockwise, causing the limiting block to abut against the limiting slot. Then, the anchor rod with stress gauge can be fixed in the pressure plate by the engagement of the collar with the slot and the support of the limiting block by the limiting slot.

[0008] In one specific implementation, a sleeve is fitted on the outer surface of the anchor rod, and one end of the sleeve is connected to the top of the pressure plate. The outer surface of the sleeve abuts against the inner surface of the shape memory alloy honeycomb plate. An elastic rubber is embedded in the inner surface of the sleeve. The sleeve is made of soft plastic. During grouting, the grout solidifies and squeezes the sleeve, causing the elastic rubber to come into close contact with the anchor rod and the stress gauge.

[0009] In the above process, the sleeve is made of polypropylene soft plastic with elastic rubber embedded inside. After grouting, the grout squeezes the sleeve, making the elastic rubber close to the anchor rod and stress gauge, realizing gapless stress transmission and eliminating measurement errors caused by loose installation.

[0010] In one specific implementation, the bottom of the collar is provided with multiple sets of sand injection holes, which are used to fill the gap between the sleeve and the anchor rod with fine sand. The inside of the sand injection holes is fitted with a plug, and the bottom of the plug is provided with a pull rope.

[0011] In the above implementation process, fine sand is filled into the gap between the sleeve and the anchor rod through the sand injection hole, thereby eliminating the tiny gap between the sleeve and the anchor rod, improving the vibration resistance performance. Moreover, the compact structural design eliminates the need to fill a large amount of fine sand, which can reduce the weight of the monitoring device to a certain extent. At the same time, the setting of the plug can block the sand injection hole, and the setting of the pull rope makes it easy to remove the plug from the sand injection hole.

[0012] In one specific implementation, one set of anchor rods has a stop block threaded to the end away from the stress gauge, and the stop block has a T-shaped structure. The outer surface of the stop block fits against the inner surface of the other end of the sleeve. The other set of anchor rods has a locking nut threaded to the end away from the stress gauge, and the locking nut is located below the pressure plate.

[0013] In the above implementation process, the stop block and the anchor rod are threaded together, and the stop block and the smaller diameter end are fitted into the sleeve to improve the stability of the anchor rod in the sleeve. The locking nut is used to further strengthen the firmness between the anchor rod and the pressure plate.

[0014] In one specific implementation, the outer surface of the sleeve is provided with uniformly arranged slurry outlet holes, and the slurry outlet holes are in the shape of an outwardly expanding trumpet.

[0015] In the above process, the grout outlet is funnel-shaped with an outward expansion angle of more than 60° to ensure that the grout spreads evenly during grouting and to avoid air bubbles from being trapped.

[0016] In one specific implementation, the top of the pressure plate is provided with equidistantly arranged grouting holes, and the grouting holes are distributed at intervals with the limiting groove.

[0017] In the above process, the grouting hole is used to fix the anchor rod and sleeve in the pre-drilled hole and then perform grouting operation to improve the firmness of the anchor rod in the hole.

[0018] In one specific implementation, a grouting groove is provided at one end of the sleeve, and the grouting groove is located outside the grouting hole.

[0019] In the above process, the grouting groove is a ring-shaped groove structure, which guides the grout to fill the gap between the sleeve and the surrounding rock first. After solidification, it forms an elastic support layer, reducing the impact of rock creep on the prestress of the anchor bolt.

[0020] In one specific implementation, the MEMS multimodal sensor also supports synchronous acquisition of multidimensional data.

[0021] In the above implementation process, MEMS multimodal sensors support multidimensional data synchronous acquisition, thereby providing a more accurate and richer data foundation for subsequent data analysis. At the same time, based on this synchronous data, AI algorithms can be used to mine the potential correlations and change patterns between various physical quantities, and more accurately judge the development trend of mine pressure. If the data of a certain physical quantity is abnormal, it can be combined with other synchronous data to quickly determine whether it is caused by sensor failure or actual mine pressure change, which facilitates timely maintenance and troubleshooting, and can improve the accuracy and reliability of mine pressure early warning.

[0022] In one specific implementation, the data from the stress gauge and MEMS multimodal sensor are both transmitted to a deployed edge server, where time-series analysis is performed using an LSTM model to generate a mine pressure risk level and feed it back to the terminal device.

