Intelligent mine management system based on Internet of Things technology

Abstract

The invention discloses a smart mine management system based on the Internet of Things technology, and relates to the technical field of smart mines, the system is composed of a plurality of function modules, and the system comprises a data acquisition module which divides a mine area into a plurality of sub-areas, collects original sensor data in real time through an Internet of Things sensor network, and sends the data to a server; the original sensor data comprises environment data, equipment state data and personnel positioning data of each sub-region; preprocessing the original sensor data, and mapping the original sensor data to corresponding geographic units based on the space coordinates of the sub-regions; the topology construction module is used for mapping each sub-region into a topology node, and constructing a weighted topology edge between adjacent topology nodes according to a mine physical structure, equipment relevance and a personnel activity path to form a dynamic topology network; and the data analysis module is used for processing the sensor data through a preset fusion analysis model and calculating the safety score of each sub-region.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent mines, and particularly to an intelligent mine management system based on Internet of Things technology. Background Art

[0002] By monitoring the mine environment (such as gas concentration, temperature, humidity, etc.) and the operating status of equipment in real time, potential safety hazards can be discovered in a timely manner and corresponding measures can be taken to effectively prevent accidents. Using data analysis technology to deeply analyze information such as mineral resource distribution and mining progress helps managers make more scientific and reasonable decisions, optimize resource allocation, and reduce costs; by means of an automated control system, human operation errors are reduced, the working efficiency of production equipment is improved, and precise production scheduling also helps to shorten non-production time and improve the overall production capacity; around the mining area, monitoring environmental changes, evaluating the impact of mining activities on the ecological environment, supporting the implementation of the concept of green mining, and promoting the sustainable development of mining enterprises.

[0003] Traditional mine management systems mostly rely on manual inspections and discrete monitoring methods, and usually independently collect parameters such as vibration, gas, and equipment status. Lack of spatio-temporal correlation analysis of multi-source data (environment, equipment, personnel) leads to fragmented risk perception and difficulty in discovering complex safety hazards in a timely manner; dividing the monitoring area based on a fixed physical structure ignores the dynamic propagation of ore body stress, the linkage relationship of equipment, and the impact of personnel flow paths on risk conduction, and it is impossible to construct a topological network reflecting the real-time state of the mine, resulting in a lag in risk assessment behind actual changes; mostly relying on a preset rule base and not dynamically adjusting in combination with real-time risk evolution, it is difficult to optimize the resource scheduling path and priority, and problems such as uneven distribution of rescue resources and conflicts between evacuation routes and risk diffusion directions often occur. Summary of the Invention

[0004] (I) Technical Problems to be Solved In view of the deficiencies of the prior art, the present invention provides an intelligent mine management system based on Internet of Things technology, which solves the problems raised in the background art.

[0005] (II) Technical Solutions To achieve the above objectives, the present invention is realized through the following technical solutions: An intelligent mine management system based on Internet of Things technology, comprising: A data acquisition module, which divides the mine area into several sub-areas, and real-time collects original sensor data through an Internet of Things sensor network. The original sensor data includes environmental data, equipment status data, and personnel positioning data of each sub-area; preprocesses the original sensor data, and maps the original sensor data to the corresponding geographical unit based on the spatial coordinates of the sub-area; Topology construction module, which maps each sub-region to topological nodes and constructs weighted topological edges between adjacent topological nodes according to the physical structure of the mine, equipment correlation and personnel activity paths to form a dynamic topological network; Data analysis module, which processes sensor data through a preset fusion analysis model, calculates the safety score of each sub-region, and maps the safety score to the eigenvalue of the corresponding topological node; Topology update module, which updates the eigenvalue of the topological node based on time-series sensor data, combines the mine geological parameters and equipment operation status, and constructs a dynamic update model of the topological structure; Scheduling and planning module, which optimizes the algorithm to generate a resource scheduling plan and an emergency response strategy according to the real-time topological network eigenvalue.

[0006] Furthermore, the data acquisition module: Deploy vibration sensors, gas sensors, temperature and humidity sensors and positioning terminals to collect ore body vibration intensity, gas concentration, temperature and humidity data and personnel location information respectively; Normalize the collected data to generate a multi-dimensional data matrix.

[0007] Furthermore, the calculation of the weight of the topological edge includes: Obtain the ore body stress correlation degree, equipment linkage frequency and personnel flow density of adjacent sub-regions, and calculate the topological edge weight. The formula is: ; Wherein, is the topological edge weight; is the ore body stress correlation degree; is the equipment linkage frequency; is the personnel flow density; and are the environmental risk values of different sub-regions respectively; is the total area of the sub-region; 、 、 and are adjustable parameters respectively; is the maximum personnel flow density.

