A digital mine safety production monitoring method and system based on the Internet of Things

Abstract

This invention discloses a digital mine safety production monitoring method and system based on the Internet of Things (IoT), relating to the field of mine safety technology. The invention constructs a multi-dimensional monitoring system by integrating multi-source data, combining sensor topology maps and 3D point cloud models to achieve three-dimensional perception and visualization of safety status. An improved LoRaWAN is used to ensure real-time data transmission. Based on an optimized CNN sensor data prediction model, residual calculation triggers a first-level early warning to actively cut off equipment power. Risk values ​​are calculated using D-S evidence theory, combined with real-time 3D stress cloud maps generated by finite element analysis, enabling intelligent matching of historical cases for second-level early warnings and dynamic escape path planning for third-level early warnings. This invention possesses adaptive optimization capabilities for model parameters and dynamic updates to the case library. Through 3D visualization, it improves communication efficiency while achieving precise risk positioning, multi-level differentiated early warnings, and scientific decision-making, constructing a closed-loop ecosystem for mine safety management.

Description

Technical Field

[0001] This invention belongs to the field of mine safety technology, and specifically relates to a digital mine safety production monitoring method and system based on the Internet of Things. Background Technology

[0002] With the continuous development of the mining industry, mine safety production faces challenges in many aspects, including complex environment, technical equipment and management level. It is still necessary to improve the level of safety production technology to ensure the healthy development of mining enterprises and the safety of workers' lives.

[0003] Chinese patent CN108757040B discloses a mine safety monitoring system; it connects the upper and lower parts of the mine through a wireless communication network, and monitors environmental parameters and personnel location information in the mining area in real time by setting up fixed monitoring devices, mobile monitoring devices, and camera devices, realizing the linkage management of the monitoring system and improving the overall level of automation and informatization; by acquiring various types of environmental data in the mining area in real time, it determines the corresponding environmental safety level and the corresponding dangerous area range, and when the environmental safety level is dangerous, it controls multiple infrared sensors in the area where the monitoring sensor group is located to activate the detection of personnel approach signals and send alarm signals to the fixed monitoring devices in the corresponding area.

[0004] Existing technologies for mine environmental monitoring are limited in scope, focusing only on environmental parameters and personnel location. They lack real-time monitoring of the mine's mechanical state and equipment operating parameters, failing to comprehensively reflect the mine's overall safety status. Furthermore, their communication and data processing capabilities are insufficient, failing to address issues such as signal attenuation, electromagnetic interference, and node conflicts in complex mine environments, potentially leading to data transmission delays or losses. Risk warning mechanisms are simplistic, triggering alarms solely based on environmental safety levels, lacking multi-level warning and tiered response capabilities, and lacking intelligent decision-making abilities such as historical case matching and dynamic escape route planning. Finally, they lack predictive analysis capabilities, failing to predict trends in sensor data and thus hindering the early detection of potential risks. Summary of the Invention

[0005] (a) Technical problems to be solved

[0006] To address the problems in related technologies, this invention provides a digital mine safety production monitoring method based on the Internet of Things. This invention solves the problems of single-dimensional mine environmental monitoring, insufficient communication and data processing capabilities, simple risk warning mechanisms, and lack of predictive analysis by multi-dimensional data fusion and accurate monitoring, efficient communication and predictive maintenance, multi-level early warning and intelligent decision-making, as well as three-dimensional visualization and mechanical analysis.

[0007] (II) Technical Solution

[0008] To solve the above-mentioned technical problems, the present invention is achieved through the following technical solution:

[0009] S1. Set data related to mine safety and deploy sensors; construct a sensor topology map and a 3D point cloud model of the mine; set response actions for level 1, level 2, and level 3 early warnings;

[0010] S2. Improve the traditional LoRaWAN model by combining it with the mining environment to obtain the improved LoRaWAN; construct a CNN sensor data prediction model, and optimize the parameters of the CNN sensor data prediction model by combining it with optimization algorithms to obtain the optimized sensor data prediction model.

[0011] S3. Collect raw sensor data through sensors and combine it with the improved LoRaWAN to obtain real-time multimodal data; input the real-time multimodal data into the optimized CNN sensor data prediction model to obtain predicted sensor data; calculate the residual between the real-time multimodal data and the predicted sensor data to obtain the residual value set.

[0012] S4. Determine whether to implement a Level 1 warning based on the residual value set; obtain the fusion risk value using DS evidence theory; obtain the stress concentration coefficient using finite element analysis, combined with the sensor topology map and the 3D point cloud model of the mine; and determine whether to implement a Level 2 or Level 3 warning based on the fusion risk value and the stress concentration coefficient.

[0013] Preferably, step S1 includes the following steps:

[0014] S11. Set up the mine environment dataset Mining Mechanics Dataset Mine Location Tag Dataset and device status parameter set

[0015] Among them, a i Let m represent the i-th type of environmental data related to mine safety, and m1 represent the total number of types of environmental data related to mine safety; b i This represents mechanical data related to mine safety, where m2 represents the total number of types of mechanical data related to mine safety; c i This represents location tag data related to mine safety, where m3 represents the total number of types of mechanical data related to mine safety.