[0023] In the above implementation process, stress gauges and MEMS multimodal sensors transmit the collected data to an edge server, and an LSTM model is constructed using AI algorithms. Then, time-series analysis can be performed on the collected data to uncover the patterns of change in the data over time, such as the growth trend, periodic changes, and abrupt changes in mine pressure. Based on the results of the time-series analysis, the LSTM model can quantify the mine pressure status into different risk levels, such as low risk, medium risk, and high risk, according to preset rules and trained parameters. The generated mine pressure risk level is then fed back to terminal devices, such as monitoring screens in the mine or handheld terminals for management personnel. Relevant personnel can obtain mine pressure risk information in real time. If a high-risk state is detected, emergency plans can be quickly activated, such as strengthening support and suspending some mining operations, to ensure safe production in the mine.

[0024] Compared with the prior art, the beneficial effects of the present invention are:

[0025] 1. This invention uses a piezoelectric thin-film sensor to accurately sense pressure and a stress gauge to monitor the axial force of anchor bolts. This allows for accurate assessment of mine pressure conditions, precise early warning of abnormal mine pressure, and reduction of the risk of accidents such as collapses and roof falls. It also avoids ecological damage and secondary resource exploitation caused by accident repairs, and prevents over-support due to inaccurate monitoring. This reduces the unnecessary use of support materials such as anchor bolts and lowers resource consumption. Furthermore, the invention uses a MEMS multimodal sensor to detect the impact of external loads on prestress, improving the accuracy of mine pressure monitoring. This reduces the probability of failing to obtain accurate mine pressure data due to reduced prestress, thereby improving the safety of coal mining.

[0026] 2. This invention, through the combination of structures such as slots, limiting slots, grooves, limiting blocks, and collars, can fix anchor bolts and stress gauges. The assembly method is relatively simple, requiring no complex tools or large amounts of auxiliary materials. Compared to complex installation processes, it reduces manpower and material consumption during installation, lowers energy consumption caused by prolonged installation work, and improves construction efficiency. Furthermore, it allows for the recovery of anchor bolts and stress gauges according to actual usage needs, reducing material loss and usage costs, improving material utilization, and indirectly reducing energy consumption and pollution during resource extraction and production.

[0027] 3. This invention utilizes elastic rubber to ensure that the device adheres closely to the anchor bolt and stress gauge under the pressure of the grout, achieving seamless stress transmission. This guarantees the accuracy of mine pressure monitoring and avoids resource waste such as over-support due to inaccurate data. Furthermore, the sleeve design allows grout to be injected into the sleeve through the grouting hole after the monitoring and early warning device is fixed in place. The grout then flows into the borehole through the grout outlet, improving the robustness of the monitoring and early warning device and reducing the use of support materials, thereby lowering energy consumption during resource extraction, production, and transportation. Attached Figure Description

[0028] Figure 1 This is a three-dimensional structural diagram of the present invention;

[0029] Figure 2 This is a cross-sectional view of the present invention;

[0030] Figure 3 This is a schematic diagram of the assembly structure of the limiting block and the limiting groove of the present invention;

[0031] Figure 4 This is a schematic diagram of the assembly structure of the anchor bolt and bearing plate of the present invention;

[0032] Figure 5 This is a schematic diagram of the assembly structure of the collar and the slot of the present invention;

[0033] Figure 6 This is a schematic diagram of the planar assembly structure of the pressure plate and the piezoelectric thin film sensor of the present invention;

[0034] Figure 7 This is a top view of the pressure plate of the present invention;

[0035] Figure 8 This is a schematic diagram of the assembly structure of the sleeve of the present invention.

[0036] In the diagram: 1. Pressure plate; 2. Anchor bolt; 3. Storage groove; 4. Piezoelectric film sensor; 5. Opening groove; 6. Shape memory alloy honeycomb plate; 7. Limiting groove; 8. Limiting block; 9. Collar; 10. Sand injection hole; 11. Plug; 12. Sleeve; 13. Elastic rubber; 14. Grouting groove; 15. Stop block; 16. Stress gauge; 17. Locking nut; 18. Grout outlet hole; 19. Grouting hole. Detailed Implementation

[0037] 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.

