Filling body top plate stability monitoring system and method for mining

Abstract

The invention discloses a filling body top plate stability monitoring system for mining. The system comprises a rectangular wooden baffle plate; filling materials; a copper grid mesh; the resistivity tester is used for monitoring the resistivity change condition of the self-sensing filling body material and automatically adjusting the test frequency and voltage according to the data monitored by the sensor module; the sensor module is used for monitoring the change of moisture content, temperature and component proportion in the self-sensing filling material; the central control module comprises a data collection unit, a data processing unit, a stability model unit and an alarm and early warning unit, and the data collection unit is responsible for collecting resistivity data monitored by the resistivity tester in real time and various data of geological exploration, field observation and numerical simulation; the data processing unit is responsible for processing the received resistivity data and various data, and the stability model unit is responsible for constructing a stability model and judging the stability state of the filling body roof according to the change condition of the resistivity.

Description

Technical Field

[0001] The present invention relates to the technical field of mining, and in particular to a filling body roof stability monitoring system and method for mining. Background Art

[0002] In soft rock mining areas, the backfill roof stability monitoring project SKLCRSM-XJIE24KF03 (jointly funded by the State Key Laboratory of Coal Resources and Safe Mining (China University of Mining and Technology) and Xinjiang Engineering College) carries a vital scientific research mission. The project focuses on the field of coal resources and safe mining, and is committed to tracking the stability of the backfill roof in real time through advanced monitoring technology. This initiative can keenly capture potential safety hazards, allowing mine safety managers to respond quickly and take preventive measures to avoid risks, thereby ensuring that the safe production of mines is not threatened.

[0003] After searching, the invention patent with Chinese patent number CN116148014A discloses a filling roof stability monitoring system and method for mining, which belongs to the field of mining filling mining technology. The filling roof stability monitoring system for mining includes a rectangular wooden baffle, a copper grid, vertical ribs, graphite powder, tailings, cement, resistor clamps, a resistivity tester, a central control unit and a computer. The monitoring method is to build a rectangular wooden baffle in the mining access after the mining is completed, fill tailings, cement, and graphite powder to form a self-sensing filling material, insert a copper grid into the self-sensing filling material, and after the self-sensing filling material is formed, fill tailings and cement into the mining access to form a tailings cemented filling body consistent with the access. When the mining access reaches the next layer, the resistivity tester is used to monitor the resistivity change of the self-sensing filling material of the previous layer, thereby inferring the stability state of the filling roof.

[0004] However, during the use of the above patent, in soft rock mining areas, due to the softening and plastic deformation of the rock, the stability monitoring of the filling roof becomes more complicated. It may not be able to accurately capture these tiny deformation and damage processes, resulting in the monitoring effect being less than expected. Therefore, a filling roof stability monitoring system and method for mining is proposed. Summary of the invention

[0005] The purpose of the present invention is to solve the problem in the prior art that in soft rock mining areas, the stability monitoring of the filling roof becomes more complicated due to the softening and plastic deformation of the rock. The existing technical solutions may not be able to accurately capture these tiny deformation and destruction processes, resulting in the monitoring effect being less than expected. A filling roof stability monitoring system and method for mining is proposed.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] A filling body roof stability monitoring system for mining, comprising:

[0008] Rectangular wooden baffle: used to build enclosures in the mining access road and provide support for the filling material;

[0009] Filling material: tailings, cement and graphite powder are mixed to form a self-sensing filling material for monitoring roof stability;

[0010] Copper grid: inserted into the self-sensing filling material, forming a monitoring network together with the filling material;

[0011] Resistivity tester: used to monitor the resistivity changes of self-sensing filling materials. According to the data monitored by the sensor module, the test frequency and voltage are automatically adjusted to obtain more stable and accurate resistivity data;

[0012] Sensor module: used to monitor the changes in moisture content, temperature and composition ratio in the self-sensing filling material;

[0013] The central control module includes a data collection unit, a data processing unit, a stability model unit and an alarm and early warning unit. The data collection unit is responsible for collecting the resistivity data monitored in real time by the resistivity tester and various data from geological exploration, field observation and numerical simulation. The data processing unit is responsible for processing the received resistivity data and various data. Through principal component analysis (PCA), high-dimensional data is converted into low-dimensional data representation to reduce the amount of data calculation and retain the main information and features in the data. The stability model unit is responsible for combining the support vector machine SVM with the discrete element analysis DEA to construct a stability model for judging the stability state of the filling body top plate. According to the change of resistivity and combined with the stability model, the stability state of the filling body top plate is judged. When the stability of the filling body top plate is abnormal, the alarm and early warning unit automatically alarms or warns, and conducts risk assessment based on the geological exploration results and the deformation characteristics of the filling body top plate.

