Coal mine collapse monitoring system and method

Abstract

The invention belongs to the technical field of coal mining, and discloses a coal mine collapse monitoring system and method, and the method comprises the steps: a deformation monitoring module carries out the deformation monitoring of a coal mine; the displacement stress monitoring module monitors rock mass displacement, stress state and strain change of a coal mine goaf or a potential collapse area; the underground water monitoring module measures the change of the underground water level in real time; the route planning module plans a cruise route of the unmanned aerial vehicle; the image acquisition module performs low-altitude flight shooting on a coal mine area; the image preprocessing module preprocesses the scanned image; the abnormity identification module performs feature extraction on the preprocessed image and identifies the abnormal condition of coal mine collapse; the comprehensive evaluation module performs multi-source data fusion and establishes a multi-parameter early warning model; when the monitoring data exceeds a preset threshold value, the alarm module triggers an early warning signal. Through multiple monitoring means, the functions of real-time continuous monitoring, intelligent analysis and evaluation and the like are achieved, and potential collapse risks can be found in time.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of coal mining, more particularly, to a coal mine collapse monitoring system and method. BACKGROUND

[0002] During the process of coal mining, collapse indeed occurs. When the underground coal seam is mined out, the roof strata of the goaf will bend and move downward under the action of its own gravity and the pressure of the overlying strata. When the tensile stress formed inside the roof strata exceeds its tensile strength limit, the roof will break and collapse. As the coal mining work progresses, the range of strata affected by mining expands continuously, eventually forming a much larger nearly elliptical collapse basin on the ground surface than the goaf. Collapse leads to ground subsidence, causing serious damage to land resources. Collapse changes the flow direction and flow rate of surface water bodies (rivers, lakes, wells, etc.), resulting in a decrease in the water volume of some ditches and springs or even drying up, affecting the normal production and life of local residents. Buildings in the collapse area may tilt, crack or even collapse due to unstable foundations, endangering the safety of residents' lives and property.

[0003] The prior art document with publication number CN114485555B provides a device for monitoring ground subsidence of a coal mine and a monitoring method. The device includes a base on the rock strata, a top cover on the base, a drive transmission structure, a displacement display screen and a monitoring terminal arranged at the top end of the base, a storage end connected to the displacement display screen, a drive transmission structure connected to a protective casing through a transmission line at the other end, a rock strata displacement probe sleeved on the outer circumference of the protective casing, a plurality of detection rods arranged on the outer circumferential surface of the rock strata displacement probe, a prism arranged on the surface of the top cover, and a solar cell panel arranged on the upper end surface of the top cover. The device realizes monitoring of each subsidence layer in the monitoring area, monitors the migration law of each rock strata above the goaf, monitors the ground subsidence of the coal mine, and enables users to intuitively understand and master the collapse conditions at each location in the area, so that protective measures can be taken for surface structures that may be affected by the collapse area or targeted treatment measures can be taken for the collapse area.

[0004] Although the prior art solution in the above can achieve the relevant beneficial effects through the structure of the prior art, it still has the following defects: 1. It cannot comprehensively monitor the collapse of the coal mine; 2. The prior art can only monitor and record the collapse that has occurred, but cannot predict and discover potential collapse risks. For areas with complex geological conditions and frequent mining activities, the prediction ability of the prior art is limited, and it is difficult to accurately assess the collapse risk.

[0005] In view of this, we propose a coal mine collapse monitoring system and method. SUMMARY

[0006] Technical problems to be solved The purpose of the present application is to provide a coal mine collapse monitoring system and method, which solves the technical problems raised in the background art, realizes comprehensive monitoring of ground deformation, underground displacement, stress change, groundwater conditions and other multi-source information in the coal mine area, realizes real-time continuous monitoring, intelligent analysis and evaluation through the integration of various monitoring means, and can timely discover potential collapse risks, thereby realizing early warning and timely response, reducing or avoiding personnel casualties and property losses.

[0007] 2. Technical solution The technical solution of the present application provides a coal mine collapse monitoring system, which comprises: Data acquisition module: collect comprehensive records of the ground conditions in the coal mine area, including topography, landform, vegetation coverage, building distribution, etc., and potential collapse risk areas. Collect images of the coal mine, including normal images and abnormal situation images, and label the images as reference samples; Deformation monitoring module: including leveling and tiltmeter monitoring; deformation monitoring of the coal mine.

[0008] Leveling: using leveling instruments such as level and laser level, periodically measuring the elevation of the coal mine area, and evaluating the collapse risk by monitoring the change of ground subsidence.

[0009] Tiltmeter monitoring: monitor the inclination change of the ground surface by installing a tiltmeter, and then analyze the possibility of collapse.

[0010] Displacement stress monitoring module: includes inclinometer, displacement meter, stress meter and strain meter, installs inclinometer and displacement meter in the coal mining area or potential collapse area, which is used to monitor the rock mass displacement, stress state and strain change of the coal mining area or potential collapse area. To evaluate the stability of rock mass and collapse risk.

[0011] Groundwater monitoring module: arrange water level monitoring wells in the coal mine area, measure the change of groundwater level in real time, analyze the influence of groundwater on the stability of coal mine. By monitoring the water quality change of groundwater, such as sand content, pH value and other parameters, evaluate the stability of groundwater system, indirectly reflect the risk of coal mine collapse.

[0012] Route planning module: plan the cruise route of the unmanned aerial vehicle; provide accurate flight route planning for the unmanned aerial vehicle cruise to ensure comprehensive coverage of the coal mine area.

[0013] Image acquisition module: including unmanned aerial vehicle, high-definition camera and laser scanner; the unmanned aerial vehicle carries laser radar to scan the coal mine; the unmanned aerial vehicle carries laser scanner and other equipment to take pictures at low altitude, and obtains high-resolution image data.

[0014] Image preprocessing module: pre-process the acquired scan images, including filtering denoising, normalization, etc. Abnormality identification module: feature extraction is performed on the pre-processed images, and machine learning or deep learning algorithms are used to identify abnormal situations, such as ground cracks, collapse pits, etc. Laser scanning technology can quickly obtain ground elevation data and generate high-precision three-dimensional terrain models to identify abnormal situations of coal mine subsidence.

[0015] Comprehensive evaluation module: multi-source data fusion is performed, and the data obtained by various monitoring means and the monitoring results of the abnormality identification module are fused and processed, and multi-source information such as ground deformation, underground displacement, stress change, and groundwater conditions is comprehensively analyzed to establish a multi-parameter early warning model, improving the accuracy and reliability of subsidence early warning.

[0016] Alarm module: including alarm, when the monitoring data exceeds the preset threshold, the system automatically triggers the early warning signal, reminding relevant personnel to take measures.

[0017] Central control unit: network connection with data acquisition module, deformation monitoring module, displacement stress monitoring module, groundwater monitoring module, route planning module, image acquisition module, image preprocessing module, abnormality identification module, comprehensive evaluation module and alarm module.

[0018] Through the above technical scheme, by comprehensively monitoring the multi-source information such as ground deformation, underground displacement, stress change, and groundwater conditions in the coal mine area, the system can timely discover potential subsidence risks and automatically trigger early warning signals when the data exceeds the preset threshold, thereby realizing early warning and timely response, reducing or avoiding personnel casualties and property losses.

[0019] As an optional solution of the present application, the deformation monitoring module monitors the deformation of the coal mine, including the following steps: Determine the monitoring area: according to the geological conditions, mining history, and past subsidence records of the coal mine, determine the key monitoring area.

[0020] Lay out monitoring points: reasonably lay out leveling points and tiltmeter monitoring points in the monitoring area. The layout of monitoring points should consider the topography, geological conditions and monitoring requirements to ensure that the deformation of the coal mine can be fully reflected. Prepare high-precision level meters, laser level meters, tiltmeters and other monitoring equipment.

