Coal mine safety mobile monitoring device based on scanning laser

Abstract

The present invention discloses a coal mine safety mobile monitoring device based on scanning laser, which relates to the field of mine safety technology. The system includes: a monitoring configuration information confirmation module, which determines the monitoring configuration information for the coal mine area space; a laser point cloud data determination module, which controls the three-dimensional laser scanner and determines the laser point cloud data; a mixed deformation model construction module, which constructs a mixed deformation model; an abnormal change distribution positioning module, which locates the abnormal change distribution; a risk control early warning module, which generates real-time risk information; a lower-level monitoring configuration determination module, which performs change trend prediction to determine the lower-level monitoring configuration; and a safety monitoring management module, which performs safety monitoring management. The present invention solves the technical problems of low accuracy of coal mine safety monitoring in the prior art and the inability to accurately predict and locate potential safety hazards, and achieves the technical effect of realizing accurate monitoring of coal mine areas, effectively preventing accidents, and ensuring coal mine production safety.

Description

Technical Field

[0001] The invention relates to the technical field of mine safety, and in particular to a coal mine safety mobile monitoring device based on scanning laser. Background Art

[0002] In the coal mining industry, ensuring production safety has always been a vital task. With the continuous increase in the depth and scale of coal mining, the geological conditions have become more complex, and the safety risks in the mining process have also increased. Traditional coal mine safety monitoring methods often have problems such as low monitoring accuracy, limited coverage, and insufficient data processing capabilities, which are difficult to meet the high requirements of modern coal mine safety production. In the past, monitoring of coal mining areas mainly relied on manual inspections and some simple sensor equipment. These methods are not only inefficient, but also easily affected by human factors and environmental interference, resulting in inaccurate and untimely monitoring results. At the same time, due to the lack of effective data integration and analysis methods, it is difficult to extract valuable information from massive monitoring data, and it is impossible to accurately predict and locate potential safety hazards.

[0003] The existing technology has the technical problem of low accuracy in coal mine safety monitoring and inability to accurately predict and locate potential safety hazards. Summary of the invention

[0004] The present application provides a coal mine safety mobile monitoring device based on scanning laser, which is used to solve the technical problems in the prior art of low accuracy of coal mine safety monitoring and inability to accurately predict and locate potential safety hazards.

[0005] In view of the above problems, the present application provides a coal mine safety mobile monitoring device based on scanning laser.

[0006] The present application provides a coal mine safety mobile monitoring device based on scanning laser, the device comprising:

[0007] A monitoring configuration information confirmation module, which is used to perform local segmentation and laser scanning configuration for the coal mine area space based on risk characteristics, and determine the monitoring configuration information, including risk distribution characteristics and risk change characteristics, and the local segmentation standard is risk homology; a laser point cloud data determination module, which controls the three-dimensional laser scanner based on the monitoring configuration information, performs three-dimensional scanning and reception of the coal mine area space, performs point cloud splicing and integration, and determines the laser point cloud data; a hybrid deformation model construction module, which is used to construct a hybrid deformation model, and the hybrid deformation model includes multiple parallel deformation decision branches guided by feature decision requirements, and uses R-tree index as the point cloud allocation call for each decision branch. Method; an abnormal change distribution positioning module, which performs branch data call and phase change analysis on the laser point cloud data based on the mixed deformation model to locate the abnormal change distribution, wherein the change analysis dimension includes spatial change and phase change, and the phase change includes vertical change and dislocation change; a risk control warning module, which generates real-time risk information and performs risk control warning based on the abnormal change distribution; a lower-level monitoring configuration determination module, which performs change trend prediction and determines the lower-level monitoring configuration based on the abnormal change distribution; a safety monitoring management module, which performs safety monitoring and management on the coal mine area space based on the real-time risk information and the lower-level monitoring configuration.

[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0009] The monitoring configuration information confirmation module is used to perform local segmentation and laser scanning configuration based on risk characteristics for the coal mine area space, and determine the monitoring configuration information; the laser point cloud data determination module controls the three-dimensional laser scanner based on the monitoring configuration information, performs three-dimensional scanning and reception of the coal mine area space, performs point cloud splicing and integration, and determines the laser point cloud data; the hybrid deformation model construction module constructs a hybrid deformation model, the hybrid deformation model includes multiple parallel deformation decision branches guided by feature decision requirements, and uses R-tree index as the point cloud allocation and calling method for each decision branch; the abnormal change distribution positioning module performs branch data calling and phase change analysis on the laser point cloud data based on the hybrid deformation model, and locates the abnormal change distribution; the risk control warning module generates real-time risk information and performs risk control warning based on the abnormal change distribution; the lower-level monitoring configuration determination module performs change trend prediction based on the abnormal change distribution to determine the lower-level monitoring configuration; the safety monitoring management module performs safety monitoring and management of the coal mine area space based on the real-time risk information and the lower-level monitoring configuration. The technical effect of achieving accurate monitoring of coal mine areas, effectively preventing accidents and ensuring safe coal mine production has been achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0011] Figure 1 A schematic diagram of the structure of a coal mine safety mobile monitoring device based on scanning laser provided in an embodiment of the present application.

[0012] Explanation of the accompanying drawings: monitoring configuration information confirmation module 10, laser point cloud data determination module 20, mixed deformation model construction module 30, abnormal change distribution positioning module 40, risk control warning module 50, lower-level monitoring configuration determination module 60, safety monitoring management module 70. DETAILED DESCRIPTION

[0013] The present application provides a coal mine safety mobile monitoring device based on scanning laser, which is used to solve the technical problems in the prior art of low accuracy of coal mine safety monitoring and inability to accurately predict and locate potential safety hazards.

[0014] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0015] Embodiment 1

[0016] like Figure 1 As shown, the present application provides a coal mine safety mobile monitoring device based on scanning laser, the device comprising:

[0017] The monitoring configuration information confirmation module 10 is used to perform local segmentation and laser scanning configuration based on risk characteristics for the coal mine area space, determine the monitoring configuration information, including risk distribution characteristics and risk change characteristics, and the local segmentation standard is risk homology.

[0018] Specifically, the monitoring configuration information confirmation module is an important part of the entire device responsible for formulating the basic configuration plan for the monitoring work of the coal mine regional space. It focuses on the specific object of the coal mine regional space. Through in-depth analysis of the area, especially based on risk characteristics, work is carried out. Risk characteristics cover a variety of factors, such as the instability of geological structure, abnormal changes in gas concentration, fluctuations in groundwater levels, stress concentration caused by mining activities, etc. Based on these risk characteristics, local segmentation is performed. Local segmentation means dividing the coal mine regional space into different local areas. The local segmentation criterion of risk homology means dividing areas with the same or similar risk sources, properties and development trends together. While completing the local segmentation, laser scanning configuration is also performed, which includes determining the layout position, scanning angle, scanning frequency and other parameters of the laser scanner. Finally, a series of monitoring configuration information is determined. Among them, the risk distribution characteristics describe the specific location and range of different risks in the coal mine regional space; the risk change characteristics reflect the trend and law of risk change over time, for example, some risks may gradually intensify, while others may tend to stabilize or weaken. It provides key basic data and configuration solutions for subsequent monitoring and analysis work, helping to improve the accuracy and effectiveness of spatial monitoring in coal mine areas.

[0019] The laser point cloud data determination module 20 controls the three-dimensional laser scanner based on the monitoring configuration information, performs three-dimensional scanning and reception of the coal mine area space, performs point cloud splicing and integration, and determines the laser point cloud data.