[0016] d i This represents the i-th equipment parameter used to reflect the equipment status, and m4 represents the total number of equipment parameters; mine environmental data such as CH4, CO concentration and temperature and humidity data; mine mechanical data such as roof displacement and rock strain data; mine positioning tag data such as personnel and equipment coordinates; equipment parameters such as equipment vibration frequency and pressure.

[0017] S12. Based on the mine environment dataset, mine mechanics dataset, mine positioning tag dataset, and equipment status parameter set, deploy sensors at key locations in the mine.

[0018] S13. Register the locations of the deployed sensors to three-dimensional coordinates to obtain a sensor topology map with spatial attributes; use a mobile laser scanner to scan the mine to obtain a three-dimensional point cloud model of the mine.

[0019] S14. Set the response action for a Level 1 warning to stop the power supply of the relevant equipment;

[0020] Set the similarity threshold to u; set the response action for the level 2 warning to match historical cases from the risk knowledge base, and use the solutions of historical cases with similarity ≥ similarity threshold u to resolve the issue. If no historical cases with similarity ≥ similarity threshold u exist, generate a risk report and send it to the safety officer, who will then conduct risk analysis and make a decision.

[0021] The response actions for the three-level early warning system are to obtain real-time personnel location and real-time three-dimensional stress cloud map, and to plan personnel escape routes.

[0022] The above steps involve setting up datasets for the mine environment, mechanics, positioning, and equipment status, and deploying sensors at key locations to create a sensor topology map with spatial attributes and a 3D point cloud model of the mine. Three levels of early warning response actions are defined, including equipment power cut-off for Level 1 warnings, historical case matching and risk report generation for Level 2 warnings, and escape route planning for Level 3 warnings. This achieves comprehensive monitoring of the mine's safety status. The three-level early warning mechanism can not only respond quickly to potential hazards and cut off equipment power to prevent accidents from escalating, but also provide solutions through historical case matching, or generate risk reports for safety officers to analyze and make decisions when no matching cases are available. In emergencies, real-time positioning and stress cloud maps can be used to plan escape routes, significantly improving the mine's emergency response capabilities and personnel safety levels.

[0023] Preferably, step S2 includes the following steps:

[0024] S21. Improve the traditional LoRaWAN to obtain the improved LoRaWAN;

[0025] S22. Construct a CNN sensor data prediction model, collect historical data and combine optimization algorithms to adjust the parameters of the CNN sensor data prediction model, and obtain an optimized CNN sensor data prediction model.

[0026] The above steps summarize and generalize the improvements made to the traditional LoRaWAN protocol, forming an improved LoRaWAN, and constructing a CNN sensor data prediction model. By collecting historical data and combining optimization algorithms to adjust the model parameters, an optimized CNN sensor data prediction model was finally obtained. The improved LoRaWAN is modified according to the real-time environment of the mine, making it more suitable for the transmission environment in the mine. The optimized CNN sensor data prediction model, through precise parameter adjustments, improves the prediction accuracy of abnormal data, provides more reliable technical support for mine safety, and enhances the system's proactive prevention and control capabilities.

[0027] Preferably, step S21 includes the following steps:

[0028] S211. Adjust the SF value according to the real-time mine signal-to-noise ratio to balance the relationship between roadway attenuation and electromagnetic interference;

[0029] S212. To address the conflict issue caused by high-density deployments within mines, a time-series allocation algorithm is used to divide time frames into fixed time slots and contention windows, prioritizing the propagation of abnormal data. The time slot allocation algorithm formula is as follows:

[0030]

[0031] Where T represents the time slot length, which is the time period allocated to each node in the TDMA frame; M represents the frame length, which is the duration of the entire TDMA frame; N represents the number of active nodes, which is the number of nodes currently participating in communication; P represents the received power, which is the signal power received by the node; and Q represents the noise power spectral density, which is the noise power per unit bandwidth.

[0032] S213. The node dynamically selects the next hop based on the routing cost function, combined with the link quality and remaining power in the mine; the routing cost function formula is as follows:

[0033]

[0034] Where C represents the routing cost, used to evaluate the cost or expense of choosing a certain path; α, β, and λ all represent weighting coefficients, used to adjust the importance of different factors in routing selection; RSSI represents the received signal strength indicator, i.e., the received signal strength; PLR represents the packet loss rate, i.e., the proportion of data packets lost during transmission; ER represents the remaining battery power, i.e., the remaining battery power of the node.

[0035] S214. An improved LoRaWAN is obtained through S211, S212, and S213;

[0036] The above steps, by adjusting the spreading factor (SF value) in real time to adapt to changes in the signal-to-noise ratio in the mine, optimizing frame division using a timing allocation algorithm, prioritizing the propagation of abnormal data, and dynamically selecting communication paths using a routing cost function, comprehensively consider signal strength, packet loss rate, and remaining node power. This effectively improves the performance and reliability of the mine communication system. Real-time adjustment of the SF value balances signal attenuation and electromagnetic interference, ensuring data transmission stability. The timing allocation algorithm resolves conflicts under high-density deployment, ensuring timely transmission of abnormal data. The introduction of the routing cost function enables nodes to intelligently select paths based on link quality and power status, reducing data transmission costs and latency, and overall improving the system's communication efficiency and response speed to mine safety risks.