[0021] Leveling: establish a leveling network in the monitoring area, including control points with known elevation and observation points with to-be-measured elevation. The layout of the leveling network should meet the measurement accuracy requirements and facilitate subsequent data processing and analysis. Use the level or laser level to measure the elevation of the monitoring points according to the specifications and requirements of the leveling.

[0022] Inclinometer monitoring: Install inclinometers at monitoring points, ensure that the instruments are securely installed, correctly oriented, and appropriate monitoring parameters are set. Inclinometers automatically and continuously monitor the inclination changes of the ground surface and record data in real time.

[0023] Data processing and analysis: Organize and process the measurement data, calculate the elevation change of each monitoring point. Through the analysis of the elevation change, evaluate the subsidence of the coal mine area and judge whether there is a risk of collapse. Process and analyze the data monitored by the inclinometer, calculate the inclination change and inclination rate of the ground surface. Through the analysis of the inclination change and rate, evaluate the inclination deformation of the coal mine area and judge whether there is a risk of collapse.

[0024] As an optional solution of the present application, the displacement stress monitoring module monitors the rock mass displacement, stress state and strain change of the coal mine goaf or potential collapse area to evaluate the stability and collapse risk of the rock mass, including the following steps: Identify monitoring targets: Determine the specific areas to be monitored (such as coal mine goaf, potential collapse area) and monitoring parameters (displacement, stress, strain).

[0025] Design monitoring scheme: According to the geological conditions, monitoring requirements and budget, select appropriate sensor types and quantities, and plan the location of the points.

[0026] Survey the site: Conduct detailed survey of the monitoring area to determine the specific installation location, taking into account factors such as geological structure, rock type, groundwater level, etc.

[0027] Sensor installation: According to the monitoring requirements, select high-precision, harsh environment-resistant (such as high temperature, high humidity, high dust) sensors. Install displacement meters and strain gauges at key locations to measure the displacement changes of the rock mass in specific directions. Seal the sensors to prevent the intrusion of groundwater, dust, etc., while ensuring that the sensors are stable and not loose.

[0028] Data acquisition: Install data collectors at appropriate locations to collect sensor data. Establish data transmission channels to ensure that data can be transmitted to the monitoring center in real time or periodically. Install and configure monitoring software, set up data reception, processing, storage and alarm functions.

[0029] System debugging: Calibrate the installed sensors again to ensure accurate and reliable measurement data. Simulate various working conditions to test the stability and accuracy of the system and adjust and optimize the parameters.

[0030] Data analysis: Start the monitoring system and begin real-time or periodic data collection. Process and analyze the collected data to evaluate the displacement, stress state and strain change of the rock mass and predict possible collapse risks.

[0031] As an optional solution of the present application, the image acquisition module carries out laser scanning on the coal mine by a UAV carrying a laser radar; the low-altitude flight shooting on the coal mine area includes the following steps: Taking off the UAV: start the UAV in a safe area and fly at a low altitude according to a preset flight path.

[0032] Laser scanning: the laser scanner carried by the UAV continuously scans the coal mine area to obtain three-dimensional point cloud data of the ground surface.

[0033] Image shooting: at the same time, the high-definition camera on the UAV shoots high-resolution image data to record the detailed situation of the coal mine area.

[0034] Data return: return the collected data to the ground control station in real time or at a fixed time to ensure the integrity and safety of the data.

[0035] As an optional solution of the present application, the anomaly recognition module extracts features from the preprocessed images and uses machine learning or deep learning algorithms to recognize abnormal situations, including the following steps: Step one: feature extraction, including image feature extraction and point cloud feature extraction; Image feature extraction: use edge detection, texture analysis, shape recognition and other methods to extract features related to ground cracks, collapse pits and other features from the preprocessed images, including color, texture and shape; Point cloud feature extraction: extract terrain features such as elevation change, slope and curvature from laser point cloud data, which helps to identify abnormal situations such as ground collapse.

[0036] Step two: build a recognition model: according to the specific application scenario and data characteristics, select appropriate machine learning or deep learning algorithms such as convolutional neural network (CNN), etc. Use labeled abnormal samples (such as known crack, collapse pit image and point cloud data) as training set to train the algorithm, so that it can recognize similar abnormal situations. Use cross-validation and other methods to evaluate the performance of the model to ensure its accuracy and stability on unknown data.

[0037] Step three: anomaly detection and recognition: input the preprocessed images and point cloud data into the trained model for anomaly detection. Input the preprocessed real-time data into the model for fast and accurate anomaly detection. Analyze the anomaly detection results based on the point cloud data and image data of the laser scanning to determine the specific type and degree of the anomaly. Analyze the model output to determine whether there are ground cracks, collapse pits and other abnormal situations, and mark the abnormal area.

[0038] Prediction result integration integrated : P integrated = w image*P image +w point_cloud *P point_cloud ; abnormal probability calculation P anomaly : ; wherein P image and P point_cloud represent the abnormal detection output scores of the model on image data and point cloud data respectively. w image and w point_cloud represent the weights of image and point cloud data. P integrated is the prediction result of the integrated image and point cloud data. a and b are parameters used to adjust the output in the sigmoid function. P anomaly is the predicted probability of abnormality.

[0039] The abnormal threshold determination D is carried out using the following formula: D = {x | x > q}; wherein q is the threshold value of abnormal detection. q can be optimized based on the performance on the validation set. D is the binary determination result of abnormality (for example, exceeding the threshold value is abnormal). x is the specific value of the abnormal probability or score output by the model, that is, P anomaly . All the abnormal probability or score x output by the model is determined as abnormal (D is true) if it is greater than the threshold value q, otherwise it is determined as normal (D is false).

[0040] Step four: monitoring report generation: according to the results of abnormal detection, a detailed monitoring report is generated, including abnormal position, type, scale and other information, which provides a scientific basis for the safety management and governance work of coal mine.

[0041] Step five: generating three-dimensional terrain model by laser scanning: using laser point cloud data, a high-precision three-dimensional terrain model is generated through three-dimensional modeling software. The model can intuitively show the terrain features of the coal mine area, which helps to more accurately identify and analyze abnormal situations. Combine the three-dimensional model with the abnormal detection results to conduct more in-depth spatial analysis and risk assessment.

[0042] As an optional scheme of the present application, the comprehensive evaluation module fuses the data obtained by various monitoring means and the monitoring results of the abnormal identification module, and comprehensively analyzes multi-source information such as ground deformation, underground displacement, stress change, underground water condition, etc., to evaluate the coal mine collapse and the potential collapse risk, including the following steps: 1. Data collection: collect multi-source data of various sensors, including ground deformation, underground displacement, stress change, underground water condition and monitoring results of the abnormal identification module and other multi-source information data; 2. Data preprocessing: including data cleaning, data calibration and data conversion; Data cleaning: remove outliers, missing values and unreasonable data.

[0043] Data calibration: Calibrate data from different sources to ensure data consistency and accuracy.

[0044] Data transformation: Converting raw data into a standard format that can be used for analysis.

[0045] 3. Data fusion processing: Ensure that all monitoring data are synchronized in time for simultaneous analysis. Match data from different monitoring points in space to form a complete monitoring network.

[0046] Compute time-synchronous interpolation (using linear interpolation of the time series): Data i time−sync (t)=Data i transformed (t1)+Data i transformed (t2)-Data i transformed (t1)×(t−t1) / (t2−t1);where t1 and t2 are the time coordinates of the original data points, and t is the time point to be interpolated. i time−sync (t) is the data of the i-th data source after time synchronization processing at time point t. i transformed (t1) is the data value of the i-th data source at the original time point t1. i transformed (t2) is the data value of the i-th data source at another original time point t2, where t2>t1. t is the time point at which interpolation is required.