[0020] Specifically, the laser point cloud data determination module 20 is an important component of the entire device responsible for acquiring and processing key data. It uses the previously determined monitoring configuration information as the basis for action. These monitoring configuration information include detailed settings about the layout position, scanning angle, scanning frequency, etc. of the laser scanner. Based on this, the three-dimensional laser scanner is accurately controlled to start working, and the scanner performs a full range of three-dimensional scanning of the coal mine area space, while receiving a large amount of data returned during the scanning process. After obtaining the original point cloud data from different scanning angles and positions, the point cloud splicing and integration operation is performed to accurately splice the point cloud data from different scanning positions and angles together to form a complete, continuous and accurate overall point cloud data that reflects the actual situation of the coal mine area space. After the above-mentioned scanning, receiving and splicing and integration process, the laser point cloud data that can describe the three-dimensional characteristics of the coal mine area space in detail is finally determined. These data provide a solid foundation for subsequent analysis, modeling and decision-making, and contribute to real-time monitoring and rapid response of coal mine areas.

[0021] The hybrid deformation model construction module 30 is used to construct a hybrid deformation model, wherein the hybrid deformation model includes a plurality of parallel deformation decision branches guided by feature decision requirements, and uses R-tree index as a point cloud allocation calling method for each decision branch.

[0022] Specifically, the hybrid deformation model construction module 30 works with the fundamental orientation of meeting specific feature decision needs. In this process, the relevant data will be deeply analyzed to fully understand the actual problem, so as to divide multiple localizations. Each localization represents an area with similar characteristics and deformation laws. For each localization, a corresponding deformation decision branch will be established. These branches operate in parallel, which means that they can process the data and tasks in their respective localizations at the same time, thereby greatly improving the processing efficiency and accuracy of the model. Moreover, the number of branches strictly corresponds to the number of localization segmentations, ensuring that each localization can be analyzed and processed specifically and accurately. When building the model, the construction process of the hybrid deformation model is explained by taking the neural network algorithm as an example, and various relevant data of the coal mine area, such as geological data, mining activity data, historical deformation data, etc. are collected. The data is cleaned and preprocessed, including data standardization, normalization, processing of missing values, etc. Feature extraction and analysis, extracting meaningful features from the preprocessed data, such as the thickness of the geological layer, rock hardness, mining depth, etc. Analyze the potential relationship between these features and deformation. Based on the similarity and correlation of features, clustering algorithms (such as K-Means) are used to divide the coal mine area into different localizations. For each localization, a deformation decision branch based on a neural network is designed. The network structure can be a multi-layer perceptron (MLP), which includes an input layer, a hidden layer, and an output layer. The input layer receives the feature data of the localization. The hidden layer performs feature transformation and nonlinear processing through the weighted summation and activation function of neurons. The output layer predicts the deformation of the localization. The weights and biases of the network are adjusted using the back propagation algorithm to minimize the error between the predicted value and the actual deformation value. For each localization of point cloud data, an R-tree index is constructed according to its spatial position. This enables the neural network branch to quickly assign relevant point cloud data when needed. The prediction results of the neural network branches of each localization are combined, and the final deformation prediction is obtained through weighted averaging or other fusion strategies. According to the difference between the fusion result and the actual situation, the structure and parameters of the neural network are adjusted. Use an independent test set data input model to calculate the mean square error, mean absolute error and other indicators between the predicted deformation and the actual deformation. Compare with other traditional methods or benchmark models to verify the performance advantages of the model. Deploy the trained model to the actual monitoring system. As new data continues to accumulate, the model is retrained regularly to update its parameters and improve prediction accuracy. Through the above steps based on the neural network algorithm, a hybrid deformation model for coal mine area deformation prediction can be constructed. Of course, in practical applications, other suitable algorithms such as support vector machines and decision trees can be selected according to specific circumstances. In terms of the allocation and call of point cloud data, an efficient R-tree indexing method is adopted. The R-tree index can quickly and accurately allocate data to the corresponding deformation decision branch based on information such as the characteristics and location of the point cloud data.When a branch needs to call point cloud data for calculation and analysis, the R-tree index can quickly provide it with the required data, greatly improving the speed and efficiency of data processing. Constructing a mixed deformation model can more accurately predict the deformation of different local areas in the coal mine area, providing a more reliable basis for safe production.

[0023] The abnormal change distribution positioning module 40 performs branch data call and time phase change analysis on the laser point cloud data based on the mixed deformation model to locate the abnormal change distribution, wherein the change analysis dimensions include spatial change and time phase change, and the time phase change includes vertical change and misalignment change.

[0024] Specifically, first, according to the settings of each decision branch in the mixed deformation model, the laser point cloud data will be targeted for branch data calls. This means that it can accurately extract data subsets related to specific localization and deformation characteristics from massive data, thereby improving the efficiency and accuracy of the analysis. After obtaining the branch data, the temporal phase change analysis will be carried out. This analysis covers multiple key dimensions. From the perspective of spatial change, it will examine the changes in the position, shape and structure of objects in three-dimensional space. For example, whether there is a local uplift or depression on the ground in a certain area, or whether the building is tilted or twisted. In terms of temporal phase change, the analysis of vertical change focuses on the displacement change of the object in the vertical direction. This includes the settlement or rise of the ground, and the change in the height of the building. The dislocation change focuses on the position deviation of the object in the horizontal direction, such as whether there is a horizontal displacement between adjacent strata, or whether different parts of the building are no longer aligned in the horizontal direction. By comprehensively considering these change dimensions, the abnormal change distribution positioning module 40 can accurately determine the distribution of abnormal changes in space. It can identify areas that are significantly different from normal change patterns, thus providing key information support for subsequent risk assessment, early warning and decision-making. Effectively monitoring and locating abnormal changes enhances the security and stability of the infrastructure or area covered by the entire monitoring system, ensuring normal production and living order.

[0025] The risk control and early warning module 50 generates real-time risk information and performs risk control and early warning based on the abnormal change distribution.

[0026] Specifically, after obtaining the abnormal change data about the coal mine area provided by the abnormal change distribution positioning module 40, a targeted analysis is immediately carried out. It will focus on analyzing the significance of these anomalies in the specific environment of the coal mine, such as whether the displacement amplitude of the coal seam exceeds the safety threshold, whether the settlement speed of the tunnel roof is too fast, or whether the deformation range of the goaf is rapidly expanding. Based on the comprehensive evaluation of these key factors, real-time and appropriate risk information for the actual situation of the coal mine is generated. This information not only clearly points out the specific risk types currently faced by the coal mine, such as roof collapse risk, gas outburst risk, water seepage risk, etc., but also estimates the possible development trend of the risk and the degree of harm that may be caused. At the moment of generating risk information, the risk control warning module 50 quickly initiates a full range of early warning measures. It will highlight the risk area and related data on the large screen of the coal mine monitoring center, and send clear and clear early warning notifications to the communication equipment of coal mine managers and underground workers, including key information such as specific descriptions of the risks, recommended emergency response measures, and evacuation routes. For serious risk situations, continuous and strong sound and light alarms will be triggered to ensure that personnel can be highly alert even in the noisy environment underground. In addition, it can also work in conjunction with key facilities such as the ventilation system and drainage system of the coal mine to automatically adjust operating parameters to reduce risks. For example, if the roof of a coal mining face is detected to be sinking at an abnormally fast speed, high-risk information about roof collapse will be immediately generated, and an emergency evacuation order will be issued to underground workers. At the same time, the ground dispatch center will be notified to stop coal mining operations in the area and start emergency rescue preparations. With its accurate risk assessment and timely and effective early warning mechanism, the risk control early warning module 50 provides a solid guarantee for the safe production of coal mines and minimizes the possibility and losses of accidents.

[0027] The lower-level monitoring configuration determination module 60 performs change trend prediction based on the abnormal change distribution to determine the lower-level monitoring configuration.