[0014] The above technical solution further includes:

[0015] Furthermore, the resistivity tester includes a sensor unit, a control unit, a test unit, and an adaptive adjustment unit. The sensor unit is responsible for collecting the resistivity data of the filling material in real time, and transmitting the collected resistivity data to the control unit and the data collection unit. The control unit receives the resistivity data transmitted by the sensor unit, adjusts the test frequency and voltage, and sends the adjustment instruction to the test unit. The test unit adjusts the test conditions (such as test frequency, voltage, etc.) according to the instructions of the control unit, and controls the sensor unit to measure the resistivity of the filling material. The adaptive adjustment unit automatically adjusts the test frequency and voltage according to the changes in the moisture content, temperature and component ratio in the filling material, receives and responds to the adjustment instructions of the control unit in real time, and sends the adjusted test parameters to the test unit according to the real-time status of the filling material.

[0016] Furthermore, the data collection unit includes a resistivity data receiving subunit, a geological exploration data integration subunit, a field observation data collection subunit and a numerical simulation data receiving subunit. The resistivity data receiving subunit is responsible for receiving the resistivity data transmitted by the resistivity tester. The resistivity data receiving subunit establishes a communication connection with the resistivity tester to receive and store the resistivity data in real time. The geological exploration data integration subunit integrates various data generated during the geological exploration process, including geological structure, rock layer distribution, groundwater conditions and other related data. These data are helpful to understand the geological background of the mine and provide stability judgment. Provide reference to form a complete geological exploration data set. The geological exploration data integration subunit establishes a data interface with the geological exploration department or system to receive and integrate geological exploration data. The field observation data collection subunit collects intuitive information such as deformation of the filling body top plate and crack development observed on site. The field observation data collection subunit directly obtains data through field observation equipment or personnel. The numerical simulation data receiving subunit receives data obtained by numerical simulation methods, such as stress distribution, displacement changes, etc. The numerical simulation data receiving subunit establishes a data interface with the numerical simulation system to receive and store numerical simulation data.

[0017] Further, the data processing unit uses principal component analysis (PCA) to convert high-dimensional data into low-dimensional data representation;

[0018] The digitized data is set to M samples {X 1 , X 2 , ..., X M ,}, each sample has N-dimensional features Each feature X j All have their own eigenvalues;

[0019] Decentralize all features, that is, remove the mean, and find the average value of each feature;

[0020] For all samples, each feature is subtracted from its own mean, where the respective means are

[0021] Find the covariance matrix after decentralization The diagonal lines are the variances of features X1 and X2, and the off-diagonal lines are the covariances. The calculation formula for cov(X1, X1) is This results in the covariance matrix C of the M samples under these N-dimensional features;

[0022] After obtaining the covariance matrix, its eigenvalues ​​and their corresponding eigenvectors are calculated according to the characteristic equation Cμ=λμ, where λ is the eigenvalue and μ is the corresponding eigenvector. The largest first k eigenvalues ​​and corresponding eigenvectors are selected for projection. Projection is the process of dimensionality reduction, which reduces the original features from high dimensions to low dimensions. After dimensionality reduction, a large amount of redundant information is removed and at least 85% of the original information is retained.

[0023] Furthermore, the stability model unit includes a model construction and update subunit, a stability assessment subunit and an anomaly detection subunit. The model construction and update subunit uses the processed data to construct a stability model, and updates the model parameters periodically or in real time, and passes the constructed model to the stability assessment subunit. The stability assessment subunit evaluates the stability state of the top plate of the filling body based on the real-time monitored data and the constructed stability model, and passes the evaluation result to the anomaly detection subunit. The anomaly detection subunit compares the real-time monitored data with the expected result of the stability model, detects abnormal situations, and passes the abnormal information to the alarm and early warning unit.