[0037] Preferably, step S22 includes the following steps:

[0038] S221. Construct a CNN sensor data prediction model and set the initial learning rate of the CNN sensor data prediction model to p1.

[0039] S222. Collect historical normal operating condition data and historical abnormal operating condition data of mine sensors to obtain training data; collect abnormal mine data corresponding to level 1, level 2, and level 3 early warnings to obtain test data.

[0040] S223. Set the training error threshold q1 and the maximum number of training iterations; use the training data to repeatedly train the CNN sensor data prediction model; after each round of training, calculate the training error q2 of the CNN sensor data prediction model and adjust the initial learning rate according to the training error;

[0041] When q2≤q1 or the maximum number of iterations is reached, a well-trained CNN sensor data prediction model with a learning rate of q2 is obtained.

[0042] S224. Test the trained CNN sensor data prediction model using test data. After the test is completed, an optimized CNN sensor data prediction model is obtained.

[0043] The above steps involve setting an initial learning rate for the model and collecting historical normal and abnormal operating condition data from the mine as training data, as well as abnormal data corresponding to various levels of early warning as test data. By setting a training error threshold and a maximum number of iterations, the model is repeatedly trained using the training data, and the learning rate is adjusted according to the training error. When the error requirement is met or the maximum number of iterations is reached, a well-trained model is obtained. Finally, the model is tested using test data to obtain an optimized CNN sensor data prediction model. This improves the accuracy of the optimized CNN sensor data prediction model, specifically targeting the prediction of sensor data in mines, providing strong support for timely early warning and emergency response, and effectively enhancing the intelligent level of mine safety production.

[0044] Preferably, step S224 includes the following steps:

[0045] S2241. Set the test accuracy threshold r1, and use the test data to test the trained CNN sensor data prediction model to obtain the test accuracy r2.

[0046] S2242. If r2≥r1, then the trained CNN sensor data prediction model is used as the optimized CNN sensor data prediction model; otherwise, the Hippo algorithm is used to find the learning rate of the trained CNN sensor data prediction model, obtain the global optimal solution, and use the global optimal solution as the learning rate of the trained CNN sensor data prediction model to obtain the optimized CNN sensor data prediction model.

[0047] By setting the test accuracy threshold through the above steps, the trained CNN sensor data prediction model is tested and verified. If the test accuracy meets the preset threshold, the model is directly used as the optimized model. If it does not meet the threshold, the learning rate of the model is optimized using an optimization algorithm to find the global optimal solution, thereby obtaining a more accurate optimized CNN sensor data prediction model. This ensures that the model's prediction performance meets the expected standards.

[0048] Preferably, step S2242, which uses the Hippo algorithm to find the learning rate of the trained CNN sensor data prediction model and obtain the globally optimal solution, includes the following steps:

[0049] S22421. Construct a hippopotamus population w, and set the size of the hippopotamus population as z. Then the hippopotamus population w = {w1, w2, ..., w...} i ,...,w z}, where w i This represents the i-th hippopotamus in the hippopotamus population;

[0050] S22422. Randomly generate an initial position group of the hippopotamus population x = {x1, x2, ..., x} based on the learning rate. i,...,x z}, where x i This represents the initial position of the i-th hippopotamus in the hippopotamus population;

[0051] S22423. Start the iteration operation and set the maximum number of optimization iterations to ε. During each iteration, update the position of each hippo in the initial position set of the hippo population according to the test accuracy. And during each iteration, obtain the global best hippo individual and the global best target in the hippo population.

[0052] S22424. When r2≥r1 or the maximum number of iterations for optimization is reached, the hippopotamus population stops the iteration operation and obtains the global optimal solution.

[0053] The above steps optimize the learning rate of the CNN sensor data prediction model using an optimization algorithm. First, a hippo population is constructed and initial positions are randomly generated. Then, the hippo positions are updated iteratively, and the globally optimal hippo individual and target are found based on the test accuracy. When the test accuracy meets a preset threshold or reaches the maximum number of iterations, iteration stops, and the globally optimal solution for the learning rate is obtained. The introduction of the hippo algorithm significantly improves the optimization performance of the CNN sensor data prediction model.

[0054] Preferably, step S3 includes the following steps:

[0055] S31. Collect various data from the mine environment dataset, mine mechanics dataset, mine positioning tag dataset, and equipment status parameter set through sensors to obtain raw sensor data; input the raw sensor data into the improved LoRaWAN for transmission to obtain real-time multimodal data with timestamp alignment.

[0056] S32. Input the real-time multimodal data into the optimized CNN sensor data prediction model to obtain the predicted sensor data;

[0057] S33. Set the residual value threshold to e, calculate the residual between the data in the predicted sensor data and the data in the real-time multimodal data, and obtain the residual value set;

[0058] The above steps involve collecting multi-modal data from the mine through sensors and using an improved LoRaWAN for efficient transmission, ensuring data timestamp alignment and real-time performance. Real-time multimodal data is then input into an optimized CNN sensor data prediction model for prediction. By calculating the residual between the predicted data and the real-time data and setting a residual value threshold, a residual value set is obtained. This enables real-time and efficient collection and processing of mine data, providing a solid data foundation for safety production monitoring.