[0028] Specifically, when the lower monitoring configuration determination module 60 receives the abnormal change distribution information, it quickly carries out in-depth analysis and processing, and uses various data analysis and prediction algorithms to process these data. For example, a time series analysis algorithm, such as a moving average method, an exponential smoothing method or an ARIMA model, is used to analyze the change pattern of abnormal change data over time, so as to predict its future short-term and medium-term trends. For data with complex patterns, machine learning algorithms, such as decision trees, random forests or support vector machines, are used. These algorithms can automatically learn hidden patterns and relationships from data and make predictions based on this. Based on the prediction of the change trend. When making trend predictions, multiple factors will be considered comprehensively. For example, not only the amplitude of abnormal changes is concerned, but also the speed, frequency of changes and the correlation with other related factors will be analyzed. According to the prediction results, the specific configuration of the lower monitoring is determined. If it is predicted that the abnormal change trend in a certain area is intensifying, the number of monitoring equipment in the area will be increased, such as installing more displacement sensors, pressure sensors, etc. For areas where rapid changes may occur, the frequency of monitoring will be increased, from the original monitoring once a day to monitoring every hour or even shorter time intervals. If it is predicted that abnormal changes may spread to adjacent areas, the adjacent areas will also be included in the key monitoring scope and the corresponding monitoring equipment will be deployed. At the same time, the transmission and storage methods of monitoring data will be adjusted according to the predicted risk level. For data in high-risk areas, ensure that they can be transmitted to the monitoring center more quickly and stably, and stored longer and more safely for subsequent detailed analysis and tracing. Through in-depth analysis and accurate prediction of the distribution of abnormal changes, the configuration of lower-level monitoring can be dynamically adjusted and optimized to achieve more accurate and effective monitoring and protection of coal mine safety.

[0029] The safety monitoring management module 70 performs safety monitoring and management on the coal mine area space based on the real-time risk information and the lower-level monitoring configuration.

[0030] Specifically, the safety monitoring management module 70 plays a vital role as the overall coordination part of the entire system. After receiving the real-time risk information and the lower-level monitoring configuration provided, a comprehensive safety monitoring management work is immediately carried out. First, the real-time risk information is deeply interpreted and evaluated. It will classify and prioritize the risks according to the severity, urgency and possible impact. For high-risk and emergency situations, the emergency response mechanism will be immediately activated, including notifying relevant personnel to evacuate the dangerous area urgently, dispatching rescue resources, etc. At the same time, the monitoring equipment and systems in the coal mine area will be managed and dispatched in an overall manner according to the lower-level monitoring configuration. It ensures that the monitoring equipment operates normally according to the predetermined configuration and collects accurate and timely data. If a monitoring device is found to be faulty or the data is abnormal, an alarm will be quickly issued and repair or replacement will be arranged. In terms of data management, real-time data from various monitoring points will be summarized and integrated, and stored and analyzed. Through data analysis, it can discover potential risk trends and patterns, and provide a basis for further optimizing monitoring configuration and safety management strategies. In addition, it is also responsible for information exchange and collaboration with other management systems within the coal mine. For example, it coordinates with the production scheduling system to suspend or adjust production operations when the risk is high; it cooperates with the personnel management system to ensure that the location and status of each employee are monitored in real time to ensure personnel safety. It also generates safety monitoring reports regularly to report the safety status of the coal mine area and the implementation of monitoring and management work to the superior management department and relevant regulatory agencies. The safety monitoring management module 70 realizes comprehensive, efficient and scientific safety monitoring and management of the coal mine area space through effective processing of risk information, reasonable use of monitoring configuration and collaboration with other systems, thus ensuring the safe production of coal mines to the greatest extent.

[0031] In a possible implementation, the laser point cloud data determination module 20 further includes:

[0032] A point cloud confidence interval setting unit is used to set a point cloud confidence interval, wherein the point cloud confidence interval includes a same-domain confidence interval and a cross-domain confidence interval, and is determined based on an extreme value of data migration confidence of the point cloud spacing.

[0033] An untrusted point cloud extraction unit is used to traverse the laser point cloud data, traverse the point cloud confidence interval matching and verification, and extract untrusted point clouds.

[0034] The valid point cloud data determining unit is used to compensate the untrusted point cloud and determine the valid point cloud data.

[0035] Specifically, the setting of the same-domain confidence interval is mainly concentrated in the same area with similar geological and physical characteristics. For example, for the same coal mining face, the point cloud spacing data obtained by multiple laser scanning of the area under normal mining conditions is deeply analyzed. Considering factors such as the thickness change of the coal seam, the stability of the roof, and the operation influence of the coal mining equipment, the mean value, standard deviation and other statistics of the point cloud spacing are calculated. Based on these statistical results, a reasonable range is determined as the same-domain confidence interval. If the point cloud spacing of the coal mining face still fluctuates within this range in subsequent monitoring, it can be considered to be in a normal state; once it exceeds this range, it may imply abnormal changes in the coal seam structure, such as accelerated roof sinking speed and coal seam fractures. The cross-domain confidence interval focuses on the relationship between the point cloud spacing between different but related areas. For example, the transition area from the coal mining face to the transportation roadway, or the transition zone from the shallow coal seam to the deep coal seam. Similarly, through the study of a large amount of historical data, the change law of the point cloud spacing between different regions is analyzed. Taking into account the continuity and difference of geological structures, changes in stress distribution and other factors, the extreme range of spacing that can reflect normal conditions is determined as the cross-domain confidence interval. When the monitored point cloud spacing crosses this cross-domain confidence interval, it may indicate potential safety hazards such as intensified fault activity and expanded goaf impact range. The determination of the extreme value of data migration confidence is based on a comprehensive consideration of the complex environment of coal mines and long-term monitoring data. This requires not only the analysis of data in normal production processes, but also the inclusion of data on possible abnormal conditions and extreme conditions. Through deep mining and simulation of these data, the maximum and minimum possible values ​​of the point cloud spacing under different conditions are found, which are used as the extreme value boundaries of confidence. These extreme value boundaries provide a scientific and accurate basis for dividing the same-domain confidence interval and the cross-domain confidence interval, so that the setting of the point cloud confidence interval can better adapt to the changing geological conditions and production environment of coal mines, thereby improving the reliability and sensitivity of safety monitoring.

[0036] Each laser point cloud data is read in turn. While reading each point cloud data, the previously set point cloud confidence intervals are checked one by one, including the same-domain confidence interval and the cross-domain confidence interval. The features of the currently read point cloud data, such as position, spacing, etc., are compared and verified with the range of the confidence interval. If the feature value of the point cloud data falls completely within the confidence interval, it is considered to be expected and reliable point cloud data. However, if it is found through comparison that the feature value of the current point cloud data exceeds the range of the confidence interval, then the point cloud will be marked and extracted and identified as an untrusted point cloud. For example, in the process of traversing the point cloud data of the coal mine tunnel, it is assumed that the same-domain confidence interval stipulates that the spacing between adjacent point clouds should be between 0.5 meters and 1 meter. When a point cloud is read, its spacing with the adjacent point cloud is 1.2 meters, which exceeds the set same-domain confidence interval, then the point cloud will be extracted as an untrusted point cloud waiting for subsequent processing and analysis. Through such a traversal, matching and verification process, point cloud data that may be abnormal or erroneous can be effectively screened out, providing a basis for further analysis and processing, thereby more accurately monitoring the safety status of coal mines.