[0038] Please see Figures 1-8 This invention provides an AI-based intelligent monitoring and early warning device for mine pressure, comprising a pressure plate 1 and anchor bolts 2. A slot is provided through the top of the pressure plate 1, and the anchor bolts 2 are located inside the slot. Multiple sets of anchor bolts 2 are provided, each set being a hollow structure. A storage slot 3 is provided on the top of the pressure plate 1, and piezoelectric thin-film sensors 4 are detachably installed inside the storage slot 3. Uniformly arranged opening slots 5 are provided on the outer surface of the pressure plate 1, and each set of opening slots 5 communicates with the storage slot 3. A circular ring-shaped... The shape memory alloy honeycomb panel 6 has a stress gauge 16 sleeved at one end of each pair of anchor rods 2 that are close to each other. One set of anchor rods 2 has a uniformly arranged MEMS multimodal sensor embedded in its outer surface. The MEMS multimodal sensor is used to monitor the influence of external load on prestress. The MEMS multimodal sensor also supports multidimensional data synchronous acquisition. The stress gauge 16 and the MEMS multimodal sensor data are transmitted to the deployed edge server. The LSTM model is used for time series analysis to generate the mine pressure risk level and feed it back to the terminal equipment.

[0039] Furthermore, the pressure plate 1, as the basic load-bearing component of the entire device, bears loads such as pressure from above the mine. Two sets of anchor bolts 2, with a hollow structure, are installed. The anchor bolts 2 play a crucial role in anchoring during mine support. The hollow structure facilitates the wiring arrangement of the stress gauge 16 and the MEMS multimodal sensor, enabling data transmission and other functions. The storage slot 3, located on top of the pressure plate 1, provides installation space for the piezoelectric thin-film sensor 4. The piezoelectric thin-film sensor 4 can convert the pressure changes borne by the pressure plate 1 into electrical signals, thereby achieving quantitative pressure sensing. The shape memory alloy honeycomb plate 6, installed on top of the pressure plate 1, can adapt to the curved deformation of the roadway roof. When the mine pressure changes, it can adjust its own pressure accordingly. The deformation adapts to a certain degree of pressure change, and at the same time transmits pressure information to the piezoelectric thin film sensor 4. The stress gauge 16 is sleeved on the end of the two sets of anchor rods 2 that are close to each other, and is used to directly measure the stress on the anchor rod 2. The MEMS multimodal sensor is used to monitor the influence of external load on prestress, and also has the ability to acquire multi-dimensional data synchronously. It can simultaneously acquire data of multiple modes such as stress, strain, and acceleration, and comprehensively perceive the mechanical environment of the anchor rod 2 from multiple dimensions, providing a rich data source for subsequent mine pressure analysis and improving the accuracy of mine pressure monitoring data. The MEMS multimodal sensor, stress gauge 16 and piezoelectric thin film sensor 4 are all existing technologies, and their working principles are not described in detail.

[0040] The top of the pressure plate 1 is provided with equidistantly arranged limiting grooves 7, and each set of limiting grooves 7 is connected to the groove opening. The limiting grooves 7 are located inside the storage groove 3 and have an L-shaped structure. The bottom of the pressure plate 1 is provided with equidistantly arranged slots, and the slots are connected to the limiting grooves 7. The width of the slots is half that of the limiting grooves 7. Each set of limiting grooves 7 contains a limiting block 8. The surfaces of multiple sets of limiting blocks 8 that are close to each other are fitted with collars 9, and the outer surface of the collars 9 abuts against the inner surface of the groove opening. The collars 9 are fitted onto the outer surface of another set of anchor rods 2 and welded for fixation. The bottom of the collar 9 is provided with multiple sets of sand injection holes 10, which are used to fill the gap between the sleeve 12 and the anchor rod 2 with fine sand. The inside of the sand injection hole 10 is fitted with a stopper 11, and the bottom of the stopper 11 is provided with a pull rope. One set of anchor rods 2 is threadedly connected to a stop block 15 at the end away from the stress gauge 16. The stop block 15 has a T-shaped structure, and the outer surface of the stop block 15 is in contact with the inner surface of the other end of the sleeve 12. The other set of anchor rods 2 is threadedly connected to a locking nut 17 at the end away from the stress gauge 16. The locking nut 17 is located below the pressure plate 1.