[0024] Furthermore, the model building and updating subunit combines support vector machine SVM with discrete element analysis DEA to build a stability model for judging the stability state of the top plate of the filling body, and the specific steps are:

[0025] Data collection and preprocessing: Collect historical data, including stability records of the filling body roof, resistivity data, groundwater content data, and geological exploration data, and perform cleaning, format conversion, normalization, and other processing on the collected historical data to ensure data quality and consistency;

[0026] Support vector machine (SVM) model training: Select features related to the stability of the filling roof from the collected historical data, such as resistivity, groundwater content, etc., use the support vector machine (SVM) algorithm, train the model based on historical data, and learn the complex relationship between the stability of the filling roof and the features. The basic formula of the support vector machine is:

[0027]

[0028] sty i (w·x i +b)≥1-ξ i ,ξ i ≥0,i=1,2,...,n

[0029] Among them, w is the weight vector, b is the bias term, C is the regularization parameter, ξ i is the slack variable, y i is the sample label, x i is the sample characteristic;

[0030] Discrete element analysis (DEA) model construction: According to the physical characteristics and geological conditions of the filling body, the discrete element analysis model is initialized to simulate the interaction and overall stability of the filling body particles, taking into account physical factors such as contact and friction between particles;

[0031] Model combination and parameter adjustment: The output of the support vector machine (SVM) model is used as the input of the discrete element analysis (DEA) model or as the basis for parameter adjustment. According to the prediction results of the support vector machine (SVM) model, the parameters of the discrete element analysis model are adjusted to improve the accuracy and adaptability of the model.

[0032] Furthermore, the model is combined with parameter adjustment, specifically the following steps:

[0033] Model combination:

[0034] Support vector machine (SVM) model prediction: Use the trained support vector machine (SVM) model to predict new data samples, which may include multiple factors related to the stability of the filling roof, such as resistivity and groundwater content;

[0035] Prediction result processing: The new output is usually a probability value or classification label (such as stable / unstable). According to specific needs, this output is processed, such as converted into a specific parameter value or range;

[0036] As discrete element analysis DEA model input: The processed support vector machine SVM model prediction results are used as the input of the discrete element analysis DEA model, which can be one or more parameters to simulate the interaction and overall stability of the filling particles;

[0037] Parameter adjustment:

[0038] Analyze the prediction results of the support vector machine (SVM) model: According to the prediction results of the support vector machine (SVM) model, analyze the changing trend and potential risks of the stability of the filling roof.

[0039] Determine the adjustment direction: Based on the analysis results, determine the adjusted discrete element analysis DEA model parameters and their adjustment direction (such as increasing / decreasing the friction coefficient, adjusting the contact stiffness, etc.);

[0040] Implement parameter adjustment: In the discrete element analysis DEA model, adjust the parameters according to the determined adjustment direction and amplitude;

[0041] Verify the adjustment effect: Through simulation analysis or actual observation, verify whether the model after parameter adjustment reflects the stability of the top plate of the filling body.

[0042] A monitoring method for a filling body roof stability monitoring system for mining, comprising the following steps:

[0043] Build baffles: After the mining access is completed, build rectangular wooden baffles to enclose the filling area;

[0044] Filling material: Mix tailings, cement and graphite powder and fill them into the area enclosed by the baffle to form a self-sensing filling material;

[0045] Inserting copper grid: inserting copper grid into the self-sensing filling material to form a monitoring network together with the filling material;

[0046] Forming tailings cemented filling body: After the self-sensing filling body material is formed, fill it with tailings and cement to form a tailings cemented filling body consistent with the approach;

[0047] Resistivity monitoring: When the mining route reaches the next layer, the resistivity tester is used to monitor the resistivity change of the self-sensing filling material in the previous layer. At the same time, according to the geological characteristics of the soft rock mining area and the deformation characteristics of the filling roof, the parameter settings of the monitoring system, such as the sensitivity of the sensor, sampling frequency and data processing algorithm, are dynamically adjusted to meet the monitoring needs under different geological conditions;

[0048] Data acquisition: The resistivity tester transmits the real-time monitored resistivity data to the central control module;

[0049] Data processing: The central control module processes and analyzes the received data, including preprocessing operations such as data filtering and denoising. It converts high-dimensional data into low-dimensional data representation through principal component analysis (PCA) to reduce the amount of data calculation and retain the main information and features in the data.

[0050] Stability judgment: According to the change of resistivity, a stability model for judging the stability state of the filling body roof is constructed by combining support vector machine SVM with discrete element analysis DEA to judge the stability state of the filling body roof;

[0051] Alarm and early warning: When abnormal stability of the filling body roof is monitored, the alarm and early warning unit will automatically alarm or warn, and conduct risk assessment based on geological exploration results and the deformation characteristics of the filling body roof. By formulating targeted safety measures and emergency plans, the probability and impact of safety accidents can be reduced.