[0037] Once the untrusted point clouds are extracted from a large amount of laser point cloud data, the next key step is to accurately compensate these untrusted point clouds to finally determine the valid point cloud data, ensure the accuracy and reliability of the data, and provide strong support for coal mine safety monitoring. The compensation process is a complex and delicate operation. First, an interpolation compensation method based on the surrounding trusted point clouds will be used. For each untrusted point cloud, multiple trusted point clouds will be searched within a certain range around it. The positions and characteristic values ​​of these trusted point clouds will be used to calculate the compensation value. The average value compensation is used to calculate the arithmetic mean of the corresponding characteristic values ​​(such as distance, coordinates, etc.) of the surrounding trusted point clouds, and the average value is used as the compensation value of the untrusted point cloud. When considering the special factors of the coal mine environment, the type of geological structure (such as folds, faults, etc.), the distribution characteristics of coal seams, the mode and intensity of mining activities, etc. will be deeply analyzed. For example, if the untrusted point cloud is located near a known fault, the possible impact of the fault on the point cloud position will be considered during compensation, and the compensation strategy will be adjusted accordingly. After the compensation operation is completed, it will not be directly identified as valid point cloud data, but will be strictly evaluated and verified. The compensated point cloud data will be compared with the set confidence interval again to check whether it meets the expected distribution law and quality standards. Cross-validation, visual inspection and other methods will also be used to ensure the accuracy of compensation. Through the above rigorous and meticulous compensation and verification process, the quality of point cloud data can be improved to the greatest extent, providing an accurate and reliable data foundation for coal mine safety monitoring.

[0038] In a possible implementation manner, the valid point cloud data determining unit further includes:

[0039] An untrusted point cloud determination unit is used to perform necessity determination on the untrusted point cloud.

[0040] The untrusted point cloud vacant unit is used to leave the untrusted point cloud vacant if the necessity threshold is not met.

[0041] An untrusted point cloud compensation unit is used to perform algorithmic compensation on the untrusted point cloud if a first necessity threshold is met, wherein the compensation method includes mean interpolation or trend interpolation.

[0042] An untrusted point cloud acquisition unit, wherein the untrusted point cloud acquisition unit is used to rescan and acquire the untrusted point cloud if a second necessity threshold is met, wherein the second necessity threshold is greater than the first necessity threshold.

[0043] Specifically, when the untrusted point cloud determination unit receives point cloud data identified as untrusted, it will quickly launch a comprehensive and in-depth evaluation process. It will analyze the precise position of the untrusted point cloud in the entire coal mine monitoring area in great detail. If this point cloud happens to be at the key roof support points of the coal mine tunnel, these support points are under tremendous pressure, and any slight change may cause serious safety problems; or it is at the precise installation position of important equipment, and the normal operation of the equipment depends on the stability of this position; or it is in a known high-risk area where geological disasters may occur, such as near faults, at the edge of goafs, etc., then the point cloud will be preliminarily identified as having extremely high potential importance. Next, the determination unit will deeply explore the degree of close correlation between the untrusted point cloud and the surrounding point clouds. It will calculate the geometric relationship between the point cloud and the adjacent point clouds, such as the distance and angle, and analyze the density distribution and change trend between them. If the deviation of this untrusted point cloud may lead to an unusual dense or sparse distribution pattern of the surrounding point clouds, destroying the uniformity and continuity that should have been there; or if it affects the spatial consistency of the overall data, such as causing abrupt turns or disconnections in areas that should have smoothly transitioned, then this will significantly increase the possibility of it being judged as necessary. At the same time, the judgment unit will also take into full account the specific goals and strict requirements of the current specific monitoring task. If the core goal of this monitoring is to accurately detect extremely small displacements of coal seams to warn of potential landslide risks in advance, then any untrusted point cloud related to coal seam displacement, even if its deviation seems small, is highly valued because of its critical impact on the monitoring target. In addition, in order to make more accurate and reliable judgments, the judgment unit will also make extensive use of rich historical data and mature statistical models. In terms of historical data, it covers the data set accumulated from multiple point cloud scans conducted in the same or similar coal mine environment in the past. These data not only include the distribution of point clouds in different regions, different mining stages, and different geological conditions, but also record the changing trends of point cloud characteristics over time. Statistical models are mathematical expressions or algorithms built based on these historical data. Common statistical models include regression analysis models, which are used to find the linear or nonlinear relationship between point cloud features and other related factors (such as mining time, geological structure changes, etc.); cluster analysis models, which group point cloud data with similar characteristics to find the category and possible causes of abnormal point clouds; and probability distribution models, such as normal distribution and Poisson distribution, which are used to evaluate whether the probability of the current untrusted point cloud is within a reasonable range. By analyzing the change patterns of these point cloud data in similar situations in the past and the actual impact on subsequent analysis and decision-making, the potential value and necessity of the current untrusted point cloud can be further evaluated.

[0044] First, the setting of the necessary threshold requires comprehensive consideration of many factors to achieve an accurate and reasonable standard. First, a detailed safety risk assessment is conducted. This includes a comprehensive analysis of various areas and structures of the coal mine. For example, for the top support structure of the main tunnel, since it may cause serious collapse accidents once a problem occurs, the risk level is extremely high, so the corresponding threshold will be set very strictly, and only a very small point cloud deviation may be allowed. In contrast, for some auxiliary channels or storage areas, since their direct impact on overall safety is relatively small, the threshold can be moderately relaxed. In terms of monitoring objectives and accuracy requirements, specific monitoring priorities need to be clarified. If the goal is to accurately detect small displacements of coal seams to prevent potential gas outbursts, the threshold may be set at the millimeter level to capture extremely subtle changes. If the main focus is on larger-scale structural deformations, the threshold can be increased accordingly. Statistical analysis of historical data is an important basis for setting thresholds. Collect a large amount of past point cloud data and process it using data analysis tools and algorithms. Calculate the mean, standard deviation, extreme value and other statistical quantities of the data, and observe the distribution of the data. If the data is normally distributed, the threshold can usually be set outside the range of the mean plus or minus a few standard deviations. For example, if the mean is 10 cm and the standard deviation is 2 cm, the threshold may be set below 6 cm or above 14 cm to identify abnormal situations. By taking all the above into consideration and after multiple simulations, tests and adjustments, a necessary threshold is finally determined that can not only ensure coal mine safety but also achieve a balance between monitoring efficiency, cost and accuracy. When the untrusted point cloud is judged to have not reached the preset necessary threshold after a rigorous necessity assessment, it indicates that the deviation or anomaly of these point clouds is considered to be relatively minor or has a negligible impact on the overall results in the current monitoring and analysis scenario. After receiving this judgment result, the vacant unit will take a series of precise operations to process these untrusted point clouds. First, it will specially mark these point clouds in the data storage structure to clarify their vacant status. This marking method ensures that in the subsequent data processing and analysis stages, the relevant algorithms and calculation processes can quickly identify and skip these point clouds, thereby improving processing efficiency. In some cases, the vacant unit will directly physically delete these untrusted point clouds that do not meet the necessary threshold from the data set. However, before executing the deletion operation, data is usually backed up first to prevent the deleted data from being restored in case the judgment criteria need to be reviewed or adjusted later.

[0045] When untrusted point clouds are judged to meet the first necessity threshold, it means that although these point clouds have deviations, they can be made more valuable for reference through appropriate compensation, thereby improving the quality and availability of the overall data. Different algorithms are used for compensation. Mean interpolation is a common method. If the interval between an untrusted point cloud and its adjacent point clouds is relatively small and the data distribution is relatively uniform, the reasonable value of the untrusted point cloud can be estimated by calculating the average value of the adjacent point clouds. However, when the interval is too large, simply using the mean may result in reduced accuracy. For some data that show obvious trends, such as exponential growth or decay trends, trend interpolation is more applicable. This method analyzes the data change trend between adjacent point clouds and infers the possible value of the untrusted point cloud based on this trend. Taking displacement monitoring in coal mines as an example, if a series of point clouds represent the displacement of a structure, there is an untrusted point cloud. If the displacement of the adjacent point clouds shows a gradually increasing exponential trend, then when compensating this untrusted point cloud, its possible displacement value will be calculated based on this exponential trend, rather than just relying on the mean of the adjacent point clouds. Through such targeted compensation methods, untrusted point clouds can be better processed, providing a more accurate and reliable data basis for subsequent analysis and decision-making.