[0041] Furthermore, during assembly, the anchor rod 2 is first passed through the slot, and the limiting block 8 is aligned with the slot. The anchor rod 2 is pushed upward, and then the anchor rod 2 is rotated clockwise, causing the limiting block 8 to abut against the limiting groove 7. Then, the anchor rod 2, equipped with the stress gauge 16, is fixed in the pressure plate 1 by the engagement of the collar 9 with the slot and the support of the limiting block 8 by the limiting groove 7. The sand injection hole 10 at the bottom of the collar 9 is used to fill the gap between the sleeve 12 and the anchor rod 2 with fine sand. Filling with fine sand can increase the friction between the anchor rod 2 and the sleeve 12, thereby improving the anchor rod 2's strength. The anchoring effect is excellent, and through the cooperation of structures such as the limiting groove 7, limiting block 8, collar 9 and slot, the pressure plate 1 and anchor rod 2 are tightly connected, which can effectively resist various forces generated by mine pressure, reduce the relative displacement between device components, and improve the stability of the overall structure in complex mine stress environment. At the same time, the threaded connection of the stop block 15 and the locking nut 17 facilitates the installation and disassembly of the anchor rod 2. The sand injection hole 10 is controlled by the stop plug 11 and the pull rope, which facilitates sand injection operation and subsequent maintenance, and reduces the difficulty and cost of device installation and maintenance.

[0042] An anchor rod 2 is fitted with a sleeve 12 on its outer surface, and one end of the sleeve 12 is connected to the top of the pressure plate 1. The outer surface of the sleeve 12 abuts against the inner surface of the shape memory alloy honeycomb plate 6. An elastic rubber 13 is embedded in the inner surface of the sleeve 12. The sleeve 12 is made of soft plastic. During grouting, the grout solidifies and squeezes the sleeve 12, causing the elastic rubber 13 to come into close contact with the anchor rod 2 and the stress gauge 16. The outer surface of the sleeve 12 is provided with uniformly arranged grout outlet holes 18, and the grout outlet holes 18 are flared outwards. The top of the pressure plate 1 is provided with equidistantly arranged grouting holes 19, and the grouting holes 19 are distributed at intervals with the limiting groove 7. One end of the sleeve 12 is provided with a grouting groove 14, and the grouting groove 14 is located outside the grouting hole 19.

[0043] Furthermore, the sleeve 12 is made of soft plastic, with one end connected to the top of the pressure plate 1. The inner surface is inlaid with elastic rubber 13. During grouting, the pressure generated by the solidification of the grout will squeeze the sleeve 12, causing the elastic rubber 13 to come into close contact with the anchor rod 2 and the stress gauge 16. This allows for better transmission of the stress on the anchor rod 2, enabling the stress gauge 16 to accurately sense stress changes. This effectively reduces losses and interference during stress transmission, improves the accuracy of the stress gauge 16's measurements, and makes the monitoring data more reliable. During grouting, the grout enters the grouting groove 14 through the grouting hole 19 and then flows out through the grout outlet hole 18. The grout outlet hole 18 is flared outwards to facilitate the grout flow and even distribution in the borehole, thus playing an anchoring and sealing role and ensuring the stability of the mine support.

[0044] The early warning method of this invention involves deploying the monitoring equipment described herein in key areas of the mine, such as fully mechanized mining faces or roadways. These devices collect data in real time via IoT technology and transmit it to a data center, forming an intelligent monitoring network covering the entire mining area. Distributed storage systems such as Hadoop and Spark are used for efficient storage of the massive monitoring data. Data preprocessing techniques, including data cleaning, noise reduction, and compression, are employed to improve data quality and the accuracy of subsequent analysis, generating mine pressure distribution maps and trend maps to facilitate managers' intuitive understanding of the mine's safety status. Then, an LSTM model is constructed based on historical mine pressure data using AI algorithms. This is combined with statistical analysis and unsupervised learning algorithms such as K-means clustering and DBSCAN to monitor data anomalies in real time, quickly identify potential mine pressure risks, and set different levels of early warning thresholds based on the actual conditions of the mine and expert experience. When monitoring data triggers early warning conditions, the system automatically generates early warning information and quickly notifies relevant personnel through various means such as SMS, email, and APP push notifications. Simultaneously, based on the early warning level, targeted emergency response suggestions and measures are provided to help the mine respond quickly to mine pressure disasters.