[0059] Preferably, step S4 includes the following steps:

[0060] S41. Determine whether there are residual values ​​in the residual value set that are greater than or equal to the residual value threshold e. If so, execute a level one warning and collect the sensor data corresponding to the residual values ​​in the residual value set that are greater than or equal to the residual value threshold e to obtain the abnormal dataset; otherwise, maintain the current state.

[0061] S42. Set the threshold for the first-level risk value as f and the threshold for the second-level risk value as g. Collect geological exploration data and equipment log data. Combine the geological exploration data and equipment log data with the abnormal data in the abnormal dataset to obtain combined abnormal data. Use the DS evidence theory to calculate the combined abnormal data to obtain the fusion risk value h.

[0062] S43. Set the stress concentration factor threshold to k; extract the mechanical data from the real-time multimodal data to obtain real-time mechanical data; superimpose the real-time mechanical data with the three-dimensional point cloud model of the mine to obtain superimposed mechanical data; use finite element analysis to calculate the superimposed mechanical data to obtain the stress concentration factor; combine the stress concentration factor with the sensor topology map to obtain a three-dimensional stress cloud map.

[0063] S44. If the fusion risk value h ≥ the first-level risk value threshold f, execute the second-level warning; if the fusion risk value h ≥ the second-level risk value threshold g or the stress concentration factor j ≥ the stress concentration factor threshold k, execute the third-level warning.

[0064] The above steps determine whether to trigger a Level 1 warning by judging whether the residual value set exceeds a preset threshold, and collect corresponding abnormal data; combine geological exploration and equipment log data, and use DS evidence theory to calculate the fused risk value; extract real-time mechanical data and overlay it with the three-dimensional model of the mine, perform finite element analysis to obtain the stress concentration coefficient, and generate a three-dimensional stress cloud map; based on the comparison results of the fused risk value and stress concentration coefficient with their respective thresholds, execute the corresponding Level 2 or Level 3 warning; achieve accurate identification and graded warning of mine safety risks; improve the comprehensiveness and accuracy of risk assessment; provide intuitive mechanical analysis basis for warning; the graded warning mechanism ensures differentiated responses for different risk levels, effectively improving the warning capability and emergency response efficiency of the mine safety monitoring system.

[0065] A digital mine safety production monitoring system based on the Internet of Things (IoT) is used to implement the aforementioned digital mine safety production monitoring method based on the Internet of Things (IoT), including a data acquisition and initialization module, a communication and prediction model optimization module, a real-time data processing and anomaly detection module, and a risk assessment and early warning triggering module.

[0066] The data acquisition and initialization module is used to define mine safety-related datasets; deploy sensors and construct topology maps and 3D point cloud models; and set the response action rules for three-level early warnings. (Key sub-modules are listed below.)

[0067] The communication and prediction model optimization module is used to improve the traditional LoRaWAN protocol and to build and optimize the CNN sensor data prediction model to obtain an optimized CNN sensor data prediction model.

[0068] The real-time data processing and anomaly detection module is used to collect multimodal data through sensors, align the data through improved LoRaWAN transmission, and use optimized CNN sensor data prediction model data to calculate the residual between real-time data and predicted values ​​to obtain a residual value set.

[0069] The risk assessment and early warning triggering module is used to determine whether to trigger a level one early warning based on the residual value set; and to determine whether to trigger a level two or three early warning by integrating DS evidence theory and finite element analysis to calculate the risk value and stress concentration coefficient.

[0070] (3) Beneficial effects

[0071] The present invention has the following beneficial effects:

[0072] This invention constructs a multi-dimensional data monitoring system by integrating mine environmental data, mechanical data, equipment operating parameters, and personnel positioning information; combined with sensor topology maps and mine 3D point cloud models, it realizes three-dimensional perception and visualization of mine safety status, comprehensively covering environmental, equipment, personnel, and geomechanical risks, and significantly improving monitoring accuracy and coverage.

[0073] This invention addresses the challenges of complex mining environments by employing an improved LoRaWAN protocol that dynamically adjusts communication parameters, timing allocation algorithms, and routing optimization. This effectively solves the problems of signal attenuation and node conflicts, ensuring the real-time performance and reliability of data transmission. Furthermore, based on a CNN sensor data prediction model and optimized algorithms, the system can predict abnormal data trends in advance and trigger a first-level early warning through residual calculation, thus shifting from passive response to proactive prevention and control.

[0074] This invention directly cuts off the power supply to the equipment through a first-level early warning to prevent the accident from escalating; a second-level early warning matches historical cases with a risk knowledge base to assist in rapid decision-making; a third-level early warning combines real-time personnel positioning with a three-dimensional stress cloud map to dynamically plan the optimal escape route; and by combining multi-source data fusion of DS evidence theory and mechanical calculations of finite element analysis, the system can quantify risk values ​​and stress concentration coefficients, providing a scientific basis for differentiated emergency responses.

[0075] By optimizing the predictive model parameters through algorithmic optimization and continuously updating the historical case library, the system possesses dynamic learning and iteration capabilities. The three-dimensional visualization interface, linked with real-time data, not only supports precise risk positioning but also provides data support for long-term mine safety planning. Compared with traditional solutions, this invention achieves breakthroughs in communication efficiency, early warning intelligence, decision-making science, and system adaptability, constructing a fully closed-loop ecosystem for mine safety management.