[0008] Furthermore, the fusion analysis model of the data analysis module includes: Extract the spatial distribution characteristics in the sensor data through a convolutional neural network to identify the abnormal vibration area of the ore body, the gas concentration gradient and the temperature and humidity hot spot; at the same time, use the long short-term memory network to capture the time-series change law of the sensor data and predict the ore body stress accumulation trend and equipment failure cycle; Perform cross-modal fusion of the spatial distribution characteristics and temporal variation characteristics to generate a comprehensive risk characterization vector for sub-regions; train the fusion analysis model based on the non-linear error weighting term between the predicted safety score and the true safety score, the consistency constraint term of the scores of adjacent sub-regions, and the dynamic weight balance mechanism.

[0009] Further, the non-linear error weighting term: Enhance the penalty for large errors through an exponential function, and the score consistency constraint term dynamically adjusts the difference tolerance of the scores of adjacent sub-regions according to the confidence of sensor data; output the real-time safety score of each sub-region, with the score range of 0, 1, and map the score result to the eigenvalue of the topological node to drive the dynamic update of the topological network.

[0010] Further, the dynamic update model is: Calculate the eigenvalue of the node at the next moment based on the historical eigenvalue of the topological node, the interaction of adjacent nodes, and the environmental risk driving factor; The continuous contribution of the historical eigenvalue is dynamically adjusted by the first update weight; the interaction of adjacent nodes is quantified by the ore body stress propagation interaction factor, which characterizes the influence of the stress difference and coupling effect between adjacent sub-regions on the eigenvalue; the environmental risk driving factor models the direct influence of the environmental risk value of the sub-region itself on the eigenvalue, and uses a logarithmic function to enhance the sensitivity of low-risk regions.

[0011] Further, iteratively optimize the learning rate parameters in the update weight and the interaction factor through the gradient descent algorithm to make the evolution trend of the topological eigenvalue output by the model match the historical sensor data; Predict the spatio-temporal propagation path of the mine risk according to the updated topological network eigenvalue, and output the risk level evolution map of each sub-region within a preset future time period.

[0012] Further, the scheduling and planning module includes: Generate a device maintenance path, a personnel evacuation route, and an emergency resource allocation plan based on the real-time topological eigenvalue, and conduct an evaluation. The formula is: ; Among them, is the evaluation value; R is the total number of moments; is the cumulative value of the path risk; is the total time; is the total distance; , and are the weights respectively, satisfying and .

[0013] (III) Beneficial effects The present invention provides an intelligent mine management system based on Internet of Things technology, having the following beneficial effects: (1) This solution integrates environmental data (vibration, gas, temperature and humidity), equipment status (operating parameters, fault signals) and personnel positioning information through an Internet of Things sensor network, realizes spatio-temporal correlation analysis of multi-modal data, breaks through the limitations of traditional single-dimensional monitoring, and accurately identifies sudden changes in ore body stress, equipment chain failures and environmental anomaly coupling risks; constructs a dynamic topology network based on the physical structure of the ore body, equipment linkage logic and personnel flow paths, and quantifies the risk conduction ability by real-time updating node eigenvalue and edge weight (such as stress correlation degree, equipment cooperation frequency), dynamically reflects the spatio-temporal evolution trend of the mine safety status, and solves the problem of static model lag.

[0014] (2) This solution uses a CNN-LSTM hybrid model to extract spatial features and perform time-series dependence analysis on sensor data, combines the non-linear error weighting and scoring consistency constraint mechanisms in the loss function, significantly improves the prediction accuracy of safety scores in complex noise environments, and reduces the false alarm rate by more than 40%; generates equipment maintenance, personnel evacuation and resource allocation plans by dynamically updating the model to predict the risk propagation path and combining multi-objective optimization algorithms, realizes the dynamic matching of emergency strategies and real-time risk evolution, and has a full-process closed-loop automation design from data collection, topology construction to scheduling planning, reduces manual decision-making links, and avoids misjudgments caused by insufficient experience or information lag. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 is a schematic diagram of the system flow of the present invention; Figure 2 is a schematic diagram of the overall method of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0016] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention. Embodiment