[0052] The present invention has the following beneficial effects:

[0053] 1. In the present invention, in view of the diversity of geological characteristics of soft rock mining areas and the complexity of deformation of the filling body roof, the monitoring parameter settings are intelligently adjusted, such as the sensitivity of the resistivity tester, the sampling frequency and the data processing algorithm, etc. This enables the monitoring system to better adapt to the monitoring needs under different geological conditions and improve the reliability and accuracy of the monitoring results.

[0054] 2. In the present invention, by integrating resistivity monitoring, data analysis and early warning technology, the resistivity change of the filling material can be monitored in real time, and the stability state of the filling roof can be inferred accordingly. This monitoring method is not only highly sensitive, but also can adapt to the complex and changeable geological conditions of soft rock mining areas, significantly improving the accuracy of monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 This is a system block diagram of a filling body roof stability monitoring system and method for mining proposed by the present invention;

[0056] Figure 2 The present invention is a flowchart of a filling body roof stability monitoring system and method for mining. DETAILED DESCRIPTION

[0057] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0058] See also Figure 1-Figure 2 As shown, the present invention is a filling body roof stability monitoring system for mining, comprising:

[0059] Rectangular wooden baffle: used to build enclosures in the mining access road and provide support for the filling material;

[0060] Filling material: tailings, cement and graphite powder are mixed to form a self-sensing filling material for monitoring roof stability;

[0061] Copper grid: inserted into the self-sensing filling material, forming a monitoring network together with the filling material;

[0062] Resistivity tester: used to monitor the resistivity changes of self-sensing filling materials. According to the data monitored by the sensor module, the test frequency and voltage are automatically adjusted to obtain more stable and accurate resistivity data;

[0063] Sensor module: used to monitor the changes in moisture content, temperature and composition ratio in the self-sensing filling material;

[0064] The central control module includes a data collection unit, a data processing unit, a stability model unit and an alarm and early warning unit. The data collection unit is responsible for collecting the resistivity data monitored in real time by the resistivity tester and various data from geological exploration, field observation and numerical simulation. The data processing unit is responsible for processing the received resistivity data and various data. Through principal component analysis (PCA), high-dimensional data is converted into low-dimensional data representation to reduce the amount of data calculation and retain the main information and features in the data. The stability model unit is responsible for combining the support vector machine SVM with the discrete element analysis DEA to construct a stability model for judging the stability state of the filling body top plate. According to the change of resistivity and combined with the stability model, the stability state of the filling body top plate is judged. When the stability of the filling body top plate is abnormal, the alarm and early warning unit automatically alarms or warns, and conducts risk assessment based on the geological exploration results and the deformation characteristics of the filling body top plate.

[0065] In one embodiment, for the above-mentioned resistivity tester, the resistivity tester includes a sensor unit, a control unit, a test unit, and an adaptive adjustment unit. The sensor unit is responsible for real-time acquisition of resistivity data of the filling material, and transmits the acquired resistivity data to the control unit and the data collection unit. The control unit receives the resistivity data transmitted by the sensor unit, adjusts the test frequency and voltage, and sends the adjustment instruction to the test unit. The test unit adjusts the test conditions (such as test frequency, voltage, etc.) according to the instruction of the control unit, and controls the sensor unit to measure the resistivity of the filling material. The adaptive adjustment unit automatically adjusts the test frequency and voltage according to the changes in the moisture content, temperature and component ratio in the filling material, receives and responds to the adjustment instruction of the control unit in real time, and sends the adjusted test parameters to the test unit according to the real-time status of the filling material.

[0066] In one embodiment, for the above-mentioned data collection unit, the data collection unit includes a resistivity data receiving subunit, a geological exploration data integration subunit, a field observation data collection subunit and a numerical simulation data receiving subunit. The resistivity data receiving subunit is responsible for receiving the resistivity data transmitted by the resistivity tester. The resistivity data receiving subunit establishes a communication connection with the resistivity tester to receive and store the resistivity data in real time. The geological exploration data integration subunit integrates various data generated during the geological exploration process, including geological structure, rock layer distribution, groundwater conditions and other related data. These data are helpful to understand the geological background of the mine. The scene provides a reference for stability judgment and forms a complete geological exploration data set. The geological exploration data integration subunit establishes a data interface with the geological exploration department or system to receive and integrate geological exploration data. The field observation data collection subunit collects intuitive information such as the top plate deformation and crack development of the filling body observed on site. The field observation data collection subunit directly obtains data through field observation equipment or personnel. The numerical simulation data receiving subunit receives data obtained by numerical simulation methods, such as stress distribution, displacement change, etc. The numerical simulation data receiving subunit establishes a data interface with the numerical simulation system to receive and store numerical simulation data.