[0076] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0077] To more clearly illustrate the technical solutions of the embodiments of the invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the invention. For those skilled in the art, the drawings can be obtained from these drawings without creative effort.

[0078] Figure 1 This is a flowchart illustrating a digital mine safety production monitoring method based on the Internet of Things according to the present invention.

[0079] Figure 2 This is a schematic diagram illustrating the process of deploying sensors and setting three-level early warning response actions in a digital mine safety production monitoring method based on the Internet of Things according to the present invention.

[0080] Figure 3 This is a schematic diagram of the process of obtaining the improved LoRaWAN in the Internet of Things-based digital mine safety production monitoring method of the present invention;

[0081] Figure 4 This is a flowchart illustrating the process of obtaining an optimized CNN sensor data prediction model in a digital mine safety production monitoring method based on the Internet of Things according to the present invention.

[0082] Figure 5 This is a schematic diagram of a module of a digital mine safety production monitoring system based on the Internet of Things according to the present invention. Detailed Implementation

[0083] The technical solutions of the embodiments of the invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the invention, and not all embodiments. Based on the embodiments of the invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the invention.

[0084] In the description of this invention, it should be understood that the terms "opening", "upper", "lower", "top", "middle", "inner", etc., which indicate orientation or positional relationship, are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the components or elements referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation on the invention.

[0085] Example 1:

[0086] Please see Figure 1 , Figure 2 , Figure 3 and Figure 4 This invention discloses a digital mine safety production monitoring method based on the Internet of Things, comprising the following steps:

[0087] S1. Set data related to mine safety and deploy sensors; construct a sensor topology map and a 3D point cloud model of the mine; set response actions for level 1, level 2, and level 3 early warnings;

[0088] S1 includes the following steps:

[0089] S11. Set up the mine environment dataset Mining Mechanics Dataset Mine Location Tag Dataset and device status parameter set

[0090] Among them, a i Let m represent the i-th type of environmental data related to mine safety, and m1 represent the total number of types of environmental data related to mine safety; b i This represents mechanical data related to mine safety, where m2 represents the total number of types of mechanical data related to mine safety; c i This represents location tag data related to mine safety, where m3 represents the total number of types of mechanical data related to mine safety.

[0091] d i This represents the i-th equipment parameter used to reflect the equipment status, and m4 represents the total number of equipment parameters; mine environmental data such as CH4, CO concentration and temperature and humidity data; mine mechanical data such as roof displacement and rock strain data; mine positioning tag data such as personnel and equipment coordinates; equipment parameters such as equipment vibration frequency and pressure.

[0092] S12. Based on the mine environment dataset, mine mechanics dataset, mine positioning tag dataset, and equipment status parameter set, deploy sensors at key locations in the mine.

[0093] S13. Register the locations of the deployed sensors to three-dimensional coordinates to obtain a sensor topology map with spatial attributes; use a mobile laser scanner to scan the mine to obtain a three-dimensional point cloud model of the mine.

[0094] S14. Set the response action for a Level 1 warning to stop the power supply of the relevant equipment;

[0095] Set the similarity threshold to u; set the response action for the level 2 warning to match historical cases from the risk knowledge base, and use the solutions of historical cases with similarity ≥ similarity threshold u to resolve the issue. If no historical cases with similarity ≥ similarity threshold u exist, generate a risk report and send it to the safety officer, who will then conduct risk analysis and make a decision.

[0096] The response actions for the three-level early warning system are to obtain real-time personnel location and real-time three-dimensional stress cloud map, and to plan personnel escape routes.

[0097] S2. Improve the traditional LoRaWAN model by combining it with the mining environment to obtain the improved LoRaWAN; construct a CNN sensor data prediction model, and optimize the parameters of the CNN sensor data prediction model by combining it with optimization algorithms to obtain the optimized sensor data prediction model.

[0098] S2 includes the following steps:

[0099] S21. Improve the traditional LoRaWAN to obtain the improved LoRaWAN;

[0100] S21 includes the following steps:

[0101] S211. Adjust the SF value according to the real-time mine signal-to-noise ratio to balance the relationship between roadway attenuation and electromagnetic interference;

[0102] S212. To address the conflict issue caused by high-density deployments within mines, a time-series allocation algorithm is used to divide time frames into fixed time slots and contention windows, prioritizing the propagation of abnormal data. The time slot allocation algorithm formula is as follows:

[0103]

[0104] Where T represents the time slot length; N represents the number of active nodes; P represents the received power, i.e., the signal power received by the node; and Q represents the noise power spectral density.

[0105] S213. The node dynamically selects the next hop based on the routing cost function, combined with the link quality and remaining power in the mine; the routing cost function formula is as follows:

[0106]

[0107] Where C represents the routing cost, α, β and λ are all weighting coefficients, RSSI represents the received signal strength indicator, PLR represents the packet loss rate, and ER represents the remaining battery power;

[0108] S214. An improved LoRaWAN is obtained through S211, S212, and S213;

[0109] S22. Improve the DS evidence theory to obtain the improved DS evidence theory;

[0110] S22. Construct a CNN sensor data prediction model, collect historical data and combine optimization algorithms to adjust the parameters of the CNN sensor data prediction model, and obtain an optimized CNN sensor data prediction model.