[0017] Please refer to Figure 1 , this embodiment provides an intelligent mine management system based on Internet of Things technology, and the system includes: The data acquisition module divides the mining area into several sub-areas, and collects raw sensor data in real time through the Internet of Things sensor network. The raw sensor data includes environmental data, equipment status data, and personnel positioning data of each sub-area; preprocesses the raw sensor data, and maps the raw sensor data to the corresponding geographical unit based on the spatial coordinates of the sub-area; The topology construction module maps each sub-area to a topology node, and constructs weighted topology edges between adjacent topology nodes according to the physical structure of the mine, equipment correlation, and personnel activity paths to form a dynamic topology network; The data analysis module processes the sensor data through a preset fusion analysis model, calculates the safety score of each sub-area, and maps the safety score to the eigenvalue of the corresponding topology node; The topology update module updates the eigenvalue of the topology node based on the time-series sensor data, and constructs a dynamic update model of the topology structure in combination with the mine geological parameters and equipment operation status; The scheduling and planning module optimizes the algorithm to generate a resource scheduling plan and an emergency response strategy according to the real-time topology network eigenvalue; Data acquisition module: Deploy vibration sensors, gas sensors, temperature and humidity sensors, and positioning terminals to collect ore body vibration intensity, gas concentration, temperature and humidity data, and personnel position information respectively; Normalize the collected data to generate a multi-dimensional data matrix; Sensor deployment: Deploy the following sensors in each sub-area: Three-axis vibration sensor: Installed on the key support structure of the roadway, with a sampling frequency of 100Hz, monitoring rock burst microseismic signals; Multi-gas detector: Real-time collection of gas (CH 4 ), carbon monoxide (CO) concentration, accuracy ±0.1ppm; Temperature and humidity sensor: Record data every 30 seconds, range -20°C to 60°C, accuracy ±0.5°C; Equipment current monitoring module: Integrated in the conveyor belt motor, collecting current fluctuations, temperature, and vibration frequency.