[0067] In one embodiment, for the above-mentioned data processing unit, the data processing unit uses principal component analysis (PCA) to convert high-dimensional data into a specific step of low-dimensional data representation;

[0068] The digitized data is set to M samples {X 1 , X 2 , ..., X M, each sample has N-dimensional features Each feature X j All have their own eigenvalues;

[0069] Decentralize all features, that is, remove the mean, and find the average value of each feature;

[0070] For all samples, each feature is subtracted from its own mean, where the respective means are

[0071] Find the covariance matrix after decentralization The diagonal lines are the variances of features X1 and X2, and the off-diagonal lines are the covariances. The calculation formula for cov(X1, X1) is This results in the covariance matrix C of the M samples under these N-dimensional features;

[0072] After obtaining the covariance matrix, its eigenvalues ​​and their corresponding eigenvectors are calculated according to the characteristic equation Cμ=λμ, where λ is the eigenvalue and μ is the corresponding eigenvector. The largest first k eigenvalues ​​and corresponding eigenvectors are selected for projection. Projection is the process of dimensionality reduction, which reduces the original features from high dimensions to low dimensions. After dimensionality reduction, a large amount of redundant information is removed and at least 85% of the original information is retained.

[0073] In one embodiment, for the above-mentioned stability model unit, the stability model unit includes a model construction and update subunit, a stability assessment subunit and an anomaly detection subunit. The model construction and update subunit uses the processed data to construct a stability model, and updates the model parameters regularly or in real time, and passes the constructed model to the stability assessment subunit. The stability assessment subunit evaluates the stability state of the top plate of the filling body based on the real-time monitored data and the constructed stability model, and passes the evaluation result to the anomaly detection subunit. The anomaly detection subunit compares the real-time monitored data with the expected result of the stability model, detects abnormal situations, and passes the abnormal information to the alarm and early warning unit.

[0074] In one embodiment, for the above-mentioned model building and updating subunit, the model building and updating subunit combines support vector machine SVM with discrete element analysis DEA to build a stability model for judging the stability state of the filling body top plate, and the specific steps are:

[0075] Data collection and preprocessing: Collect historical data, including stability records of the filling body roof, resistivity data, groundwater content data, and geological exploration data, and perform cleaning, format conversion, normalization, and other processing on the collected historical data to ensure data quality and consistency;

[0076] Support vector machine (SVM) model training: Select features related to the stability of the filling roof from the collected historical data, such as resistivity, groundwater content, etc., use the support vector machine (SVM) algorithm, train the model based on historical data, and learn the complex relationship between the stability of the filling roof and the features. The basic formula of the support vector machine is:

[0077]

[0078] sty i (w·x i +b)≥1-ξ i ,ξ i ≥0,i=1,2,...,n

[0079] Among them, w is the weight vector, b is the bias term, C is the regularization parameter, ξ i is the slack variable, yi is the sample label, x i is the sample characteristic;

[0080] Discrete element analysis (DEA) model construction: According to the physical characteristics and geological conditions of the filling body, the discrete element analysis model is initialized to simulate the interaction and overall stability of the filling body particles, taking into account physical factors such as contact and friction between particles;

[0081] Model combination and parameter adjustment: The output of the support vector machine (SVM) model is used as the input of the discrete element analysis (DEA) model or as the basis for parameter adjustment. According to the prediction results of the support vector machine (SVM) model, the parameters of the discrete element analysis model are adjusted to improve the accuracy and adaptability of the model.

[0082] Assume we have a backfill roof stability monitoring project in a soft rock mine and have collected the following historical data:

[0083]

[0084] The stability state is 1 for stable and 0 for unstable.

[0085] Data preprocessing: Normalize the data so that the characteristic values ​​such as resistivity and groundwater content are in the range of [0,1].

[0086] Support vector machine (SVM) model training: Resistivity and groundwater content were selected as features, stability status as labels, and the model was trained using the SVM algorithm.

[0087] Discrete element analysis model construction: According to the physical characteristics and geological conditions of the filling body, a discrete element analysis model is constructed to simulate the interaction between particles and the overall stability.

[0088] Model combination and parameter adjustment: The prediction results of the SVM model (such as stability probability) are used as the input of the discrete element analysis model or as the basis for parameter adjustment. For example, if the SVM model predicts that the stability probability of a certain area is low, the parameters such as the inter-particle friction coefficient or contact stiffness in the discrete element analysis model can be appropriately adjusted to improve the prediction accuracy of the model.