[0046] When the untrusted point cloud is evaluated as meeting the second necessity threshold, it indicates that the importance of these point clouds is extremely high, and the required accuracy and reliability cannot be achieved only by compensating the existing data. The second necessity threshold here is set higher than the first necessity threshold because rescanning acquisition is a relatively costly and complex means of obtaining the most direct and accurate data. Therefore, rescanning acquisition will only be initiated when the importance of the point cloud for monitoring and analysis reaches a high level. In actual operation, the untrusted point cloud acquisition unit will rearrange the laser scanning device to conduct targeted scanning of the relevant area. This process requires careful planning, including determining parameters such as the scanning angle, position and frequency to ensure that the required point cloud data can be accurately obtained. For example, in coal mine safety monitoring, if the area where some untrusted point clouds are located is a key part that is about to be mined on a large scale, or an area where some potential safety hazards have already appeared, and these point clouds meet the second necessity threshold, then rescanning acquisition will be initiated. By obtaining new and more accurate point cloud data, it can provide more powerful support for coal mine safety production decisions. The untrusted point cloud acquisition unit ensures the acquisition of high-quality point cloud data in critical situations through strict necessity threshold judgment and precise rescanning and acquisition operations, thereby improving the reliability and effectiveness of the entire monitoring and analysis system.

[0047] In a possible implementation, the laser point cloud data determination module 20 further includes:

[0048] A device coordinate system determining unit is used to determine the distribution position of the three-dimensional laser scanner, and to determine the device coordinate system with the distribution position as the center, and the device coordinate system corresponds to the distribution position one by one.

[0049] A global coordinate system determining unit is used to determine a global coordinate system based on the coal mine regional space.

[0050] A coordinate transformation relationship determining unit is configured to determine a coordinate transformation relationship based on the global coordinate system and the device coordinate system.

[0051] A point cloud stitching and integration unit is used to perform point cloud stitching and integration based on the coordinate transformation relationship.

[0052] Specifically, first of all, in order to accurately determine the distribution position of the 3D laser scanner, a variety of high-precision positioning technologies and measurement methods are used. Combined with satellite positioning systems (such as GPS or Beidou) to obtain approximate geographic location information, and then use more sophisticated indoor positioning technologies, such as laser ranging, inertial navigation systems, or visual positioning based on specific markers, to further improve the accuracy of position determination, accurate to the centimeter or even millimeter level. Once the extremely accurate distribution position of the scanner is obtained, a dedicated device coordinate system is constructed with this position as the core. This device coordinate system usually takes the position of the scanner as the origin, and the direction of the coordinate axis is determined according to the orientation of the scanner and preset rules. For example, the scanning direction of the scanner can be defined as the positive direction of one of the coordinate axes, while the other two coordinate axes are based on the right-hand rule. Since the placement position of each 3D laser scanner in the coal mine is unique, a completely corresponding device coordinate system is established for each distribution position. This means that each device coordinate system has unique characteristics, reflecting the specific position and direction of the corresponding scanner in the coal mine space. For example, if a scanner is placed in a tunnel in a coal mine, and its distribution position is measured as (X1, Y1, Z1), then the device coordinate system constructed with this position as the origin will be able to accurately describe the relative position and direction relationship of the point cloud acquired by the scanner in the local space. Through this one-to-one device coordinate system construction method, the point cloud data collected by each scanner can be accurately expressed and recorded in its own independent and clear coordinate system, providing a clear and accurate basis for subsequent data processing and analysis.

[0053] The global coordinate system determination unit plays a vital role in the entire system. It establishes a unified and authoritative spatial reference framework for the measurement and analysis of the entire coal mine area. When determining the global coordinate system, the spatial structure of the coal mine area will first be thoroughly surveyed and mapped. This includes obtaining the topographic data of the coal mine, clarifying the location of the mine shaft, the distribution direction of the tunnels, the scope of the mining face, and the layout of various facilities (such as ventilation equipment, transportation tracks, etc.). Next, the origin of the global coordinate system is selected by comprehensively considering various factors. The main shaft of the coal mine is used as the origin because it is a key location for the entry and exit of personnel and materials and has clear and stable characteristics. Or the geometric center of the coal mine is selected to ensure the balance of the coordinate system in the entire area. The direction of the coordinate axis is usually set according to the direction of the geographic North Pole or the extension direction of the main tunnel of the coal mine. For example, the direction of the geographic North Pole is used as the positive direction of the Y axis to ensure consistency and compatibility with external geographic information. The positive direction of the X axis can be along the main extension direction of the main tunnel to better reflect the layout of the internal structure of the coal mine. The scale of the coordinate axis will be carefully determined according to the measurement accuracy requirements and the actual scale of the coal mine. If it is necessary to be accurate to smaller details, the scale may be set to a smaller value, such as 1 cm or 1 mm. For large-scale coal mining areas, in order to avoid the complexity of calculation and representation caused by excessive values, a scale of 1 meter or even larger may be selected. In the determination process, the conversion relationship with the internationally used geographic coordinate system or industry standard coordinate system will also be fully considered to enable seamless data exchange and integration with other related geographic information systems or databases.

[0054] The coordinate transformation relationship determination unit is responsible for accurately establishing the transformation rules between the global coordinate system and the device coordinate system. First, the various parameter characteristics of the global coordinate system and the device coordinate system are analyzed in depth. This includes obtaining detailed information such as the origin position of the two coordinate systems, the direction vector of the coordinate axis, and the scale unit. For the origin position, the coordinate difference of the origin of the device coordinate system in the global coordinate system is calculated, and this difference will be used as part of the translation amount. The comparison of the coordinate axis direction is more complicated. The rotation parameters are determined by calculating the angle and rotation order between the device coordinate system axis direction vector and the global coordinate system axis direction vector. For example, it is necessary to rotate around a certain axis by a certain angle first, then rotate around another axis, and so on, and finally determine the complete rotation order and angle. When considering the scale unit, the proportional relationship between the two is calculated. If the scale of the device coordinate system is smaller or larger, it needs to be scaled accordingly. Next, mathematical models and algorithms are used to construct specific transformation formulas or matrices. A common method is to use a rotation matrix to represent the rotation of the coordinate axis, combined with the translation vector and scaling factor to comprehensively describe the entire transformation process. In actual calculations, multiple known corresponding points in the device coordinate system and the global coordinate system are repeatedly verified and the conversion parameters are optimized to improve the accuracy of the conversion.

[0055] Fusion of point cloud data from different device coordinate systems into a unified and complete whole. When receiving point cloud data collected by different devices, obtain the previously determined coordinate transformation relationship, and use these precise transformation rules to transform the point cloud data in each device coordinate system into the global coordinate system one by one. During the transformation process, for each point, the corresponding rotation, translation and scaling parameters are calculated to obtain an accurate new position in the global coordinate system. Then, the point cloud splicing begins. Since there are overlapping areas in the point clouds collected by different devices, these overlapping parts need to be processed by specific algorithms. The common method is to match and fuse based on features such as point distance and normal vector. For adjacent point cloud data, detailed comparison and analysis will be performed. If two points are very close in spatial position and their attributes (such as color, intensity, etc.) are also similar, they are considered to be overlapping points. For overlapping points, average, weighted average or selection of better values ​​are used to integrate them to eliminate duplication and inconsistency. At the same time, the spliced ​​point cloud will be smoothed to eliminate mutations and discontinuities, so that the spliced ​​point cloud presents a natural and continuous effect both visually and numerically. During the entire splicing and integration process, quality inspection and optimization are continuously carried out. The accuracy and reliability of the splicing results are ensured by calculating the deviation between the spliced ​​point cloud and the real scene, checking the integrity and consistency of the point cloud, etc. Finally, a seamlessly spliced, complete and accurate comprehensive point cloud dataset that reflects the spatial form and characteristics of the coal mine area is generated, providing strong support for subsequent analysis, modeling and decision-making.