[0047] Perform spatial matching resampling (using bilinear resampling): Data i spatial−match (x,y)=BilinearInterpolate(Data i time−sync ,x,y); where BilinearInterpolate is the bilinear interpolation function. i spatial−match (x,y) is the data of the i-th data source at the spatial coordinate (x,y) after spatial matching processing.

[0048] A weighted fusion algorithm is used for multi-source data: different weights are assigned to each data source based on its reliability and accuracy. Multi-source information such as ground deformation, underground displacement, stress changes, and groundwater conditions are used to assess coal mine subsidence and potential subsidence risks. ; where λ i is the decay rate of data source i (to consider the degree of new and old data, the decay rate λ i is a constant less than 1 but close to 1, used to achieve exponential decay effect.), t i is the data collection time. WeightedFusion(t) is the weighted fusion result at time point t. w i is the weight assigned to the i-th data source. λ i is the decay rate of the i-th data source, used for exponential weighted moving average. t i is the time point of data collection.

[0049] Risk assessment: According to the comprehensive data after weighted fusion, combined with historical data and expert experience, set reasonable warning threshold. When the monitoring data reaches or exceeds the warning threshold, trigger the warning mechanism, prompt the existence of coal mine collapse or potential collapse risk. When the monitoring data reaches or exceeds the warning threshold, trigger the warning mechanism. According to the warning result and actual situation, feedback and adjust the warning model. Continuously optimize the warning parameters and threshold settings, improve the adaptability and accuracy of the warning system.

[0050] 4、Comprehensive analysis: including ground deformation analysis, underground displacement analysis, stress change analysis and underground water condition analysis; Ground deformation analysis: analyze the deformation characteristics such as ground subsidence and uplift, evaluate the deformation rate and trend. Combined with geological conditions, judge whether the deformation is caused by coal mining.

[0051] Underground displacement analysis: monitor the moving direction and speed of underground rock strata, evaluate its influence on ground stability. Analyze the correlation between underground displacement and ground deformation.

[0052] Stress change analysis: monitor the stress change of coal seam and roof and floor rock strata, evaluate the mining stress distribution and evolution law. Analyze the influence of stress change on rock stability and collapse risk.

[0053] Underground water condition analysis: analyze the rise and fall of underground water level, water quality change, etc., evaluate its influence on rock stability and collapse risk. Combined with hydrogeological conditions, judge the correlation between underground water and collapse.

[0054] 5. Establish a multi-parameter early warning model: According to the comprehensive analysis results, determine the key indicators that can reflect the collapse risk (such as ground deformation rate, underground displacement, stress change, groundwater level change and monitoring results of abnormal identification module, etc.). According to historical data and expert experience, set reasonable early warning threshold for each early warning indicator. Adopt decision tree machine learning algorithm to build multi-parameter early warning model. The model should be able to consider the interaction and mutual influence between parameters, improve the accuracy and reliability of early warning.

[0055] 6. Evaluation and feedback: Real-time processing and analysis of monitoring data, timely evaluation of collapse risk. When the monitoring data reaches or exceeds the early warning threshold, trigger the early warning mechanism. According to the early warning results and actual situation, feedback and adjust the early warning model. Continuously optimize the early warning parameters and threshold settings, improve the adaptability and accuracy of the early warning system.

[0056] Through the above technical scheme, the comprehensive evaluation module can realize comprehensive and accurate evaluation of coal mine collapse and potential collapse risk, and provide strong guarantee for coal mine safety production.

[0057] The present application provides a kind of coal mine collapse monitoring method, comprising the following steps: S1, deformation monitoring module carries out deformation monitoring to coal mine.

[0058] S2, displacement stress monitoring module installs inclinometer and displacement meter in coal mine goaf or potential collapse area, monitors the displacement of rock mass, stress state and strain change in coal mine goaf or potential collapse area. To evaluate the stability and collapse risk of rock mass.

[0059] S3, groundwater monitoring module measures the change of groundwater level in real time, analyzes the influence of groundwater on coal mine stability.

[0060] S4, route planning module plans the cruise route of unmanned aerial vehicle; Ensure comprehensive coverage of coal mine area.

[0061] S5, image acquisition module carries high-definition camera and laser radar through unmanned aerial vehicle to carry out laser scanning on coal mine; unmanned aerial vehicle carries out cruise monitoring according to the planned route; unmanned aerial vehicle carries laser scanner and other equipment, carries out low-altitude flight shooting on coal mine area, and obtains high-resolution image data.

[0062] S6, image preprocessing module pre-processes the scanned images obtained, including filtering denoising, normalization, etc. S7, abnormal identification module extracts features from preprocessed images, and identifies abnormal conditions using machine learning or deep learning algorithms, monitors ground cracks, collapse pits and other features. Laser scanning technology can quickly obtain ground elevation data and generate high-precision three-dimensional terrain model, which is suitable for monitoring complex terrain. Identify abnormal conditions of coal mine collapse.

[0063] S8. The comprehensive assessment module performs multi-source data fusion, integrates the data obtained by various monitoring methods and the monitoring results of the anomaly identification module, comprehensively analyzes multi-source information such as ground deformation, underground displacement, stress changes, groundwater conditions, etc., establishes a multi-parameter early warning model, and improves the accuracy and reliability of collapse warning.

[0064] S9. When the monitoring data exceeds the preset threshold, the alarm module automatically triggers a warning signal to remind relevant personnel to take countermeasures.

[0065] 3. Beneficial effects One or more technical solutions provided in the technical solution of this application have at least the following technical effects or advantages: 1. By comprehensively monitoring multiple sources of information such as ground deformation, underground displacement, stress changes, and groundwater conditions in coal mine areas, the present invention enables the system to promptly detect potential collapse risks and automatically trigger an early warning signal when the data exceeds a preset threshold, thereby achieving early warning and timely response, reducing or avoiding casualties and property losses.

[0066] 2. The system integrates multiple monitoring methods, including leveling, inclinometer monitoring, displacement and stress monitoring, groundwater monitoring, and drone laser scanning. These methods each have unique advantages and complement each other to improve monitoring accuracy and comprehensiveness. At the same time, the application of drone technology makes monitoring complex terrain easier and more efficient.

[0067] 3. Through the comprehensive evaluation module, the system can integrate and process multi-source data, use machine learning or deep learning algorithms for intelligent analysis and evaluation, establish a multi-parameter early warning model, and improve the accuracy and reliability of collapse warnings.

[0068] 4. By integrating multiple monitoring methods, realizing real-time continuous monitoring, intelligent analysis and evaluation and other functions, it provides a strong guarantee for the safe production of coal mines and has significant social and economic benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 This is a flow chart of a coal mine subsidence monitoring method disclosed in a preferred embodiment of the present application; Figure 2 This is a structural diagram of a coal mine subsidence monitoring system disclosed in a preferred embodiment of the present application. DETAILED DESCRIPTION

[0070] The present application is further described in detail below with reference to the accompanying drawings.

[0071] Reference Figure 1 and Figure 2The embodiment of the application provides a coal mine collapse monitoring system, which comprises: The data acquisition module collects comprehensive records of the surface conditions of the coal mine area, including topography, landform, vegetation coverage, building distribution, etc., and potential collapse risk areas. Collect images of the coal mine, including normal images and abnormal situation images, and label the images as reference samples; label the topography, landform, vegetation, etc. of the coal mine area in the normal state as the baseline data. Detailed labeling is performed on the discovered surface cracks, collapse pits, ground subsidence, etc. including location, size, shape, etc. The collected image data is cleaned to remove duplicate, blurred or invalid images to ensure the quality of the reference samples. The reference samples are subjected to data enhancement through rotation, scaling, cropping, etc. to improve the generalization ability of the model.