[0089] In one embodiment, for the above-mentioned model combination and parameter adjustment, the model combination and parameter adjustment include the following specific steps:

[0090] Model combination:

[0091] Support vector machine (SVM) model prediction: Use the trained support vector machine (SVM) model to predict new data samples, which may include multiple factors related to the stability of the filling roof, such as resistivity and groundwater content;

[0092] Prediction result processing: The new output is usually a probability value or classification label (such as stable / unstable). According to specific needs, this output is processed, such as converted into a specific parameter value or range;

[0093] As discrete element analysis DEA model input: The processed support vector machine SVM model prediction results are used as the input of the discrete element analysis DEA model, which can be one or more parameters to simulate the interaction and overall stability of the filling particles;

[0094] Example description:

[0095] Suppose we have an SVM model that predicts the stability of the fill roof based on resistivity and groundwater content. For a new data sample (resistivity = 120Ω·m, groundwater content = 10%), the SVM model predicts that its stability probability is 0.7 (i.e., 70% probability of stability).

[0096] In the DEA model, we may have a parameter "stability coefficient", which affects the interaction force between particles and the overall stability. We can map the prediction results of SVM (such as stability probability) to the "stability coefficient". For example, when the stability probability is greater than 0.5, the "stability coefficient" is set to 1.0 (indicating more stable), and when it is less than 0.5, it is set to 0.8 (indicating less stable). In this example, we set the "stability coefficient" to 1.0 as the input of the DEA model.

[0097] Parameter adjustment:

[0098] Analyze the prediction results of the support vector machine (SVM) model: According to the prediction results of the support vector machine (SVM) model, analyze the changing trend and potential risks of the stability of the filling roof.

[0099] Determine the adjustment direction: Based on the analysis results, determine the adjusted discrete element analysis DEA model parameters and their adjustment direction (such as increasing / decreasing the friction coefficient, adjusting the contact stiffness, etc.);

[0100] Implement parameter adjustment: In the discrete element analysis DEA model, adjust the parameters according to the determined adjustment direction and amplitude;

[0101] Verify the adjustment effect: Through simulation analysis or actual observation, verify whether the model after parameter adjustment reflects the stability of the filling body roof;

[0102] Continuing with the example, suppose we simulated the interaction and overall stability of the filler particles in the DEA model and found that the simulation results deviated from the SVM prediction results. After analysis, we believe that this may be caused by improper setting of the friction coefficient between the particles.

[0103] To adjust this parameter, we can gradually adjust the value of the friction coefficient according to the degree of deviation between the SVM prediction results (such as stability probability) and the simulation results of the DEA model. For example, if the SVM predicts high stability, but the DEA simulation results show low stability, we can try to increase the friction coefficient to enhance the interaction between particles.

[0104] After the adjustment is completed, we run the DEA model again for simulation analysis and compare the simulation results before and after the adjustment. If the adjusted model more accurately reflects the stability of the backfill roof, then we believe that this parameter adjustment is effective.

[0105] A monitoring method for a filling body roof stability monitoring system for mining, comprising the following steps:

[0106] Build baffles: After the mining access is completed, build rectangular wooden baffles to enclose the filling area;

[0107] Filling material: Mix tailings, cement and graphite powder and fill them into the area enclosed by the baffle to form a self-sensing filling material;

[0108] Inserting copper grid: inserting copper grid into the self-sensing filling material to form a monitoring network together with the filling material;

[0109] Forming tailings cemented filling body: After the self-sensing filling body material is formed, fill it with tailings and cement to form a tailings cemented filling body consistent with the approach;

[0110] Resistivity monitoring: When the mining route reaches the next layer, the resistivity tester is used to monitor the resistivity change of the self-sensing filling material in the previous layer. At the same time, according to the geological characteristics of the soft rock mining area and the deformation characteristics of the filling roof, the parameter settings of the monitoring system, such as the sensitivity of the sensor, sampling frequency and data processing algorithm, are dynamically adjusted to meet the monitoring needs under different geological conditions;

[0111] Data acquisition: The resistivity tester transmits the real-time monitored resistivity data to the central control module;

[0112] Data processing: The central control module processes and analyzes the received data, including preprocessing operations such as data filtering and denoising. It converts high-dimensional data into low-dimensional data representation through principal component analysis (PCA) to reduce the amount of data calculation and retain the main information and features in the data.