[0045] Furthermore, sensor deployment: First, stress gauges 16, denoted as Pi(t), and displacement sensors, denoted as Dj(t), are deployed in key areas of the mine, such as fully mechanized mining faces and roadways, where i and j are sensor numbers and t is time;

[0046] Data collection: Sensor data is transmitted to the data center in real time via IoT technology, forming a time series dataset {Pi(t), Dj(t)};

[0047] Data preprocessing: Data cleaning, noise reduction, compression and other techniques are used. The processed data is denoted as {P^i(t), D^j(t)} and a distributed storage system is used to efficiently store the preprocessed data.

[0048] AI algorithm model development: Based on historical mine pressure data {P^i(t), D^j(t)}, key features are selected or extracted, and then an LSTM model is constructed for prediction. The prediction results are then integrated with those of decision trees and support vector machines.

[0049] Statistical analysis: Outliers were detected using a threshold method, i.e., when |Pi(t)−μPi|>σPi×k, it was considered an outlier, where μPi and σPi are the mean and standard deviation of Pi(t), respectively, and k is a coefficient;

[0050] Early warning threshold setting: Based on expert experience and the actual conditions of the mine, different levels of early warning thresholds are set;

[0051] Real-time warning: An warning is triggered when the monitored data P^i(t) or D^j(t) is greater than the preset value.

[0052] The working principle and usage process of this invention are as follows: Before use, multiple sets of MEMS multimodal sensors are first arranged on the surface of the anchor rod 2. Then, the anchor rod 2 is inserted into the slot, and the limiting block 8 is aligned with the slot. Then, the anchor rod 2 is pushed upward until the limiting block 8 moves to the top of the slot. Then, the anchor rod 2 is rotated clockwise so that the limiting block 8 can fall into the limiting groove 7. Next, the anchor rod 2 is fixed by locking nut 17, and the stop block 15 is inserted into the top opening of the sleeve 12. The stop block 15 is rotated so that the stop block 15 is tightened clockwise with the top of the anchor rod 2. Then, fine sand is filled into the sleeve 12 through the sand injection hole 10, and the sand injection hole 10 is blocked by the stop plug 11. Through the compact structure, not only is the overall weight of the monitoring device reduced, but also the loss and interference in the stress transmission process are reduced.

[0053] Then, the assembled anchor rod 2 and sleeve 12 are inserted into the pre-drilled hole, and the pressure plate 1 is fixed on the top plate. Then, grout is injected into the grouting groove 14 through the grouting hole 19. The injected grout will flow into the hole through the grout outlet 18. During the solidification process, the grout will squeeze the inner wall of the sleeve 12, thereby pushing the elastic rubber 13 closer to the anchor rod 2 and stress gauge 16 to achieve gapless stress transmission.

[0054] Furthermore, when recovering the anchor rod 2 and stress gauge 16, first remove the stopper 11 to collect the fine sand inside the sleeve 12, then remove the locking nut 17, and rotate the anchor rod 2 counterclockwise so that the top of the anchor rod 2 is no longer connected to the stop block 15. Then rotate the anchor rod 2 to align the limiting block 8 with the slot, and then pull the anchor rod 2 down. Since the elastic rubber 13 has a certain elasticity, it can avoid the probability of friction between the anchor rod 2 or stress gauge 16 and the sleeve 12, reducing wear. Then the anchor rod 2 and stress gauge 16 can be removed from the sleeve 12 and the pressure plate 1.