[0056] In a possible implementation, the abnormal change distribution positioning module 40 further includes:

[0057] A point cloud information distribution determining unit is used to index and call the first point cloud data of the first deformation decision branch to determine the point cloud information distribution, wherein each distribution point contains a feature multi-tuple.

[0058] The transition information distribution determining unit is used to traverse the point cloud information distribution, interactively place point cloud data and perform time phase transition analysis to determine the transition information distribution.

[0059] The abnormal change distribution positioning unit is used to traverse the change information distribution, make a change over-limit judgment, and locate the abnormal change distribution, wherein the over-limit judgment includes the spatiotemporal vector over-limit and trend over-limit, and the over-limit judgment standard is the coal mine safety standard.

[0060] Specifically, according to specific indexing rules and calling mechanisms, the first point cloud data is accurately obtained from the first deformation decision branch. These data are usually a collection of point clouds from different acquisition devices, different times or different locations. After extracting the first point cloud data, the unit begins to parse these data in depth and carefully to accurately determine the distribution of the point cloud in the entire space. It considers many key factors such as the spatial position relationship of each point, the density distribution, and the overall directional characteristics. The feature tuple contained in each distribution point is a detailed description of the rich information of the point. This tuple covers many key features, such as the three-dimensional coordinates (X, Y, Z) of the point, which are used to accurately determine its position in space; color information (RGB value), which can reflect the surface characteristics or material of the object; reflection intensity value, which helps to judge the material and lighting conditions of the object; normal vector, which can indicate the orientation of the surface where the point is located; and may even include texture features, temperature values, etc., depending on the capabilities of the acquisition device and application requirements. Through the careful analysis and integration of these feature multi-groups, the point cloud information distribution determination unit can construct a comprehensive and accurate point cloud information distribution model, which describes the physical position distribution of the point cloud in space and reveals the distribution patterns of various attribute characteristics contained therein, providing a solid foundation and important reference basis for subsequent in-depth analysis, processing and application.

[0061] Each element in the point cloud information distribution is carefully checked one by one. While traversing, it interacts closely with the upper point cloud data, which may be a collection of related point clouds collected previously with a time sequence. Then, the linear regression algorithm is used to analyze the change of the point cloud position and organize the point cloud position data collected at different times. Time is used as the independent variable and the position coordinates of the point (such as the X coordinate) are used as the dependent variable. Then, by calculating the mean, variance and other statistics of these data points, a preliminary understanding of the distribution of the data is obtained. Next, according to the principle of linear regression, the correlation coefficient between the independent variable (time) and the dependent variable (position coordinate) is calculated to determine the closeness of the linear relationship between them. After that, the least squares method is used to determine the parameters of the linear regression equation, that is, the slope and intercept of the straight line. By continuously adjusting the parameters, the sum of squares of the error between the actual position data and the position data predicted by the regression equation is minimized. For the change of the shape of the object, a similar linear relationship is established between the key parameters describing the shape (such as length, angle, etc.) and time, and the above calculation steps are repeated to obtain the change trend of the shape parameters over time. For the changes in the color, reflection intensity and other characteristics of the point cloud, we also use time as the independent variable and the characteristic value as the dependent variable to perform linear regression analysis to obtain the change pattern of the characteristics over time. Finally, through the linear regression analysis of the position, shape and characteristics of a large number of points, the distribution of the change information of the point cloud is fully determined.

[0062] The abnormal change distribution positioning unit plays a vital monitoring and early warning role in the entire system. It systematically checks each data in the change information distribution one by one. When checking, the change limit judgment will be carried out according to the strict coal mine safety standards. This judgment covers two key aspects of time and space: vector limit and trend limit. For the judgment of vector limit, the vector parameters such as displacement, velocity and acceleration of each point in the spatial and temporal dimensions are accurately calculated. These parameters are compared with the allowable range specified in the coal mine safety standards. If the vector value of a point exceeds the upper or lower limit set by the safety standard, it will be marked as a possible abnormality. The judgment of trend limit is more complicated. The change trend of the point will be analyzed, such as observing the change curve of displacement, characteristic value, etc. over a period of time. If the trend shows a sharp rise, sharp drop or unstable fluctuation, and exceeds the definition range of normal trend in the coal mine safety standard, it will also be identified as abnormal. Once the limit is exceeded, the unit will quickly locate the distribution area where these abnormal points are located. Through precise coordinate positioning and data association, the specific location and range of abnormal changes are determined. The abnormal change distribution positioning unit will accurately identify this area and issue an alarm in time so that appropriate safety measures can be taken to prevent possible coal mine accidents.

[0063] In a possible implementation manner, the transition information distribution determining unit further includes:

[0064] The upper point cloud data is point cloud distribution data of a previous phase monitoring node, and the previous phase monitoring node includes at least one phase node.

[0065] Specifically, the point cloud data refers to the point cloud distribution data obtained from each phase monitoring node before the current monitoring time. The previous phase monitoring node is the node obtained by monitoring before the current moment in the time series. These nodes are not a single specific moment, but contain at least one phase node. This means that the previous phase monitoring node may be composed of monitoring data from multiple different time points. For example, it may include monitoring data from yesterday, last week, last month or even earlier. By comparing and analyzing the current point cloud data with the point cloud distribution data of these previous phase monitoring nodes containing multiple past time points, we can have a more comprehensive and in-depth understanding of the changes and trends of the monitored objects in the time dimension, thereby providing a richer and more accurate basis for subsequent analysis and decision-making.

[0066] In a possible implementation manner, the transition information distribution determining unit further includes:

[0067] The first transition distribution determining unit performs vertical mapping and difference measurement on the point cloud information distribution and the upper point cloud data based on spatial position and feature homology to determine a first transition distribution.

[0068] The second transition distribution determining unit performs grid division on the point cloud information distribution and the upper point cloud data based on the point cloud approximation, performs misalignment mapping and difference measurement in the same grid domain, and determines a second transition distribution.

[0069] The third change distribution determination unit is used to spatially determine the first neighborhood feature of the point cloud information distribution and the second neighborhood feature of the upper point cloud data, and determine the third change distribution based on the neighborhood information difference, wherein the neighborhood feature is the relative change of the neighborhood point cloud.

[0070] An abnormal transition information determining unit is configured to determine abnormal transition information of the first point cloud data based on the first transition distribution, the second transition distribution, and the third transition distribution.

[0071] Specifically, according to the principle of spatial position and feature homology. The consistency of spatial position means that when comparing the point cloud information distribution and the upper point cloud data, the points at the same spatial coordinate position will be matched. Feature homology requires that the features with the same properties or sources are compared. For example, if it is a feature about the color of the surface of an object, then only the specific homologous feature of color is compared. Next, a vertical mapping operation is performed. This means that in the vertical direction of the space, a one-to-one correspondence is directly established between the points in the point cloud information distribution at the current moment and the points at the same position in the (upper) point cloud data at the previous moment. Then, the difference between such corresponding points is quantified by differential measurement. The differential measurement involves calculating the difference in the position coordinates of the points, the change in the feature values, etc. For the position coordinates, the Euclidean distance is calculated to represent the size of the displacement. For feature values, such as changes in the RGB values ​​of the color, changes in the reflection intensity, etc., the degree of change is measured by direct subtraction or calculation of the absolute value of the difference. Through this vertical mapping and differential measurement calculation of a large number of points, all the difference results are combined to finally determine the first transition distribution.