[0072] The deformation monitoring module includes leveling and tiltmeter monitoring; the coal mine is subjected to deformation monitoring.

[0073] Leveling: using leveling instruments such as level and laser level, periodic elevation measurement of the coal mine area is performed to evaluate the collapse risk by monitoring the change of ground subsidence. This method is suitable for large area and flat terrain monitoring.

[0074] Tiltmeter monitoring: by installing a tiltmeter to monitor the inclination change of the surface, and then analyzing the possibility of collapse. The tiltmeter can be continuously monitored in real time, and is suitable for steep terrain monitoring.

[0075] Displacement stress monitoring module: includes inclinometer, displacement meter, stress meter and strain meter, inclinometer and displacement meter are installed in the coal mine goaf or potential collapse area, which is used to monitor the rock mass displacement, stress state and strain change of the coal mine goaf or potential collapse area. To evaluate the stability and collapse risk of rock mass. The stress state and strain change of the rock are measured by the stress meter and strain meter to understand the stress condition and stress concentration area of the rock, so as to evaluate the stability and collapse risk of the rock. These sensors can be installed in the underground rock stratum to realize real-time continuous monitoring.

[0076] Groundwater monitoring module: water level monitoring wells are arranged in the coal mine area to measure the change of groundwater level in real time, and analyze the influence of groundwater on the stability of the coal mine. Sudden drop or rise of groundwater level may be a precursor of collapse. By monitoring the water quality change of groundwater, such as sand content, pH value and other parameters, the stability of the groundwater system is evaluated, which indirectly reflects the risk of coal mine collapse. Abnormal change of groundwater level is often one of the precursors of collapse.

[0077] Route planning module: the cruising route of the unmanned aerial vehicle is planned; accurate flight route planning is provided for the unmanned aerial vehicle cruising to ensure comprehensive coverage of the coal mine area.

[0078] Image acquisition module: including unmanned aerial vehicle, high-definition camera and laser scanner; unmanned aerial vehicle carries laser radar to carry out laser scanning on coal mine; unmanned aerial vehicle carries laser scanner and other equipment to carry out low-altitude flight shooting on coal mine area, and high-resolution image data is obtained.

[0079] Image preprocessing module: pre-processing the obtained scanning image, including filtering denoising, normalization, etc. Abnormal identification module: feature extraction is carried out on the pre-processed image, and machine learning or deep learning algorithm is used to identify abnormal conditions, such as ground cracks and collapse pits. Laser scanning technology can quickly obtain ground elevation data and generate high-precision three-dimensional terrain model, which is suitable for monitoring complex terrain. The abnormal situation of coal mine collapse is identified.

[0080] Comprehensive evaluation module: multi-source data fusion is carried out, the data obtained by various monitoring means and the monitoring results of the abnormal identification module are fused and processed, and multi-source information such as ground deformation, underground displacement, stress change and underground water condition is comprehensively analyzed, a multi-parameter early warning model is established, and the accuracy and reliability of collapse early warning are improved.

[0081] Alarm module: including alarm, when the monitoring data exceeds the preset threshold, the system automatically triggers the early warning signal, reminding the relevant personnel to take measures.

[0082] Central control unit: network connection with data acquisition module, deformation monitoring module, displacement stress monitoring module, underground water monitoring module, route planning module, image acquisition module, image preprocessing module, abnormal identification module, comprehensive evaluation module and alarm module.

[0083] In this technical scheme, by comprehensively monitoring the multi-source information such as ground deformation, underground displacement, stress change and underground water condition of coal mine area, the system can timely find the potential collapse risk, and automatically trigger the early warning signal when the data exceeds the preset threshold, so as to realize early warning and timely response, reduce or avoid personnel casualty and property loss. Through the comprehensive evaluation module, the system can fuse and process multi-source data, use machine learning or deep learning algorithm for intelligent analysis and evaluation, establish multi-parameter early warning model, and improve the accuracy and reliability of collapse early warning. By integrating various monitoring means, realizing real-time continuous monitoring, intelligent analysis and evaluation and other functions, it provides strong guarantee for the safety production of coal mine, has obvious social and economic benefits.

[0084] Further, the deformation monitoring module monitors the deformation of the coal mine, including the following steps: Determine the monitoring area: according to the geological conditions of the coal mine, the mining history, the past collapse record and other factors, determine the key monitoring area.

[0085] Deploy monitoring points: Leveling and inclinometer monitoring points should be strategically located within the monitoring area. The placement of monitoring points should take into account topographical and geological conditions, as well as monitoring requirements, to ensure they fully reflect the deformation of the coal mine. Prepare high-precision levels, laser levels, inclinometers, and other monitoring equipment, and inspect their integrity and accuracy.

[0086] Leveling: Establish a leveling network within the monitoring area, including control points with known elevations and observation points whose elevations are to be measured. The layout of the leveling network should meet measurement accuracy requirements and facilitate subsequent data processing and analysis. Using a level or laser level, measure the elevations of the monitoring points in accordance with leveling specifications and requirements. Care should be taken to eliminate all sources of error during the measurement process to ensure the accuracy and reliability of the measured data.

[0087] Tiltmeter Monitoring: Install a tiltmeter at the monitoring point, ensuring it is securely mounted and oriented correctly, and setting appropriate monitoring parameters. The tiltmeter automatically and continuously monitors changes in the ground's tilt and records the data in real time. During monitoring, carefully check the instrument's operating status and data transmission to ensure the continuity and integrity of the monitored data.

[0088] Data Processing and Analysis: Measured data is collated and processed to calculate the elevation change at each monitoring point. This analysis assesses the subsidence of the coal mine area and determines whether there is a risk of collapse. Inclinometer data is processed and analyzed to calculate the surface tilt change and rate. This analysis assesses the deformation of the coal mine area and determines whether there is a risk of collapse.

[0089] Furthermore, the displacement and stress monitoring module monitors the displacement, stress state, and strain changes of the rock mass in the coal mine goaf or potential collapse area to assess the stability and collapse risk of the rock mass, including the following steps: Clarify monitoring objectives: determine the specific areas that need to be monitored (such as coal mine goafs, potential collapse areas) and monitoring parameters (displacement, stress, strain).

[0090] Design a monitoring plan: Based on geological conditions, monitoring requirements, and budget, select the appropriate sensor type (inclinometer, displacement meter, stress gauge, strain gauge) and quantity, and plan the location of the sensors.

[0091] Survey site: Conduct a detailed survey of the monitoring area to determine the specific installation location, taking into account factors such as geological structure, rock type, and groundwater level.

[0092] Sensor installation: Drill holes at selected locations, ensuring hole diameters meet sensor installation requirements. Prepare necessary tools, materials, and equipment for installation. Select high-precision, harsh environment-resistant (e.g., high temperature, high humidity, high dust) sensors based on monitoring needs. Fix inclinometers in drilled holes, ensuring measurement direction aligns with intended monitoring direction and performing preliminary calibration. Install displacement meters at key locations to measure rock mass displacement changes in specific directions. Embed stress and strain meters into rock, ensuring they accurately measure internal stress and strain changes.

[0093] Seal sensors to prevent groundwater, dust, and other intrusions, while ensuring sensor stability and tightness.

[0094] Data collection: Install data collectors at appropriate locations to collect sensor data. Establish data transmission channels to ensure real-time or periodic data transmission to the monitoring center. Install and configure monitoring software, setting up data reception, processing, storage, and alarm functions.

[0095] System debugging: Perform secondary calibration on installed sensors to ensure accurate and reliable measurement data. Simulate various working conditions to test system stability and accuracy, adjusting and optimizing parameters.