[0111] S22 includes the following steps:

[0112] S221. Construct a CNN sensor data prediction model and set the initial learning rate of the CNN sensor data prediction model to p1.

[0113] S222. Collect historical normal operating condition data and historical abnormal operating condition data of mine sensors to obtain training data; collect abnormal mine data corresponding to level 1, level 2, and level 3 early warnings to obtain test data.

[0114] S223. Set the training error threshold q1 and the maximum number of training iterations; use the training data to repeatedly train the CNN sensor data prediction model; after each round of training, calculate the training error q2 of the CNN sensor data prediction model and adjust the initial learning rate according to the training error;

[0115] When q2≤q1 or the maximum number of iterations is reached, a well-trained CNN sensor data prediction model with a learning rate of q2 is obtained.

[0116] S224. Test the trained CNN sensor data prediction model using test data. After the test is completed, an optimized CNN sensor data prediction model is obtained.

[0117] S224 includes the following steps:

[0118] S2241. Set the test accuracy threshold r1, and use the test data to test the trained CNN sensor data prediction model to obtain the test accuracy r2.

[0119] S2242. If r2≥r1, then the trained CNN sensor data prediction model is used as the optimized CNN sensor data prediction model; otherwise, the Hippo algorithm is used to find the learning rate of the trained CNN sensor data prediction model to obtain the global optimal solution. The global optimal solution is used as the learning rate of the trained CNN sensor data prediction model to obtain the optimized CNN sensor data prediction model.

[0120] The steps in S2242, which use the Hippo algorithm to find the learning rate of the trained CNN sensor data prediction model and obtain the globally optimal solution, include the following:

[0121] S22421. Construct a hippopotamus population w, and set the size of the hippopotamus population as z. Then the hippopotamus population w = {w1, w2, ..., w...} i ,...,w z}, where w i This represents the i-th hippopotamus in the hippopotamus population;

[0122] S22422. Randomly generate an initial position group of the hippopotamus population x = {x1, x2, ..., x} based on the learning rate. i ,...,x z}, where x i This represents the initial position of the i-th hippopotamus in the hippopotamus population;

[0123] S22423. Start the iteration operation and set the maximum number of optimization iterations to ε. During each iteration, update the position of each hippo in the initial position set of the hippo population according to the test accuracy. And during each iteration, obtain the global best hippo individual and the global best target in the hippo population.

[0124] S22424. When r2≥r1 or the maximum number of iterations for optimization is reached, the hippopotamus population stops the iteration operation and obtains the global optimal solution.

[0125] S3. Collect raw sensor data through sensors and combine it with the improved LoRaWAN to obtain real-time multimodal data; input the real-time multimodal data into the optimized CNN sensor data prediction model to obtain predicted sensor data; calculate the residual between the real-time multimodal data and the predicted sensor data to obtain the residual value set.

[0126] S3 includes the following steps:

[0127] S31. Collect various data from the mine environment dataset, mine mechanics dataset, mine positioning tag dataset, and equipment status parameter set through sensors to obtain raw sensor data; input the raw sensor data into the improved LoRaWAN for transmission to obtain real-time multimodal data with timestamp alignment.

[0128] S32. Input the real-time multimodal data into the optimized CNN sensor data prediction model to obtain the predicted sensor data;

[0129] S33. Set the residual value threshold to e, calculate the residual between the data in the predicted sensor data and the data in the real-time multimodal data, and obtain the residual value set;

[0130] S4. Determine whether to implement a Level 1 warning based on the residual value set; obtain the fusion risk value using DS evidence theory; obtain the stress concentration coefficient using finite element analysis, combined with the sensor topology map and the 3D point cloud model of the mine; determine whether to implement a Level 2 or Level 3 warning based on the fusion risk value and the stress concentration coefficient.

[0131] S4 includes the following steps:

[0132] S41. Determine whether there are residual values ​​in the residual value set that are greater than or equal to the residual value threshold e. If so, execute a level one warning and collect the sensor data corresponding to the residual values ​​in the residual value set that are greater than or equal to the residual value threshold e to obtain the abnormal dataset; otherwise, maintain the current state.

[0133] S42. Set the threshold for the first-level risk value as f and the threshold for the second-level risk value as g. Collect geological exploration data and equipment log data. Combine the geological exploration data and equipment log data with the abnormal data in the abnormal dataset to obtain combined abnormal data. Use the DS evidence theory to calculate the combined abnormal data to obtain the fusion risk value h.

[0134] S43. Set the stress concentration factor threshold to k; extract the mechanical data from the real-time multimodal data to obtain real-time mechanical data; superimpose the real-time mechanical data with the three-dimensional point cloud model of the mine to obtain superimposed mechanical data; use finite element analysis to calculate the superimposed mechanical data to obtain the stress concentration factor; combine the stress concentration factor with the sensor topology map to obtain a three-dimensional stress cloud map.

[0135] S44. If the fusion risk value h ≥ the first-level risk value threshold f, execute the second-level warning. If the fusion risk value h ≥ the second-level risk value threshold g or the stress concentration coefficient j ≥ the stress concentration coefficient threshold k, execute the third-level warning.