[0072] Based on the key concept of point cloud approximation, the current point cloud information distribution and the previous point cloud data are gridded. Grid division is like building regular small boxes in space, which regularly divides the entire space area into many cells of the same size. When dividing, the density and distribution characteristics of the data are fully considered to ensure that the size and number of grids can reasonably cover and describe the characteristics of the point cloud. Then, in each identical grid domain, the dislocation mapping operation is performed. This is not strictly in accordance with the way of complete position correspondence, but within a certain tolerance range, to find points with approximate positions for matching and mapping. The purpose of this dislocation mapping is to deal with the slight deviations and uncertainties that may occur in the acquisition or processing of point cloud data. Then, differential measurement is performed. For the point pairs that complete the dislocation mapping in the same grid domain, the differences between them are calculated. These differences include the offset of the position, the degree of change of the eigenvalue, etc. By synthesizing and analyzing the differential measurement of all point pairs in the grid, the second transition distribution is finally determined.

[0073] Determine a fixed distance threshold. For example, consider points within 5 unit lengths from the current point as its neighborhood points. For a certain point A in the point cloud information distribution, take A as the center and calculate the distance between other points and A. Mark points with a distance less than 5 unit lengths as neighborhood points of A. Perform statistics and analysis on the features of these neighborhood points, such as calculating the average position, average color value, and variance of the position of the neighborhood points, to obtain the first neighborhood feature of point A. Similarly, for the corresponding point B in the upper point cloud data, determine its neighborhood points in the same way and calculate the second neighborhood feature. Then, compare the first neighborhood feature of point A with the second neighborhood feature of point B. For each feature, calculate the difference between them. For example, calculate the coordinate difference of the average position of the neighborhood points, the difference of the color value, the difference of the position variance, etc. Summarize and analyze the neighborhood information differences of all points. If a certain difference exceeds the preset threshold, it is considered that there is a significant change at this position. Comprehensively consider the changes of all points and finally determine the third change distribution. This distribution shows the pattern and degree of variation of point cloud data in space while taking into account neighborhood characteristics.

[0074] After receiving the results of the first, second and third transition distributions, an in-depth analysis is conducted from the perspective of coal mine safety. For the first transition distribution, the focus is on checking the changes in key parts such as coal mine tunnels and mining faces based on spatial position and feature homology. If the coal seam position or geological characteristics in a certain area change abnormally, beyond the range caused by normal mining, it may be a potential safety hazard. For the second transition distribution, special attention is paid to the changes in the spatial structure of the coal mine presented by the point cloud approximation and grid division. For example, if the displacement of the roof and floor or the misalignment mapping of the supporting structure is significantly abnormal in the same grid domain, and the difference measurement exceeds the safety threshold, this is likely to indicate the risk of mine pressure change or support failure. For the third transition distribution, the stability of the local area reflected by the relative changes of the neighborhood point cloud is carefully explored. For example, if the relative changes of coal seams and rock strata in a neighborhood show an abnormal concentration or dispersion trend, it may imply the danger of local collapse or outburst. Combining the results of these three distributions, a comprehensive analysis method suitable for coal mine safety assessment is adopted, combining multiple factors such as the geological conditions of the coal mine, mining progress, and equipment operation status to perform weighted average calculations. At the same time, a strict threshold range is set based on long-term accumulated coal mine safety data and expert experience. Once a location or area shows obvious abnormal changes in all three distributions, or the degree of change in one of the distributions seriously exceeds the safety threshold, it will be quickly determined as an abnormal change with coal mine safety risks. Through this rigorous and comprehensive analysis, the abnormal change information determination unit can accurately identify the abnormal change information in the first point cloud data of the coal mine, provide timely early warnings for coal mine safety production, and help take effective prevention and response measures to avoid accidents.

[0075] In a possible implementation, the lower-level monitoring configuration determination module 60 further includes:

[0076] A multi-level risk control monitoring point determination unit is used to determine a change prediction distribution, and based on the change prediction distribution, determine the multi-level risk control monitoring points.

[0077] A lower-level monitoring configuration determination unit, wherein the lower-level monitoring configuration determination unit performs scanning position and scanning parameter control feature configuration based on the multi-level risk control monitoring points to determine the lower-level monitoring configuration, wherein the lower-level monitoring configuration is identified by a monitoring trigger time.

[0078] A laser scanning management unit performs laser scanning management on the three-dimensional laser scanner based on the lower-level monitoring configuration.

[0079] Specifically, detailed geological information is obtained from geological surveys, such as the hardness of the strata, the distribution of rock cracks, the complexity of the geological structure, etc., and this information is quantified into specific values. Data from various real-time monitoring equipment in the mine are collected, such as stress values ​​at different locations, displacement of rock strata, gas concentration, etc. Then, the change prediction distribution is determined according to the following steps: Standardize the collected geological data to make different types of data comparable. Calculate the comprehensive risk index of each monitoring point, taking into account the geological information and real-time monitoring data. Calculate by assigning different weights to different data and then weighted summing them. Sort the monitoring points according to the comprehensive risk index. Using the sliding window method, several adjacent monitoring points are used as a window to calculate the average and standard deviation of the comprehensive risk index of the monitoring points in the window. According to the average and standard deviation, the monitoring area is divided into different risk levels, high risk, medium risk and low risk areas, so as to obtain the change prediction distribution. Next, multi-level risk control monitoring points are determined based on the change prediction distribution: In the high-risk area, several monitoring points with the highest comprehensive risk index are selected as the first-level risk control monitoring points. In the medium-risk area, a certain number of monitoring points are selected from high to low according to the comprehensive risk indicators as the second-level risk control monitoring points. In the low-risk area, relatively few monitoring points are selected as the third-level risk control monitoring points. Through the above steps, multi-level risk control monitoring points can be determined more accurately, providing effective support for safety monitoring and risk control of coal mines.

[0080] After receiving information from multi-level risk control monitoring points, it will determine the specific scanning location. It will expand a certain range outward with multi-level risk control monitoring points as the core to ensure full coverage of the areas that may be affected. For each determined scanning location, its spatial coordinates will be accurately marked so that subsequent scanning equipment can accurately locate it. In terms of the configuration of scanning parameter control features, it will be carefully set according to different risk levels and monitoring needs. For the first-level risk control monitoring point and its surrounding areas, a high-resolution scanning mode will be set to obtain more accurate detailed information; at the same time, the scanning frequency will be increased to ensure that small changes can be captured in time. For lower-level risk control monitoring points, the resolution and scanning frequency will be appropriately reduced to balance the accuracy of monitoring and resource consumption. In addition, other related parameter control features, such as the scanning angle range, laser intensity, etc., will be configured to adapt to different monitoring environments and targets. After completing the configuration of the scanning position and scanning parameter control features, the monitoring trigger time of each location is clearly marked. The determination of this time will take into account a variety of factors, including the urgency of the risk, the operating efficiency of the monitoring equipment, and the data processing capabilities. For higher-risk areas, a shorter trigger interval may be set to achieve more real-time monitoring; while for relatively stable areas, the trigger time can be appropriately extended to save resources. For example, for a coal mining area that is about to undergo large-scale mining, the first-level risk control monitoring point will be set to perform a high-resolution full scan every hour, while the surrounding second-level and third-level risk control monitoring points will be scanned every two hours and half a day respectively, and the scanning parameters will be adjusted accordingly to achieve the effect of ensuring close monitoring of key areas and rational use of resources.