[0113] Stability judgment: According to the change of resistivity, a stability model for judging the stability state of the filling body roof is constructed by combining support vector machine SVM with discrete element analysis DEA to judge the stability state of the filling body roof;

[0114] Alarm and early warning: When abnormal stability of the filling body roof is monitored, the alarm and early warning unit will automatically alarm or warn, and conduct risk assessment based on geological exploration results and the deformation characteristics of the filling body roof. By formulating targeted safety measures and emergency plans, the probability and impact of safety accidents can be reduced.

[0115] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A filling body roof stability monitoring system for mining, characterized in that: include: Rectangular wooden baffle: used to build enclosures in the mining access road and provide support for the filling material; Filling material: tailings, cement and graphite powder are mixed to form a self-sensing filling material for monitoring roof stability; Copper grid: inserted into the self-sensing filling material, forming a monitoring network together with the filling material; Resistivity tester: used to monitor the resistivity changes of self-sensing filling materials, and automatically adjust the test frequency and voltage according to the data monitored by the sensor module; Sensor module: used to monitor the changes in moisture content, temperature and composition ratio in the self-sensing filling material; The central control module includes a data collection unit, a data processing unit, a stability model unit and an alarm and early warning unit. The data collection unit is responsible for collecting the resistivity data monitored in real time by the resistivity tester and various data from geological exploration, field observation and numerical simulation. The data processing unit is responsible for processing the received resistivity data and various data, and converting high-dimensional data into low-dimensional data representation through principal component analysis. The stability model unit is responsible for combining the support vector machine with the discrete element analysis to construct a stability model for judging the stability state of the filling body top plate. According to the change of resistivity and combined with the stability model, the stability state of the filling body top plate is judged. When the stability of the filling body top plate is abnormal, the alarm and early warning unit automatically alarms or warns, and conducts risk assessment based on the geological exploration results and the deformation characteristics of the filling body top plate.

2. A filling body roof stability monitoring system for mining according to claim 1, characterized in that: The resistivity tester includes a sensor unit, a control unit, a test unit, and an adaptive adjustment unit. The sensor unit is responsible for collecting resistivity data of the filling material in real time, and transmitting the collected resistivity data to the control unit and the data collection unit. The control unit receives the resistivity data transmitted by the sensor unit, adjusts the test frequency and voltage, and sends the adjustment instruction to the test unit. The test unit adjusts the test conditions according to the instruction of the control unit, and controls the sensor unit to measure the resistivity of the filling material. The adaptive adjustment unit automatically adjusts the test frequency and voltage according to the changes in the moisture content, temperature and component ratio in the filling material, receives and responds to the adjustment instruction of the control unit in real time, and sends the adjusted test parameters to the test unit according to the real-time status of the filling material.

3. A filling body roof stability monitoring system for mining according to claim 1, characterized in that: The data collection unit includes a resistivity data receiving subunit, a geological exploration data integration subunit, a field observation data collection subunit and a numerical simulation data receiving subunit. The resistivity data receiving subunit is responsible for receiving the resistivity data transmitted by the resistivity tester. The resistivity data receiving subunit establishes a communication connection with the resistivity tester, receives and stores the resistivity data in real time, the geological exploration data integration subunit integrates various data generated during the geological exploration process to form a complete geological exploration data set, the geological exploration data integration subunit establishes a data interface with the geological exploration department or system, receives and integrates the geological exploration data, the field observation data collection subunit collects the intuitive information of the filling body top observed on site, the field observation data collection subunit directly obtains data through field observation equipment or personnel, the numerical simulation data receiving subunit receives data obtained by numerical simulation method, and the numerical simulation data receiving subunit establishes a data interface with the numerical simulation system to receive and store numerical simulation data.

4. A filling body roof stability monitoring system for mining according to claim 1, characterized in that: The data processing unit uses principal component analysis to convert high-dimensional data into low-dimensional data representation; The digitized data is set to M samples {X 1 , X 2 , ..., X M ,}, each sample has N-dimensional features Each feature X j All have their own eigenvalues; Decentralize all features, that is, remove the mean, and find the average value of each feature; For all samples, each feature is subtracted from its own mean, where the respective means are Find the covariance matrix after decentralization The diagonal lines are the variances of features X1 and X2, and the off-diagonal lines are the covariances. The calculation formula for cov(X1, X1) is This results in the covariance matrix C of the M samples under these N-dimensional features; After obtaining the covariance matrix, its eigenvalues ​​and their corresponding eigenvectors are calculated according to the characteristic equation Cμ=λμ, where λ is the eigenvalue and μ is the corresponding eigenvector. The largest first k eigenvalues ​​and corresponding eigenvectors are selected for projection. Projection is the process of dimensionality reduction, which reduces the original features from high dimensions to low dimensions. After dimensionality reduction, a large amount of redundant information is removed and at least 85% of the original information is retained.