[0136] Example 2:

[0137] Please see Figure 5 A digital mine safety production monitoring system based on the Internet of Things (IoT) is used to implement the above-mentioned digital mine safety production monitoring method based on the Internet of Things (IoT), including a data acquisition and initialization module, a communication and prediction model optimization module, a real-time data processing and anomaly detection module, and a risk assessment and early warning triggering module.

[0138] The data acquisition and initialization module is used to define mine safety-related datasets; deploy sensors and construct topology maps and 3D point cloud models; and set the response action rules for three-level early warnings. (Key sub-modules are listed below.)

[0139] The communication and prediction model optimization module is used to improve the traditional LoRaWAN protocol and to build and optimize the CNN sensor data prediction model to obtain an optimized CNN sensor data prediction model.

[0140] The real-time data processing and anomaly detection module is used to collect multimodal data through sensors, align the data through improved LoRaWAN transmission, and use optimized CNN sensor data prediction model data to calculate the residual between real-time data and predicted values ​​to obtain a residual value set.

[0141] The risk assessment and early warning triggering module is used to determine whether to trigger a level one early warning based on the residual value set; and to determine whether to trigger a level two or three early warning by integrating DS evidence theory and finite element analysis to calculate the risk value and stress concentration coefficient.

[0142] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0143] The preferred embodiments of the invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention.

Claims

1. A digital mine safety production monitoring method based on the Internet of Things, characterized in that, Includes the following steps: S1. Set data related to mine safety and deploy sensors; construct a sensor topology map and a 3D point cloud model of the mine; set response actions for level 1, level 2, and level 3 early warnings; S2. Improve the traditional LoRaWAN model by combining it with the mining environment to obtain the improved LoRaWAN; construct a CNN sensor data prediction model, and optimize the parameters of the CNN sensor data prediction model by combining it with optimization algorithms to obtain the optimized sensor data prediction model. S3. Collect raw sensor data through sensors and combine it with improved LoRaWAN to obtain real-time multimodal data; input the real-time multimodal data into the optimized CNN sensor data prediction model to obtain predicted sensor data; Calculate the residual between real-time multimodal data and predicted sensor data to obtain a set of residual values; S3 includes the following steps: S31. Collect various data from the mine environment dataset, mine mechanics dataset, mine positioning tag dataset, and equipment status parameter set through sensors to obtain raw sensor data; input the raw sensor data into the improved LoRaWAN for transmission to obtain real-time multimodal data with timestamp alignment. S32. Input the real-time multimodal data into the optimized CNN sensor data prediction model to obtain the predicted sensor data; S33. Set the residual value threshold to e, calculate the residual between the data in the predicted sensor data and the data in the real-time multimodal data, and obtain the residual value set; S4. Determine whether to implement a Level 1 warning based on the residual value set; obtain the fusion risk value using DS evidence theory; obtain the stress concentration coefficient using finite element analysis, combined with the sensor topology map and the 3D point cloud model of the mine; determine whether to implement a Level 2 or Level 3 warning based on the fusion risk value and the stress concentration coefficient. S4 includes the following steps: S41. Determine whether there are residual values ​​in the residual value set that are greater than or equal to the residual value threshold e. If so, execute a level one warning and collect the sensor data corresponding to the residual values ​​in the residual value set that are greater than or equal to the residual value threshold e to obtain the abnormal dataset; otherwise, maintain the current state. S42. Set the threshold for the first-level risk value as f and the threshold for the second-level risk value as g. Collect geological exploration data and equipment log data. Combine the geological exploration data and equipment log data with the abnormal data in the abnormal dataset to obtain combined abnormal data. Use the DS evidence theory to calculate the combined abnormal data to obtain the fusion risk value h. S43. Set the stress concentration factor threshold to k; extract the mechanical data from the real-time multimodal data to obtain real-time mechanical data; superimpose the real-time mechanical data with the three-dimensional point cloud model of the mine to obtain superimposed mechanical data; use finite element analysis to calculate the superimposed mechanical data to obtain the stress concentration factor; combine the stress concentration factor with the sensor topology map to obtain a three-dimensional stress cloud map. S44. If the fusion risk value h ≥ the first-level risk value threshold f, execute the second-level warning. If the fusion risk value h ≥ the second-level risk value threshold g or the stress concentration coefficient j ≥ the stress concentration coefficient threshold k, execute the third-level warning.