[0096] Data analysis: Start the monitoring system and begin real-time or periodic data collection. Process and analyze collected data to evaluate rock mass displacement, stress state, and strain changes, predicting potential collapse risks.

[0097] Further, the groundwater monitoring module measures changes in groundwater level in real time, analyzes the impact of groundwater on coal mine stability, and evaluates the stability of the groundwater system, including the following steps: Data collection: Collect basic data such as geological and hydrogeological maps, investigation reports, geological drilling, and hydrogeological drilling from land, water conservancy, and environmental protection departments. Collect large-scale construction project geotechnical engineering investigation data and reports from coal mine enterprises. Understand dynamic monitoring data such as groundwater level and quality in the evaluation area, as well as groundwater development and utilization. Collect drinking water source protection area division and pollution source distribution data provided by local environmental protection departments.

[0098] Site investigation: Conduct site investigation in the coal mine area to understand basic conditions such as topography, geological structure, and rock layer distribution. Identify potential hydrogeological problem areas, such as possible aquifers, faults, and karst.

[0099] Monitoring well site selection: Based on data collection and site investigation results, reasonably select and arrange monitoring wells in the coal mine area. Monitoring wells should be placed in or near aquifers as much as possible to accurately monitor groundwater level changes.

[0100] Drilling Operation: Conduct drilling operations according to design requirements to ensure that the drilling depth reaches the target aquifer.

[0101] During drilling, core data, water level changes, and other data should be recorded. Install appropriate well pipes and perform cementing treatment to prevent well wall collapse and groundwater pollution. Water level monitoring sensors and other necessary monitoring equipment should be installed in the well pipes. Debug the installed monitoring equipment to ensure accurate and stable monitoring of groundwater levels and water quality parameters.

[0102] Regular Monitoring: Regularly measure the water level and sample and analyze the water quality of the monitoring well. The monitoring frequency should be determined according to the actual situation, and it is generally recommended to monitor at least once a month. Accurately record the water level data and water quality parameters (such as sediment content, pH value, etc.) of each monitoring. Organize and store the monitoring data for subsequent analysis.

[0103] Data Analysis: Analyze the trend of groundwater level changes and identify abnormal changes (such as sudden drops or rises). Analyze the changes in water quality parameters and evaluate the stability of the groundwater system. Analyze the impact of groundwater changes on coal mine stability in combination with geological structures, mining activities, and other factors.

[0104] Calculate the water level change trend Ttrend: Ttrend = (∑ n t=2 ΔW level,t ) / (n−1); ΔW level,t = W level,t − W level,t-1；其中，n是观测次数。ΔWlevel,t is the water level change between the tth observation and the previous observation. W level,t is the water level at time point t. W level,t−1 is the water level at time point t−1, which is the water level of the previous observation. Ttrend is the average trend of water level changes, indicating the general direction and rate of water level changes during the entire monitoring period.

[0105] Identify abnormal changes A anomaly : A anomaly = ∑ n t=1 [(W level,t − W - level ) / σ level ] 2 ; where W level,t is the water level of the tth observation, Wˉ level is the average of all observed water levels, lσ level is the standard deviation of water level observations, and n is the number of observations. Abnormal changes can be identified by calculating the deviation of water level changes from the average water level. If the sum of squared deviations exceeds a certain threshold, it can be considered as an abnormal change.

[0106] Risk assessment: Based on monitoring data and analysis results, assess the stability of the coal mine area groundwater system. Identify potential collapse risk areas and factors.

[0107] Assess groundwater system stability S stability : S stability = (∑ n t=1 d stable,t ) / D total *100%; where d stable,t is the number of stable days recorded at time point t, d stable,t =1 if the water level change on the same day is within the acceptable range, otherwise 0. D total is the total number of monitoring days. S stability is the percentage of groundwater system stability, representing the proportion of stable days to the total number of monitoring days.

[0108] Risk assessment value R risk,i : R risk,i =w1*R geology,i +w2*R hydro,i +w3*R activity,i ; where R geology,i , R hydro,i and R activity,i are the risk scores of geology, hydrogeological conditions and mining activities for the i-th region, w1, w2, w3 are the weights of risk scores of geology, hydrogeological conditions and mining activities.

[0109] Potential collapse risk area identification R collapse : R collapse =∑ k i=1 R risk,i ; where R risk,i is the evaluation value of the i-th risk area, based on factors such as geology, hydrogeological conditions and mining activities. R collapse is the sum of potential collapse risk area identification, representing the overall risk of the entire coal mine area. k is the total number of risk areas.

[0110] For abnormal changes in groundwater level, develop corresponding measures such as strengthening drainage, grouting reinforcement, etc. Strengthen hydrogeological monitoring and early warning work in coal mine area, timely discover and handle potential safety hazards.

[0111] In this technical solution, effective monitoring and analysis of coal mine area groundwater level can be realized, providing scientific basis for assessing the impact of groundwater on coal mine stability, and providing strong support for preventing and responding to coal mine collapse and other geological disasters.

[0112] Further, the route planning module plans the cruise route of the unmanned aerial vehicle; including the following steps: 1. Requirement analysis and target setting: Determine the main purpose of the unmanned aerial vehicle cruise, such as monitoring the safety of coal mines, measuring terrain changes, etc. Clearly define the scope of the coal mine area that the unmanned aerial vehicle needs to cover, including boundaries, important monitoring points, prohibited flight areas, etc. Consider the endurance, flight speed, maximum height, load capacity of the unmanned aerial vehicle, etc. to ensure that the planned route is within the capabilities of the unmanned aerial vehicle.

[0113] 2. Data collection: Obtain high-precision topographic maps, satellite images or laser radar data of the coal mine area. Mark out buildings, towers, trees and other obstacles that may affect flight in the area. Obtain real-time or predicted weather data, including wind direction, wind speed, rainfall probability, etc. Obtain specific information about the coal mine: such as mine location, transportation routes, mining area boundaries, etc.

[0114] 3. Path planning: Grid-based algorithms: such as A* algorithm or Dijkstra algorithm, divide the area into grids and calculate the optimal path between each grid. Use heuristic search algorithms: such as genetic algorithm, particle swarm optimization algorithm, to find efficient paths in complex terrain. Adjust the route in real time during the flight of the unmanned aerial vehicle, such as avoiding sudden obstacles or optimizing flight altitude.

[0115] 4. Path generation and optimization: Generate a preliminary flight path that covers the entire coal mine area based on the algorithm. Adjust the path through the algorithm to ensure that the coal mine area is fully covered and repeated flights are minimized. Optimize flight altitude, speed and heading to maximize the endurance of the unmanned aerial vehicle. Ensure that the path avoids all known obstacles and dangerous areas.

[0116] 5. Path verification and simulation: Use FSX (Flight Simulator X) flight simulation software tools to simulate flight according to the planned path and check the feasibility and safety of the path. Evaluate key indicators such as flight time, energy consumption, coverage range, etc. and make necessary adjustments to the path.

[0117] 6. Real-time adjustment and monitoring: Equip the unmanned aerial vehicle with a real-time monitoring system, including GPS positioning, cameras, sensors, etc. to monitor and adjust in real time during flight. Dynamically adjust: adjust the flight path according to real-time data (such as weather changes, newly discovered obstacles).

[0118] 7. Deployment and execution: Confirm that the unmanned aerial vehicle is in good condition and all devices are working properly. Upload the optimized flight path to the unmanned aerial vehicle control system. Start the unmanned aerial vehicle and execute the cruise mission according to the planned path.

[0119] 8. Data Collection and Analysis: Collect images, videos, and sensor data collected by drones during their patrols. Process and analyze the collected data to extract useful information, such as coal mine condition assessment and safety hazard identification.