[0018] Data preprocessing: Timestamp alignment: Align multi-source data by the edge computing node according to a 5-second time window; Noise filtering: Use wavelet transform to remove mechanical interference noise in the vibration signal; The calculation of the weight of the topology edge includes: Obtain the ore body stress correlation degree, equipment linkage frequency, and personnel flow density of adjacent sub-areas, calculate the topology edge weight, and the formula is: ; Among them, is the topological edge weight; is the stress correlation degree of the ore body; is the equipment linkage frequency; is the personnel flow density; and are the environmental risk values of different sub-regions respectively; is the total area of the sub-region; , , and are adjustable parameters respectively; is the maximum personnel flow density; Equipment linkage frequency: The equipment linkage frequency is calculated by counting the number of collaborative operations of equipment in adjacent sub-regions within a preset time window. Specifically: when the equipment operation parameters (such as current, vibration frequency) of sub-region u and the equipment parameters of sub-region u+1 synchronously meet the threshold conditions (such as current exceeding the limit and vibration frequency suddenly rising) within the time window (such as ±5 seconds), it is recorded as one linkage event; after accumulating the number of events and dividing by the length of the time window (such as 1 hour), the linkage frequency is obtained; It should be noted that: although high-frequency linkage reflects strong equipment collaboration, it will also increase the vulnerability of the system (such as a single equipment failure is likely to trigger a chain reaction); therefore, it is necessary to balance the risk conduction ability through weight adjustment; when increases, rapidly approaches 0, and the denominator term is approximately 1. At this time, the stress correlation degree of the ore body in the formula dominates the weight, and the influence of the equipment linkage frequency weakens; this avoids excessive weight inflation in high-frequency linkage scenarios, resulting in distorted risk assessment; Sub-region division and data collection: Basis for division: The mine area is divided into several sub-regions according to the physical structure of the roadway. Each sub-region covers a roadway section about 50 meters long, and it is ensured that adjacent sub-regions are connected through the main roadway or ventilation shaft; Sensor deployment: The following Internet of Things devices are deployed in each sub-region: Ore body stress sensor: Installed on the roof and side walls of the roadway to monitor rock microseismic signals and stress changes in real time; Equipment status monitoring terminal: Integrated into key equipment such as conveyor belts and ventilators to collect parameters such as operating current, temperature, and vibration frequency; Personnel positioning beacon: Through the smart bracelets worn by miners, the position is tracked in real time using UWB (Ultra-Wideband) technology with an accuracy of 0.3 meters; Topological node mapping and edge construction: Node Generation: Map each sub-region to a node in the topological network. Node attributes include: Static attributes: roadway gradient, rock stratum type, equipment type; Dynamic attributes: real-time stress value, equipment operation status (normal / warning / fault), personnel density; Topological Edge Construction Rules: Physical connectivity: If two sub-regions are directly connected by a roadway, a topological edge is automatically constructed; Equipment linkage: If there is a material transportation dependency between the conveyor belt in sub-region A and the crusher in sub-region B, an equipment linkage edge is constructed; Personnel flow path: Based on historical positioning data, if the personnel movement frequency from sub-region C to D exceeds 10 times per hour, a personnel flow edge is constructed; Dynamic Topological Network Update: Update Trigger Conditions: Regular update: Recalculate the edge weights every 5 minutes; Event-driven update: When a device failure or gas concentration exceeds the limit occurs in a certain sub-region, immediately update the associated edge weights; Visualization display: Display the dynamic topological network through a 3D digital twin platform. The edge weights are represented by color gradients (green → red) to indicate the risk conduction intensity, and the node sizes reflect the real-time personnel density; The fusion analysis model of the data analysis module includes: Extract the spatial distribution features in the sensor data through a convolutional neural network to identify abnormal ore body vibration regions, gas concentration gradients, and temperature and humidity hotspots; at the same time, use a long short-term memory network to capture the temporal variation laws of the sensor data and predict the ore body stress accumulation trend and equipment failure cycle; Perform cross-modal fusion on the spatial distribution features and temporal variation features to generate a comprehensive risk characterization vector for the sub-region; train the fusion analysis model based on the non-linear error weighting term between the predicted safety score and the real safety score, the scoring consistency constraint term between adjacent sub-regions, and the dynamic weight balance mechanism; Non-linear error weighting term: Enhance the penalty for large errors through an exponential function. The scoring consistency constraint term dynamically adjusts the difference tolerance of the scores between adjacent sub-regions according to the confidence of the sensor data; output the real-time safety score for each sub-region, with the score range of 0, 1, and map the scoring results to the feature values of the topological nodes to drive the dynamic update of the topological network; The dynamic update model is: Calculate the node feature values at the next moment based on the historical feature values of the topological nodes, the interaction between adjacent nodes, and the environmental risk driving factors; The continuity contribution of the historical eigenvalue is dynamically adjusted by the first update weight; the interaction between adjacent nodes is quantified by the ore body stress propagation interaction factor, which characterizes the influence of stress differences and coupling effects between adjacent sub-regions on the eigenvalue; the environmental risk driving factor models the direct impact of the sub-region's own environmental risk value on the eigenvalue, and uses the logarithmic function to enhance the sensitivity of low-risk regions; The learning rate parameters in the update weight and interaction factor are iteratively optimized through the gradient descent algorithm to make the model output match the evolution trend of the topological eigenvalue of the historical sensor data; Based on the updated topological network eigenvalue, predict the spatio-temporal propagation path of mine risks, and output the risk level evolution map of each sub-region within a preset future time period; The scheduling and planning module includes: Based on the real-time topological eigenvalue, generate the equipment maintenance path, personnel evacuation route and emergency resource allocation plan, and conduct an evaluation. The formula is: ; Where, is the evaluation value; R is the total number of moments; is the cumulative value of path risk; is the total time; is the total distance; , and are the weights respectively, satisfying and ; For example: Abnormal detection in the rock burst warning and equipment linkage: The vibration sensor in sub-region P detects a high-frequency microseismic signal (>200Hz), and the CNN identifies that its spatial distribution is concentrated on the roadway roof; the LSTM analysis shows that the current of the conveyor belt motor fluctuates and increases by 30% within 10 minutes.

[0019] Score generation: The output safety score of the fusion analysis model = 0.25 (red alert), and the eigenvalue = 0.75.

[0020] Topological linkage: According to the eigenvalue, the topological update module increases the edge weight between sub-region P and adjacent region Q to 0.9 (originally 0.5), reflecting that the rock burst risk may spread along the high-stress correlation path; The scheduling and planning module immediately generates instructions: 1. Shut down the conveyor belt equipment in sub-region P and its associated areas; 2. Evacuate the personnel in this area to the preset safe path (avoiding area Q); 3. Start the support robot to reinforce the roof.

[0021] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. Those of ordinary skill in the art will recognize that the units and algorithm steps of the examples described in connection with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in hardware or software depends on the specific application and design constraints of the technical solution.

[0022] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units. They may be located in one place or distributed over multiple network units. Some or all of the units can be selected according to actual needs to achieve the objectives of the solution of this embodiment.

[0023] As described above, the above are only specific embodiments of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered within the protection scope of the present application.