[0055] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A mine pressure intelligent monitoring and early warning device based on AI algorithm, comprising a pressure plate (1) and anchor rods (2), wherein a slot is provided through the top of the pressure plate (1), the anchor rods (2) are located inside the slot, and multiple sets of anchor rods (2) are provided, each set of anchor rods (2) being a hollow structure, characterized in that, The top of the pressure plate (1) is provided with a storage groove (3), and the inside of the storage groove (3) is detachably installed with uniformly arranged piezoelectric thin film sensors (4). The outer surface of the pressure plate (1) is provided with uniformly arranged opening grooves (5), and each set of opening grooves (5) is connected to the storage groove (3). The top of the pressure plate (1) is equipped with a circular shape memory alloy honeycomb plate (6). Each pair of anchor rods (2) has a stress gauge (16) attached to one end of each pair of anchor rods (2). One pair of anchor rods (2) has a uniformly arranged MEMS multimodal sensor embedded in its outer surface. The MEMS multimodal sensor is used to monitor the influence of external loads on prestress.

2. The intelligent monitoring and early warning device for mine pressure based on AI algorithm according to claim 1, characterized in that, The top of the pressure plate (1) is provided with equidistantly arranged limiting grooves (7), and each set of limiting grooves (7) is connected to the groove opening. The limiting grooves (7) are located inside the storage groove (3). The limiting grooves (7) have an L-shaped structure. The bottom of the pressure plate (1) is provided with equidistantly arranged slots, and the slots are connected to the limiting grooves (7). The width of the slots is half that of the limiting grooves (7). Each set of limiting grooves (7) has a limiting block (8) placed inside. Multiple sets of limiting blocks (8) are fitted with collars (9) on their surfaces that are close to each other. The outer surface of the collars (9) abuts against the inner surface of the groove opening. The collars (9) are fitted onto the outer surface of another set of anchor rods (2) and welded and fixed.

3. The intelligent monitoring and early warning device for mine pressure based on AI algorithm according to claim 1, characterized in that, The outer surface of the anchor rod (2) is fitted with a sleeve (12), and one end of the sleeve (12) is connected to the top of the pressure plate (1). The outer surface of the sleeve (12) abuts against the inner surface of the shape memory alloy honeycomb plate (6). An elastic rubber (13) is embedded in the inner surface of the sleeve (12). The sleeve (12) is made of soft plastic. During grouting, the grout solidifies and squeezes the sleeve (12) so that the elastic rubber (13) is close to the anchor rod (2) and the stress gauge (16).

4. The intelligent monitoring and early warning device for mine pressure based on AI algorithm according to claim 2, characterized in that, The bottom of the collar (9) is provided with multiple sets of sand injection holes (10), and the sand injection holes (10) are used to fill the gap between the sleeve (12) and the anchor rod (2) with fine sand. The inside of the sand injection hole (10) is fitted with a plug (11), and the bottom of the plug (11) is provided with a pull rope.

5. The intelligent monitoring and early warning device for mine pressure based on AI algorithm according to claim 3, characterized in that, One set of anchor rods (2) has a stop block (15) threaded to the end away from the stress gauge (16), and the stop block (15) has a T-shaped structure. The outer surface of the stop block (15) is in contact with the inner surface of the other end of the sleeve (12). The other set of anchor rods (2) has a locking nut (17) threaded to the end away from the stress gauge (16), and the locking nut (17) is located below the pressure plate (1).

6. The intelligent monitoring and early warning device for mine pressure based on AI algorithm according to claim 3, characterized in that, The outer surface of the sleeve (12) is provided with uniformly arranged slurry outlet holes (18), and the slurry outlet holes (18) are in the shape of an outwardly expanding trumpet.

7. The intelligent monitoring and early warning device for mine pressure based on AI algorithm according to claim 1, characterized in that, The top of the pressure plate (1) is provided with equidistant grouting holes (19), and the grouting holes (19) and the limiting groove (7) are distributed at intervals.

8. The intelligent monitoring and early warning device for mine pressure based on AI algorithm according to claim 3, characterized in that, One end of the sleeve (12) is provided with a grouting groove (14), and the grouting groove (14) is located outside the grouting hole (19).

9. The intelligent monitoring and early warning device for mine pressure based on AI algorithm according to claim 1, characterized in that, The MEMS multimodal sensor also supports multidimensional data synchronous acquisition.

10. The intelligent monitoring and early warning device for mine pressure based on AI algorithm according to claim 1, characterized in that, The stress gauge (16) and MEMS multimodal sensor data are transmitted to the deployed edge server, where LSTM model is used for time series analysis to generate the mine pressure risk level and feed it back to the terminal equipment.

Images