[0120] 9. Feedback and Optimization: Evaluate the effectiveness of cruise missions, including coverage completeness and data quality. Continuously optimize the route planning module based on feedback to improve efficiency and accuracy.

[0121] In this technical solution, an efficient and reliable drone cruise route planning module can be built to ensure full coverage of the coal mine area and provide precise flight guidance.

[0122] Furthermore, the image acquisition module uses a drone carrying a laser radar to perform laser scanning of the coal mine; low-altitude flight photography of the coal mine area includes the following steps: Take off the drone: Start the drone in a safe area and fly it at low altitude according to the preset flight path.

[0123] Laser scanning: The laser scanner carried by the drone continuously scans the coal mine area to obtain three-dimensional point cloud data of the surface.

[0124] Image capture: At the same time, the high-definition camera on the drone captures high-resolution image data to record the detailed conditions of the coal mine area.

[0125] Data return: The collected data is sent back to the ground control station in real time or at a fixed time to ensure the integrity and security of the data.

[0126] Furthermore, the image preprocessing module preprocesses the acquired scanned image, including the following steps: Step 1: Data import and organization: Import the data sent back from the drone (including images and point cloud data) into professional image processing software.

[0127] Step 2: Filtering and denoising: Apply algorithms such as Gaussian filtering and median filtering to remove noise from the image and improve image quality. De-noise the laser point cloud data to remove outliers caused by environmental interference and improve data accuracy.

[0128] Step 3: Normalization: Adjust the brightness, contrast and other parameters of the image to make the image data within a unified range for easy subsequent processing. Perform coordinate conversion and scale unification on the laser point cloud data to ensure data consistency and comparability.

[0129] Furthermore, the anomaly recognition module extracts features from the preprocessed image and uses machine learning or deep learning algorithms to identify anomalies, including the following steps: Step 1: Feature Extraction: This includes image feature extraction and point cloud feature extraction. Image Feature Extraction: Utilize edge detection, texture analysis, shape recognition, and other methods to extract features related to ground cracks, collapse pits, and other relevant features from pre-processed images. This includes color, texture, and shape. Point Cloud Feature Extraction: Extract terrain features such as elevation changes, slope, curvature, and other features from laser point cloud data. These features help identify abnormal conditions such as ground subsidence.

[0130] Step 2: Build Recognition Model: Based on specific application scenarios and data characteristics, select appropriate machine learning or deep learning algorithms such as convolutional neural networks (CNN) and others. Use labeled abnormal samples (such as known crack, collapse pit images and point cloud data) as training sets to train the algorithm to recognize similar abnormal conditions. Use cross-validation and other methods to evaluate model performance to ensure accuracy and stability on unknown data.

[0131] Step 3: Abnormality Detection and Recognition: Input pre-processed images and point cloud data into the trained model for abnormality detection. Input real-time data into the model for fast and accurate abnormality detection. Combine laser scanning point cloud data and image data to interpret the abnormality detection results and determine the specific type and extent of the anomaly. Analyze the model output to determine whether there are ground cracks, collapse pits, and other abnormal conditions, and mark the abnormal areas.

[0132] Predicted results integration P integrated : P integrated = w image * P image + w point_cloud * P point_cloud ; Abnormal probability calculation P anomaly : ; In the formula, P image and P point_cloud represent the model's abnormality detection output score for image data and the model's abnormality detection output score for point cloud data, respectively. w image and w point_cloud represent the weights of image and point cloud data. P integrated is the integrated prediction result of image and point cloud data. α and β are parameters used to adjust the output in the sigmoid function. P anomaly is the predicted probability of abnormality.

[0133] Anomaly threshold determination D is performed using the following formula: D = {x | x > θ}; where θ is the threshold value for anomaly detection. θ can be optimized based on performance on a validation set. D is the binary determination result of the anomaly (e.g., exceeding the threshold is an anomaly). x is the anomaly probability or score output by the model, i.e., P anomaly The specific numerical value. All model output anomaly probabilities or scores x are determined as anomalies (D is true) if they are greater than the threshold θ, otherwise they are determined as normal (D is false).

[0134] Step four: monitoring report generation: according to the results of anomaly detection, generate detailed monitoring report, including abnormal position, type, scale and other information, provide scientific basis for safety management and governance work of coal mine.

[0135] Step five: generate three-dimensional terrain model by laser scanning: use laser point cloud data to generate high-precision three-dimensional terrain model through three-dimensional modeling software. The model can intuitively show the terrain features of the coal mine area, which helps to more accurately identify and analyze abnormal situations. Combine the three-dimensional model with the anomaly detection results to conduct more in-depth spatial analysis and risk assessment.

[0136] Further, the comprehensive evaluation module fuses the data obtained by various monitoring means and the monitoring results of the anomaly recognition module, and comprehensively analyzes multi-source information such as ground deformation, underground displacement, stress change, and underground water conditions to evaluate coal mine subsidence and potential subsidence risks, including the following steps: 1. Data collection: collect multi-source data from various sensors, including ground deformation, underground displacement, stress change, underground water conditions, and monitoring results of the anomaly recognition module; 2. Data preprocessing: including data cleaning, data calibration and data conversion; Data cleaning: remove outliers, missing values and unreasonable data.

[0137] Data calibration: calibrate data from different sources to ensure consistency and accuracy of data.

[0138] Data conversion: convert raw data into a standard format for analysis.

[0139] 3. Data fusion processing: ensure that all monitoring data are synchronized in time for synchronous analysis. Match the data of different monitoring points in space to form a complete monitoring network.

[0140] Time synchronization interpolation (linear interpolation of time series): Data i time−sync(t)=Data itransformed (t1)+Data itransformed (t2)-Data i transformed (t1)×(t−t1) / (t2−t1); Where t1 and t2 are the time coordinates of the original data points, and t is the time point that needs to be interpolated. i time −sync(t)是在时间点t处，第i个数据源经过时间同步处理后的数据。Data itransformed (t1) is the data value of the i-th data source at the original time point t1. i transformed (t2) is the data value of the i-th data source at another original time point t2, where t2>t1. t is the time point at which interpolation is required.

[0141] Spatial matching resampling (using bilinear resampling): Data i spatial−match(x,y)=BilinearInterpolate(Data itime−sync ,x,y); where BilinearInterpolate is a bilinear interpolation function. i spatial−match (x,y) is the data of the i-th data source at the spatial coordinate (x,y) after spatial matching processing.

[0142] A weighted fusion algorithm is used for multi-source data: different weights are assigned to each data source based on its reliability and accuracy. Multi-source information such as ground deformation, underground displacement, stress changes, and groundwater conditions are used to assess coal mine subsidence and potential subsidence risks. Weighted convergence (using exponentially weighted moving average): ; Among them, λ i is the decay rate of data source i (used to consider the newness of the data, the decay rate λ i is a constant less than 1 but close to 1, used to achieve exponential decay effect. ), t i is the data acquisition time. WeightedFusion(t) is the weighted fusion result at time point t. i is the weight assigned to the i-th data source. i is the decay rate of the ith data source, used in the exponentially weighted moving average. i is the time point of data collection.

[0143] Risk assessment: Based on the integrated data after weighted fusion, combined with historical data and expert experience, set reasonable warning thresholds. When the monitoring data reaches or exceeds the warning threshold, trigger the warning mechanism, indicating the existence of coal mine collapse or potential collapse risk. When the monitoring data reaches or exceeds the warning threshold, trigger the warning mechanism. According to the warning results and actual situation, feedback and adjust the warning model. Continuously optimize the warning parameters and threshold settings to improve the adaptability and accuracy of the warning system.

[0144] 4. Comprehensive analysis: including ground deformation analysis, underground displacement analysis, stress change analysis and groundwater condition analysis; Ground deformation analysis: analyze the deformation characteristics such as ground subsidence and uplift, and evaluate the deformation rate and trend. Combined with geological conditions, judge whether the deformation is caused by coal mining.