Claims

1. A smart mine management system based on Internet of Things technology, characterized by: The system includes: The data acquisition module divides the mine area into several sub-areas and collects raw sensor data in real time through the IoT sensor network. The raw sensor data includes environmental data, equipment status data, and personnel location data of each sub-area. The raw sensor data is pre-processed and mapped to the corresponding geographic unit based on the spatial coordinates of the sub-area. The topology construction module maps each sub-area into a topological node, and constructs weighted topological edges between adjacent topological nodes according to the mine's physical structure, equipment relevance, and personnel activity paths to form a dynamic topological network; The data analysis module processes the sensor data through a preset fusion analysis model, calculates the safety score of each sub-area, and maps the safety score to the characteristic value of the corresponding topological node; The topology update module updates the characteristic values ​​of topological nodes based on time-series sensor data and builds a dynamic update model of the topological structure by combining mining geological parameters and equipment operating status; The scheduling and planning module generates resource scheduling plans and emergency response strategies based on the real-time topology network feature values ​​and optimization algorithms.

2. According to claim 1, a smart mine management system based on Internet of Things technology is characterized by: Data acquisition module: Deploy vibration sensors, gas sensors, temperature and humidity sensors and positioning terminals to collect ore body vibration intensity, gas concentration, temperature and humidity data and personnel location information; The collected data are normalized to generate a multidimensional data matrix.

3. According to claim 1, a smart mine management system based on Internet of Things technology is characterized by: The weight calculation of topological edges includes: Obtain the stress correlation degree of the ore body, equipment linkage frequency and personnel flow density of adjacent sub-areas, and calculate the topological edge weight. The formula is: ； in, is the topological edge weight; is the stress correlation of the ore body; is the device linkage frequency; is the density of personnel flow; and are the environmental risk values ​​of different sub-regions; is the total area of ​​the sub-region; , , and They are adjustable parameters respectively; The maximum flow density of people.

4. According to claim 3, a smart mine management system based on Internet of Things technology is characterized in that: The fusion analysis model of the data analysis module includes: The spatial distribution characteristics of sensor data are extracted through convolutional neural networks to identify abnormal vibration areas of the ore body, gas concentration gradients, and temperature and humidity hot spots. At the same time, the long-short-term memory network is used to capture the temporal variation of sensor data and predict the stress accumulation trend of the ore body and the equipment failure cycle. The spatial distribution features and the temporal variation features are cross-modally fused to generate a comprehensive risk characterization vector for the sub-region; the fusion analysis model is trained based on the nonlinear error weighting term between the predicted safety score and the actual safety score, the consistency constraint term of the scores of adjacent sub-regions, and the dynamic weight balancing mechanism.

5. According to claim 4, a smart mine management system based on Internet of Things technology is characterized in that: Nonlinear error weighting term: The penalty for large errors is enhanced by an exponential function, and the score consistency constraint dynamically adjusts the tolerance for differences in scores of adjacent sub-areas according to the confidence of sensor data. The real-time safety score of each sub-area is output, with a score range of 0 and 1, and the score result is mapped to the characteristic value of the topological node to drive the dynamic update of the topological network.

6. The intelligent mine management system based on Internet of Things technology according to claim 4 is characterized in that: The dynamic update model is: Based on the historical eigenvalues ​​of topological nodes, the interactions between adjacent nodes, and environmental risk driving factors, the node eigenvalues ​​at the next moment are calculated; The continuity contribution of the historical characteristic value is dynamically adjusted by a first update weight; The interaction between adjacent nodes is quantified by the interaction factor of ore body stress propagation, which characterizes the influence of stress difference and coupling effect between adjacent sub-regions on the eigenvalue; The environmental risk driving factor is modeled by the direct impact of the sub-region's own environmental risk value on the characteristic value, and the logarithmic function is used to enhance the sensitivity of the low-risk area.

7. The intelligent mine management system based on Internet of Things technology according to claim 6 is characterized in that: The learning rate parameters in the update weights and interaction factors are iteratively optimized through the gradient descent algorithm to make the model output match the evolution trend of the topological feature values ​​of the historical sensor data; According to the updated topological network characteristic values, the spatiotemporal propagation path of mine risks is predicted, and the risk level evolution map of each sub-area in the future preset time period is output.

8. The intelligent mine management system based on Internet of Things technology according to claim 1 is characterized by: The scheduling and planning module includes: Based on the real-time topology feature values, equipment maintenance paths, personnel evacuation routes, and emergency resource allocation plans are generated and evaluated. The formula is: ； in, is the evaluation value; R is the total number of moments; is the cumulative value of path risk; is the total time; is the total distance; , and are weights respectively, satisfying and .

Images