[0081] According to the scanning position specified in the configuration, the scanner's rotation angle, translation direction and distance are precisely controlled to ensure that the scanner's laser beam can accurately illuminate the predetermined monitoring point. During the scanning process, the scanner parameters are adjusted strictly according to the scanning parameter control features set in the configuration. For example, if the configuration requires high-resolution scanning, the scanner's transmission frequency and receiving accuracy are correspondingly increased to obtain more detailed point cloud data. If a specific scanning angle range is required, the scanner's rotation range is precisely controlled to avoid missing or repeated scanning. At the same time, pay close attention to the setting of the monitoring trigger time. Before the trigger time is reached, the scanner is preheated, calibrated and initialized in advance to ensure that efficient and accurate scanning can be started immediately at the trigger time. After the scan is completed, the acquired data is promptly subjected to preliminary quality checks, such as data integrity, noise level, etc. For abnormal situations that may occur during the scanning process, such as scanner failure, data deviation caused by external interference, etc., the laser scanning management unit can quickly make judgments and take corresponding countermeasures. For example, automatically start a backup scanner, adjust scanning parameters to compensate for interference effects, or issue an alarm to notify maintenance personnel to handle on-site. Through comprehensive and meticulous management of the 3D laser scanner, the laser scanning management unit can ensure that the scanner operates efficiently and stably in accordance with the requirements of the lower-level monitoring configuration, providing a strong guarantee for obtaining accurate and reliable monitoring data.

[0082] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0083] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

[0084] This specification and drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. A coal mine safety mobile monitoring device based on scanning laser, characterized in that: The device comprises: A monitoring configuration information confirmation module, which is used to perform local segmentation and laser scanning configuration based on risk characteristics for the coal mine area space, and determine monitoring configuration information, including risk distribution characteristics and risk change characteristics, and the local segmentation standard is risk homology; A laser point cloud data determination module, which controls a three-dimensional laser scanner based on the monitoring configuration information to perform three-dimensional scanning and reception of the coal mine area space, perform point cloud splicing and integration, and determine laser point cloud data; A hybrid deformation model construction module, wherein the hybrid deformation model construction module is used to construct a hybrid deformation model, wherein the hybrid deformation model includes a plurality of parallel deformation decision branches guided by feature decision requirements, and uses an R-tree index as a point cloud allocation calling method for each decision branch; An abnormal change distribution positioning module, which performs branch data call and spatiotemporal change analysis on the laser point cloud data based on the mixed deformation model to locate abnormal change distribution, wherein the spatiotemporal change analysis dimension includes spatial change and temporal phase change, and the temporal phase change includes vertical change and dislocation change; A risk control and early warning module, which generates real-time risk information and performs risk control and early warning based on the abnormal change distribution; A lower-level monitoring configuration determination module, which performs change trend prediction based on the abnormal change distribution to determine the lower-level monitoring configuration; A safety monitoring and management module, which performs safety monitoring and management on the coal mine area space based on the real-time risk information and the lower-level monitoring configuration; The branch data calling and spatiotemporal change analysis of the laser point cloud data include: A point cloud information distribution determining unit, the point cloud information distribution determining unit is used to index and call the first point cloud data of the first deformation decision branch to determine the point cloud information distribution, wherein each distribution point includes a feature multi-tuple; A change information distribution determination unit, the change information distribution determination unit is used to traverse the point cloud information distribution, interact with the point cloud data and perform spatiotemporal change analysis to determine the change information distribution; An abnormal transition distribution positioning unit, the abnormal transition distribution positioning unit is used to traverse the transition information distribution, perform transition over-limit judgment, and locate the abnormal transition distribution, wherein the over-limit judgment includes spatiotemporal vector over-limit and trend over-limit, and the over-limit judgment standard is the coal mine safety standard; The interactively loading point cloud data and performing spatiotemporal change analysis include: A first transition distribution determining unit, which determines a first transition distribution by performing vertical mapping and difference measurement on the point cloud information distribution and the upper point cloud data based on spatial position and feature homology; A second transition distribution determining unit, which performs grid division on the point cloud information distribution and the upper point cloud data based on the point cloud approximation, performs misalignment mapping and difference measurement in the same grid domain, and determines a second transition distribution; A third change distribution determining unit, the third change distribution determining unit is used to spatially determine a first neighborhood feature of the point cloud information distribution and a second neighborhood feature of the upper point cloud data, and determine a third change distribution based on a neighborhood information difference, wherein the neighborhood feature is a relative change of the neighborhood point cloud; An abnormal transition information determining unit is configured to determine abnormal transition information of the first point cloud data based on the first transition distribution, the second transition distribution, and the third transition distribution.

2. The device according to claim 1, characterized in that After the laser point cloud data is determined, the method includes: A point cloud confidence interval setting unit, the point cloud confidence interval setting unit is used to set a point cloud confidence interval, the point cloud confidence interval includes a same-domain confidence interval and a cross-domain confidence interval, and is determined by a data migration confidence extreme value based on the point cloud spacing; An untrusted point cloud extraction unit, the untrusted point cloud extraction unit is used to traverse the laser point cloud data, traverse the point cloud confidence interval matching and verification, and extract untrusted point clouds; The valid point cloud data determining unit is used to compensate the untrusted point cloud and determine the valid point cloud data.

3. The device according to claim 2, characterized in that Compensating the untrusted point cloud includes: An untrusted point cloud determination unit, the untrusted point cloud determination unit is used to perform necessity determination on the untrusted point cloud; An untrusted point cloud vacancy unit, wherein the untrusted point cloud vacancy unit is used to leave the untrusted point cloud vacant if the necessity threshold is not met; An untrusted point cloud compensation unit, wherein the untrusted point cloud compensation unit is used to perform algorithm compensation on the untrusted point cloud if a first necessity threshold is met, wherein the compensation method includes mean interpolation or trend interpolation; An untrusted point cloud acquisition unit, wherein the untrusted point cloud acquisition unit is used to rescan and acquire the untrusted point cloud if a second necessity threshold is met, wherein the second necessity threshold is greater than the first necessity threshold.

4. The device according to claim 1, characterized in that Before the point cloud stitching and integration is performed, the following steps are included: A device coordinate system determining unit, the device coordinate system determining unit is used to determine the distribution position of the three-dimensional laser scanner, determine the device coordinate system with the distribution position as the center, and the device coordinate system corresponds to the distribution position one by one; A global coordinate system determining unit, wherein the global coordinate system determining unit determines a global coordinate system based on the coal mine regional space; a coordinate transformation relationship determining unit, wherein the coordinate transformation relationship determining unit determines a coordinate transformation relationship based on the global coordinate system and the device coordinate system; A point cloud stitching and integration unit is used to perform point cloud stitching and integration based on the coordinate transformation relationship.

5. The device according to claim 1, characterized in that The upper point cloud data is point cloud distribution data of a previous phase monitoring node, and the previous phase monitoring node includes at least one phase node.

6. The device according to claim 1, characterized in that The determining of the lower-level monitoring configuration includes: A multi-level risk control monitoring point determination unit, the multi-level risk control monitoring point determination unit is used to determine a change prediction distribution, and based on the change prediction distribution, determine a multi-level risk control monitoring point; A lower-level monitoring configuration determination unit, wherein the lower-level monitoring configuration determination unit performs scanning position and scanning parameter control feature configuration based on the multi-level wind control monitoring points to determine the lower-level monitoring configuration, wherein the lower-level monitoring configuration is identified by a monitoring trigger time; A laser scanning management unit performs laser scanning management on the three-dimensional laser scanner based on the lower-level monitoring configuration.

Images