[0145] Underground displacement analysis: monitor the moving direction and speed of underground rock strata, and evaluate its influence on ground stability. Analyze the correlation between underground displacement and ground deformation.

[0146] Stress change analysis: monitor the stress change of coal seam and roof and floor rock strata, and evaluate the stress distribution and evolution law caused by mining. Analyze the influence of stress change on rock stability and collapse risk.

[0147] Groundwater condition analysis: analyze the rise and fall of groundwater level, water quality change, etc., and evaluate its influence on rock stability and collapse risk. Combined with hydrogeological conditions, judge the correlation between groundwater and collapse.

[0148] 5. Establish a multi-parameter warning model: according to the results of comprehensive analysis, determine the key indicators that can reflect the collapse risk (such as ground deformation rate, underground displacement, stress change, groundwater level change and monitoring results of anomaly recognition module, etc.). According to historical data and expert experience, set reasonable warning thresholds for each warning indicator. Use decision tree machine learning algorithm to build a multi-parameter warning model. The model should be able to consider the interaction and mutual influence between parameters, improve the accuracy and reliability of warning.

[0149] 6. Evaluation and feedback: real-time processing and analysis of monitoring data, timely evaluation of collapse risk. When the monitoring data reaches or exceeds the warning threshold, trigger the warning mechanism. According to the warning results and actual situation, feedback and adjust the warning model. Continuously optimize the warning parameters and threshold settings to improve the adaptability and accuracy of the warning system.

[0150] In this technical solution, the comprehensive evaluation module can realize comprehensive and accurate evaluation of coal mine collapse and potential collapse risk, providing strong guarantee for coal mine safety production.

[0151] The present application provides a coal mine collapse monitoring method, comprising the following steps: S1, the deformation monitoring module monitors the deformation of the coal mine.

[0152] S2, the displacement stress monitoring module installs inclinometers and displacement meters in the coal mine goaf or potential subsidence area to monitor the rock mass displacement, stress state and strain change of the coal mine goaf or potential subsidence area. To evaluate the stability and subsidence risk of rock mass.

[0153] S3, the groundwater monitoring module measures the change of groundwater level in real time, analyzes the influence of groundwater on the stability of coal mine.

[0154] S4, the route planning module plans the cruising route of the unmanned aerial vehicle; ensure comprehensive coverage of the coal mine area.

[0155] S5, the image acquisition module carries high-definition camera and laser radar through unmanned aerial vehicle to carry out laser scanning on coal mine; the unmanned aerial vehicle carries out cruising monitoring according to the planned route; the unmanned aerial vehicle carries laser scanner and other equipment, carries out low-altitude flight shooting on coal mine area, and obtains high-resolution image data.

[0156] S6, the image preprocessing module pre-processes the obtained scanning image, including filtering denoising, normalization, etc. S7, the anomaly recognition module extracts features from the pre-processed image, and uses machine learning or deep learning algorithm to identify abnormal conditions, monitor ground cracks, subsidence pits and other features. Laser scanning technology can quickly obtain ground elevation data, generate high-precision three-dimensional terrain model, and is suitable for monitoring of complex terrain. Identify the abnormal situation of coal mine subsidence.

[0157] S8, the comprehensive evaluation module carries out multi-source data fusion, fuses the data obtained by various monitoring means and the monitoring results of the anomaly recognition module, comprehensively analyzes the multi-source information such as ground deformation, underground displacement, stress change and groundwater condition, establishes multi-parameter early warning model, and improves the accuracy and reliability of subsidence early warning.

[0158] S9, when the monitoring data exceeds the preset threshold, the alarm module automatically triggers the early warning signal, reminding the relevant personnel to take measures.

[0159] The coal mine subsidence monitoring system of the present invention operates as follows: a deformation monitoring module monitors coal mine deformation. A displacement and stress monitoring module installs inclinometers and displacement meters in coal mine goafs or potential subsidence areas to monitor rock mass displacement, stress state, and strain changes in these areas to assess rock mass stability and subsidence risk. A groundwater monitoring module measures groundwater level changes in real time and analyzes the impact of groundwater on coal mine stability. A route planning module plans drone routes to ensure comprehensive coverage of the coal mine area. An image acquisition module uses drones equipped with high-definition cameras and lidar to perform laser scanning of the coal mine. The drones, equipped with laser scanners and other equipment, conduct low-altitude photography of the coal mine area to acquire high-resolution image data. An image preprocessing module preprocesses the acquired scanned images, including filtering, denoising, and normalization. An anomaly recognition module extracts features from the preprocessed images and uses machine learning or deep learning algorithms to identify anomalies and monitor features such as ground cracks and subsidence pits. Laser scanning technology can rapidly acquire surface elevation data and generate high-precision three-dimensional terrain models, making it suitable for monitoring complex terrain. Identify abnormal coal mine collapse conditions. The comprehensive assessment module integrates multi-source data, combining data collected by various monitoring methods with the monitoring results of the anomaly identification module. It comprehensively analyzes multi-source information such as ground deformation, underground displacement, stress changes, and groundwater conditions to establish a multi-parameter early warning model, improving the accuracy and reliability of collapse warnings. When monitoring data exceeds a preset threshold, the alarm module automatically triggers a warning signal, prompting relevant personnel to take countermeasures.

[0160] By comprehensively monitoring multiple sources of information, including ground deformation, underground displacement, stress changes, and groundwater conditions in coal mining areas, the system can promptly detect potential collapse risks and automatically trigger warning signals when data exceeds preset thresholds, thereby achieving early warning and timely response, reducing or avoiding casualties and property losses. Through a comprehensive assessment module, the system can integrate and process multi-source data, apply machine learning or deep learning algorithms for intelligent analysis and evaluation, and establish a multi-parameter warning model to improve the accuracy and reliability of collapse warnings.

[0161] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments can still be modified, or some of the technical features thereof can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A coal mine subsidence monitoring method, characterized in that: The following steps are involved: S1, deformation monitoring module performs deformation monitoring on coal mines; S2, displacement and stress monitoring module monitors the displacement, stress state and strain changes of rock mass in coal mine goaf or potential collapse area; S3, groundwater monitoring module measures the changes of groundwater level in real time; S4, the route planning module plans the cruising route of the UAV; S5, the image acquisition module uses the drone to perform patrol monitoring along the planned route; The drone is equipped with a laser scanner and a high-definition camera to conduct low-altitude flight photography of the coal mining area; S6, the image preprocessing module preprocesses the acquired scanned image; S7, the anomaly recognition module extracts features from the preprocessed images and uses machine learning algorithms to identify anomalies and monitor ground cracks and collapse pit features; Identify abnormal conditions of coal mine subsidence; S8, the comprehensive assessment module integrates multi-source data, comprehensively analyzes multi-source information on ground deformation, underground displacement, stress changes, and groundwater conditions, establishes a multi-parameter early warning model, and improves the accuracy and reliability of collapse warnings; S9. When the monitoring data exceeds the preset threshold, the alarm module automatically triggers a warning signal.

2. The coal mine subsidence monitoring method according to claim 1, characterized in that: Step S1 includes the following steps: S11. Determine monitoring areas: Determine key monitoring areas based on the geological conditions, mining history, and previous subsidence records of the coal mine; S12. Layout of monitoring points: Rationally layout leveling points and inclinometer monitoring points within the monitoring area; S13. Leveling: Establish a leveling network within the monitoring area and use a level or laser level to measure the elevation of monitoring points; S14, Inclinometer Monitoring: Install an inclinometer at the monitoring point, which automatically and continuously monitors the tilt changes of the surface and records the data in real time; S15. Data processing and analysis: Organize and process the measurement data, calculate the elevation change of each monitoring point; evaluate the settlement of the coal mine area by analyzing the elevation change; process and analyze the data monitored by the inclinometer, calculate the tilt change and tilt rate of the surface; evaluate the tilt deformation of the coal mine area by analyzing the tilt change and rate; determine whether there is a risk of collapse.