2. The method for monitoring digital mine safety production based on the Internet of Things according to claim 1, characterized in that, S1 includes the following steps: S11. Set up the mine environment dataset Mining Mechanics Dataset Mine Location Tag Dataset and device status parameter set Among them, a i Let m represent the i-th type of environmental data related to mine safety, and m1 represent the total number of types of environmental data related to mine safety; b i This represents mechanical data related to mine safety, where m2 represents the total number of types of mechanical data related to mine safety; c i This represents location tag data related to mine safety; m3 represents the total number of types of mechanical data related to mine safety; d i This represents the i-th device parameter used to reflect the device status, and m4 represents the total number of device parameters. S12. Based on the mine environment dataset, mine mechanics dataset, mine positioning tag dataset, and equipment status parameter set, deploy sensors at key locations in the mine. S13. Register the locations of the deployed sensors to three-dimensional coordinates to obtain a sensor topology map with spatial attributes; use a mobile laser scanner to scan the mine to obtain a three-dimensional point cloud model of the mine. S14. Set the response action for a Level 1 warning to stop the power supply of the relevant equipment; Set the similarity threshold to u; set the response action for the level 2 warning to match historical cases from the risk knowledge base, and use the solutions of historical cases with similarity ≥ similarity threshold u to resolve the issue. If no historical cases with similarity ≥ similarity threshold u exist, generate a risk report and send it to the safety officer, who will then conduct risk analysis and make a decision. The response actions for the three-level early warning system are to obtain real-time personnel location and real-time three-dimensional stress cloud map, and to plan personnel escape routes.

3. The method for monitoring digital mine safety production based on the Internet of Things according to claim 1, characterized in that, S2 includes the following steps: S21. Improve the traditional LoRaWAN to obtain the improved LoRaWAN; S22. Construct a CNN sensor data prediction model, collect historical data and combine it with optimization algorithms to adjust the parameters of the CNN sensor data prediction model, and obtain an optimized CNN sensor data prediction model.

4. The method for monitoring digital mine safety production based on the Internet of Things according to claim 3, characterized in that, S21 includes the following steps: S211. Adjust the SF value according to the real-time mine signal-to-noise ratio to balance the relationship between roadway attenuation and electromagnetic interference; S212. To address the conflict issue caused by high-density deployments within mines, a time-series allocation algorithm is used to divide time frames into fixed time slots and contention windows, prioritizing the propagation of abnormal data. The time slot allocation algorithm formula is as follows: Where T represents the time slot length, N represents the number of active nodes, P represents the received power, i.e. the signal power received by the node, and Q represents the noise power spectral density. S213. The node dynamically selects the next hop based on the routing cost function, combined with the link quality and remaining power in the mine; the routing cost function formula is as follows: Where C represents the routing cost, α, β and λ are all weighting coefficients, RSSI represents the received signal strength indicator, PLR represents the packet loss rate, and ER represents the remaining battery power; S214. An improved LoRaWAN is obtained through S211, S212, and S213.

5. The method for monitoring digital mine safety production based on the Internet of Things according to claim 3, characterized in that, S22 includes the following steps: S221. Construct a CNN sensor data prediction model and set the initial learning rate of the CNN sensor data prediction model to p1. S222. Collect historical normal operating condition data and historical abnormal operating condition data of mine sensors to obtain training data; collect abnormal mine data corresponding to level 1, level 2, and level 3 early warnings to obtain test data. S223. Set the training error threshold q1 and the maximum number of training iterations; use the training data to repeatedly train the CNN sensor data prediction model; after each round of training, calculate the training error q2 of the CNN sensor data prediction model and adjust the initial learning rate according to the training error; When q2≤q1 or the maximum number of iterations is reached, a well-trained CNN sensor data prediction model with a learning rate of q2 is obtained. S224. Test the trained CNN sensor data prediction model using test data. After the test is completed, an optimized CNN sensor data prediction model is obtained.

6. The method for monitoring digital mine safety production based on the Internet of Things according to claim 5, characterized in that, S224 includes the following steps: S2241. Set the test accuracy threshold r1, and use the test data to test the trained CNN sensor data prediction model to obtain the test accuracy r2. S2242. If r2≥r1, then the trained CNN sensor data prediction model is used as the optimized CNN sensor data prediction model; otherwise, the Hippo algorithm is used to find the learning rate of the trained CNN sensor data prediction model to obtain the global optimal solution. The global optimal solution is used as the learning rate of the trained CNN sensor data prediction model to obtain the optimized CNN sensor data prediction model.

7. The method for monitoring digital mine safety production based on the Internet of Things according to claim 6, characterized in that, The steps in S2242, which use the Hippo algorithm to find the learning rate of the trained CNN sensor data prediction model and obtain the globally optimal solution, include the following: S22421. Construct a hippopotamus population w, and set the size of the hippopotamus population as z. Then the hippopotamus population w = {w1, w2, ..., w...} i ,...,w z }, where w i This represents the i-th hippopotamus in the hippopotamus population; S22422. Randomly generate an initial position group of the hippopotamus population x = {x1, x2, ..., x} based on the learning rate. i ,...,x z }, where x i This represents the initial position of the i-th hippopotamus in the hippopotamus population; S22423. Start the iteration operation and set the maximum number of optimization iterations to ε. During each iteration, update the position of each hippo in the initial position set of the hippo population according to the test accuracy. And during each iteration, obtain the global best hippo individual and the global best target in the hippo population. S22424. When r2≥r1 or the maximum number of iterations for optimization is reached, the hippopotamus population stops the iteration operation and obtains the global optimal solution.

8. A digital mine safety production monitoring system based on the Internet of Things (IoT), used to implement the digital mine safety production monitoring method based on the IoT as described in any one of claims 1-7, characterized in that, The system includes a data acquisition and initialization module, a communication and prediction model optimization module, a real-time data processing and anomaly detection module, and a risk assessment and early warning triggering module.

Images