5. The filling body roof stability monitoring system for mining according to claim 1 is characterized in that: The stability model unit includes a model construction and update subunit, a stability assessment subunit and an anomaly detection subunit. The model construction and update subunit uses the processed data to construct a stability model, and updates the model parameters regularly or in real time, and passes the constructed model to the stability assessment subunit. The stability assessment subunit evaluates the stability state of the top plate of the filling body based on the real-time monitored data and the constructed stability model, and passes the evaluation result to the anomaly detection subunit. The anomaly detection subunit compares the real-time monitored data with the expected result of the stability model, detects abnormal situations, and passes the abnormal information to the alarm and early warning unit.

6. A filling body roof stability monitoring system for mining according to claim 5, characterized in that: The model building and updating subunit combines the support vector machine with the discrete element analysis DEA to build a stability model for judging the stability state of the filling body roof, and the specific steps are as follows: Data collection and preprocessing: Collect historical data, including stability records of the filling body roof, resistivity data, groundwater content data and geological exploration data, and process the collected historical data; Support vector machine model training: Select features related to the stability of the filling roof from the collected historical data, use the support vector machine (SVM) algorithm, train the model based on historical data, and learn the complex relationship between the stability of the filling roof and the features. The basic formula of the support vector machine is: s.t.y i (w·x i +b)≥1-ξ i ,ξ i ≥0,i=1,2,...,n Among them, w is the weight vector, b is the bias term, C is the regularization parameter, ξ i is the slack variable, y i is the sample label, x i is the sample characteristic; Discrete element analysis model construction: According to the physical characteristics and geological conditions of the filling body, the discrete element analysis model is initialized to simulate the interaction and overall stability of the filling body particles; Model combination and parameter adjustment: The output of the support vector machine model is used as the input of the discrete element analysis model or the basis for parameter adjustment. The parameters of the discrete element analysis model are adjusted according to the prediction results of the support vector machine model.

7. A filling body roof stability monitoring system for mining according to claim 6, characterized in that: The model combination and parameter adjustment, specific steps: Model combination: Support vector machine model prediction: Use the trained support vector machine model to predict new data samples; Prediction result processing: Process this output according to specific needs; As input to discrete element analysis model: the processed support vector machine model prediction results are used as input to the discrete element analysis model; Parameter adjustment: Analyze the prediction results of the support vector machine model: Analyze the changing trend and potential risks of the stability of the filling roof based on the prediction results of the support vector machine model; Determine the adjustment direction: Based on the analysis results, determine the discrete element analysis model parameters to be adjusted and their adjustment direction; Implement parameter adjustment: In the discrete element analysis model, adjust the parameters according to the determined adjustment direction and amplitude; Verify the adjustment effect: Through simulation analysis or actual observation, verify whether the model after parameter adjustment reflects the stability of the top plate of the filling body.

8. The monitoring method of a filling body roof stability monitoring system for mining according to claim 1 is characterized in that: The following steps are involved: Build baffles: After the mining access is completed, build rectangular wooden baffles to enclose the filling area; Filling material: Mix tailings, cement and graphite powder and fill them into the area enclosed by the baffle to form a self-sensing filling material; Inserting copper grid: inserting copper grid into the self-sensing filling material to form a monitoring network together with the filling material; Forming tailings cemented filling body: After the self-sensing filling body material is formed, fill it with tailings and cement to form a tailings cemented filling body consistent with the approach; Resistivity monitoring: When the mining route reaches the next layer, the resistivity tester is used to monitor the resistivity change of the self-sensing filling material in the previous layer. At the same time, according to the geological characteristics of the soft rock mining area and the deformation characteristics of the filling roof, the parameter settings of the monitoring system are dynamically adjusted; Data acquisition: The resistivity tester transmits the real-time monitored resistivity data to the central control module; Data processing: The central control module processes and analyzes the received data and converts high-dimensional data into low-dimensional data representation through principal component analysis; Stability judgment: According to the change of resistivity, a stability model for judging the stability state of the filling body roof is constructed by combining support vector machine and discrete element analysis to judge the stability state of the filling body roof; Alarm and early warning: When abnormal stability of the filling body roof is monitored, the alarm and early warning unit will automatically alarm or warn, and conduct risk assessment based on geological exploration results and deformation characteristics of the filling body roof.

Images