3. The coal mine subsidence monitoring method according to claim 1, characterized in that: Step S2 includes the following steps: S21. Clarify monitoring objectives: determine the specific areas and monitoring parameters that need to be monitored; S22. Design monitoring plan: Select appropriate sensor type and quantity and plan the location of the sensors based on geological conditions, monitoring requirements, and budget; S23. Site Survey: Conduct a detailed survey of the monitoring area to determine the specific installation location; S24. Sensor installation: Install displacement gauges and strain gauges at key locations based on monitoring requirements; S25, Data Collection: Install data collectors; establish data transmission channels to ensure that data can be transmitted to the monitoring center in real time or regularly; S26, System debugging: Perform secondary calibration on the installed sensors to ensure that the measurement data is accurate and reliable; S27. Data analysis: Process and analyze the collected data to evaluate the displacement, stress state and strain changes of the rock mass and predict possible collapse risks.

4. The coal mine subsidence monitoring method according to claim 1, characterized in that: Step S3 includes the following steps: S31. Data Collection: Collect basic geological data from land, water resources, and environmental protection departments; collect geotechnical engineering survey data and reports for large-scale construction projects in the coal mining area; and obtain dynamic monitoring data on groundwater levels and water quality in the assessment area, as well as the status of groundwater development and utilization. S32. On-site investigation: Conduct on-site investigation of the coal mine area to identify areas with potential hydrogeological problems; S33. Monitoring well site selection: Based on the data collection and on-site investigation results, monitor wells are reasonably located within the coal mine area; S34, Drilling Construction: Carry out drilling construction according to design requirements to ensure that the drilling depth reaches the target aquifer; install water level monitoring sensors in the well pipe and debug the installed monitoring equipment; S35. Regular monitoring: Regularly measure water levels and perform water quality sampling and analysis in monitoring wells; S36. Data Analysis: Analyze the changing trends of groundwater levels and identify abnormal changes; analyze the changes in water quality parameters and assess the stability of the groundwater system; S37. Risk Assessment: Based on monitoring data and analysis results, assess the stability of the groundwater system in the coal mine area; identify potential collapse risk areas and factors; S38. Formulate appropriate response measures for abnormal changes in groundwater levels.

5. The coal mine subsidence monitoring method according to claim 1, characterized in that: Step S7 includes the following steps: S71, feature extraction: extracting features from the preprocessed image data; S72. Build a recognition model: Based on the specific application scenario and data characteristics, select a convolutional neural network (CNN) model. Use labeled abnormal samples as a training set to train the model so that it can recognize similar abnormal situations. Use cross-validation to evaluate model performance. S73. Anomaly Detection and Identification: Combine the point cloud data and image data from laser scanning to interpret the anomaly detection results and determine the specific type and extent of the anomaly. Analyze the model output to determine whether there are abnormal ground cracks and collapse pits, and mark the abnormal areas. S74, monitoring report generation: Generate a detailed monitoring report based on the results of anomaly detection; S75. Laser scanning generates a three-dimensional terrain model.

6. The coal mine subsidence monitoring method according to claim 5, characterized in that: Step S71 includes image feature extraction and point cloud feature extraction; using edge detection method to extract features related to ground cracks and collapse pits from the preprocessed image; extracting terrain features from the laser point cloud data, including elevation changes, slope and curvature.

7. The coal mine subsidence monitoring method according to claim 5, characterized in that: In step S75, a high-precision three-dimensional terrain model is generated using the laser point cloud data through three-dimensional modeling software; the three-dimensional model is combined with the anomaly detection results to conduct a more in-depth spatial analysis and risk assessment.

8. The coal mine subsidence monitoring method according to claim 1, characterized in that: Step S8 includes the following steps: S81, data collection: collecting monitoring results of various sensors and anomaly recognition modules; S82, data preprocessing: including data cleaning, data calibration and data conversion; S83, data fusion processing: spatially match data from different monitoring points to form a complete monitoring network; use weighted fusion algorithm to fuse multi-source data; S84. Comprehensive analysis: including ground deformation analysis, underground displacement analysis, stress change analysis and groundwater condition analysis; S85. Establish a multi-parameter early warning model: Based on the comprehensive analysis results, determine the key indicators that can reflect the collapse risk. Based on historical data and expert experience, set reasonable warning thresholds for each warning indicator. Use a decision tree machine learning algorithm to build a multi-parameter early warning model. S86. Evaluation and feedback: Process and analyze monitoring data in real time to promptly assess collapse risks; trigger the early warning mechanism when monitoring data reaches or exceeds the early warning threshold; provide feedback and adjust the early warning model based on the early warning results and actual conditions.

9. The coal mine subsidence monitoring method according to claim 1, characterized in that: Step S83 includes the following steps: S831, calculating time synchronization interpolation processing; S832, perform spatial matching resampling: S833. Use a weighted fusion algorithm for multi-source data: fuse the monitoring data of ground deformation, underground displacement, stress change, groundwater conditions, and anomaly identification modules to assess coal mine collapse and potential collapse risks. S834. Risk assessment: Based on the weighted fusion of comprehensive data, combined with historical data and expert experience, set a reasonable warning threshold; when the monitoring data reaches or exceeds the warning threshold, the warning mechanism is triggered to indicate the existence of coal mine collapse or potential collapse risk.

10. A coal mine subsidence monitoring system comprising: Central control unit, data acquisition module, deformation monitoring module, displacement stress monitoring module, groundwater monitoring module, route planning module, image acquisition module, image preprocessing module, anomaly recognition module, comprehensive evaluation module and alarm module; characterized by: Data acquisition module: Collects and comprehensively records the surface conditions of the coal mine area, including topography, landforms, vegetation cover, building distribution, and potential collapse risk areas; collects images of the coal mine, including normal images and abnormal images, and annotates the images as reference samples; Deformation monitoring module: including leveling and inclinometer monitoring; deformation monitoring of coal mines; Displacement and stress monitoring module: including inclinometers, displacement meters, stress meters and strain meters, to monitor the displacement, stress state and strain changes of rock mass in coal mine goafs or potential collapse areas; Groundwater monitoring module: real-time measurement of groundwater level changes and analysis of the impact of groundwater on coal mine stability; Route planning module: plans the drone's cruising route to ensure full coverage of the coal mine area; Image acquisition module: including drones, high-definition cameras and laser scanners; drones equipped with laser radars perform laser scanning of coal mines; drones equipped with laser scanners and high-definition cameras perform low-altitude flight photography of coal mine areas; Image preprocessing module: preprocesses the acquired scanned images, including filtering, denoising and normalization; Anomaly Recognition Module: This module extracts features from pre-processed images and uses machine learning algorithms to identify anomalies, monitor ground cracks and collapse pit features, and identify anomalies in coal mine collapses. Comprehensive assessment module: This module integrates multi-source data, comprehensively analyzes ground deformation, underground displacement, stress changes, and groundwater status information, establishes a multi-parameter early warning model, and improves the accuracy and reliability of collapse warnings; Alarm module: includes an alarm. When the monitoring data exceeds the preset threshold, the system automatically triggers an early warning signal to remind relevant personnel to take countermeasures; Central control unit: network connected with data acquisition module, deformation monitoring module, displacement stress monitoring module, groundwater monitoring module, route planning module, image acquisition module, image preprocessing module, anomaly recognition module, comprehensive evaluation module and alarm module.

Images