Risk early warning method, system and device for fully-mechanized coal mining of coal mine and storage medium

Abstract

The invention discloses a risk early warning method, system and device for fully mechanized coal mining of a coal mine and a storage medium, and realizes real-time monitoring and accurate control of the state of a bracket through an intelligent monitoring, early warning and processing mechanism, thereby effectively protecting the life, health and safety of workers. The method comprises the steps of extracting target features according to historical data; a risk early warning model is trained according to the target features to obtain a target risk early warning model, and the target risk early warning model is used for predicting risk information according to the stress state of the support; collecting working data in real time, wherein the working data comprises working face top plate pressure data, bracket supporting stress state data and bottom plate condition data; inputting the working data into the target risk early warning model to obtain stress state information of the support, and predicting risk information according to the stress state information; and determining early warning levels according to the risk information, wherein the early warning levels comprise primary early warning, intermediate early warning and advanced early warning.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a risk early warning method, system, device and storage medium for comprehensive mining of coal mines. Background Art

[0002] In the fully mechanized mining face of a coal mine, the support is the core equipment for supporting the roof and maintaining the safety of the working space. Its stability and reliability are directly related to the safety and efficiency of coal mine production. The support adapts to the mining progress of the coal seam and the changes in the roof through lifting and moving to ensure the stability of the working space and the safety of the operators.

[0003] However, in the actual production process, the comprehensive mining face of coal mines often faces complex and changeable geological conditions, such as high roof pressure, soft bottom plate, large coal seam inclination, etc. These unfavorable factors increase the stress burden of the support, making the support "crushed to death" due to excessive force during the support process. The so-called "crushed to death" means that the support cannot be raised or lowered or moved normally, resulting in the inability to continue production operations. In severe cases, it may even cause safety accidents such as support collapse and roof falling.

[0004] Currently, the existing methods for dealing with the problem of stent "crushing" mainly rely on manual judgment and manual operation. Manual judgment is often based on experience and cannot conduct real-time monitoring and risk assessment of the stress state of the stent. It is difficult to detect the potential risk of stent "crushing" in advance, which poses a serious threat to personnel life safety. Summary of the invention

[0005] In order to solve the above technical problems, the present application provides a risk warning method, system, device and storage medium for comprehensive mining in coal mines.

[0006] The technical solution provided in this application is described below: The first aspect of the present application provides a risk early warning method for fully-mechanized mining in a coal mine, the method comprising: Acquiring historical data, wherein the historical data includes historical support force data, historical working surface top plate pressure data, and historical bottom plate condition data; Build a risk warning model based on deep learning network; extracting target features according to the historical data; Training the risk warning model according to the target feature to obtain a target risk warning model, wherein the target risk warning model is used to predict risk information according to the stress state of the bracket; Collecting working data in real time, the working data including working face top plate pressure data, support support stress state data and bottom plate condition data; Inputting the working data into the target risk warning model to obtain the stress state information of the bracket, and predicting the risk information according to the stress state information; Determine a warning level according to the risk information, wherein the warning level includes primary warning, intermediate warning and advanced warning; When the risk information is determined to be a primary warning or a mid-level warning, an audible and visual alarm is sounded, and the bracket is reinforced or the position of the bracket is adjusted according to the stress state of the bracket; When it is determined that the risk information is a high-level warning, an audible and visual alarm and voice broadcast are used, and a maintenance plan is determined according to the stress state information of the bracket, so that the bracket is processed according to the maintenance plan.

[0007] Optionally, extracting target features according to the historical data includes: Separating a support support force data sequence and a top plate pressure data sequence from the historical data; Performing outlier removal processing and noise interference removal processing on the bracket support force data sequence and the top plate pressure data sequence to obtain processed data; Normalizing the processed data to obtain normalized data; Obtaining the support force variation trend characteristics and the top plate pressure fluctuation range characteristics according to the normalized data; The support force variation trend characteristics of the bracket and the top plate pressure fluctuation range characteristics are integrated to obtain the target characteristics.

[0008] Optionally, inputting the working data into the target risk warning model to obtain stress state information of the support, and predicting risk information according to the stress state information, including: Preprocessing the working data to obtain preprocessed data; Inputting the preprocessed data into the target risk warning model to obtain the stress state characteristics of the stent; Determining the stress level according to the stress state characteristics; Determining whether the force level meets the preset requirement level; If yes, then generating first stress state information of the bracket according to the stress level; if no, then generating second stress state information of the bracket according to the stress level; Predict risk information according to the first stress state information or the second stress state information.

[0009] Optionally, a warning level is determined according to the risk information, and the warning level includes primary warning, intermediate warning and advanced warning, including: Determining the risk type in the risk information, wherein the risk type includes support instability, roof collapse or floor uplift; Calculating a risk index corresponding to each risk type in the risk information; Calculate the risk index corresponding to each risk type to obtain a total risk index; Set primary, intermediate and advanced thresholds; When the total risk index is less than the initial threshold, a primary warning is determined according to the risk information; When the total risk index is greater than the initial threshold but less than the intermediate threshold, it is determined as an intermediate warning according to the risk information; When the total risk index is greater than or equal to the advanced threshold, it is determined as an advanced warning according to the risk information.

[0010] Optionally, calculating a risk index corresponding to each risk type in the risk information includes: defining an initial risk index for each of said risk types; Calculate the risk index corresponding to each risk type in the risk information according to the target formula; The target formula is: Ri = Ki × Pi × Ii, where Ri is the risk index corresponding to the risk type, Ki is the initial risk index of the risk type, Pi and Ii are index coefficients, and the index coefficients are in the range of [0,1].

[0011] Optionally, the risk index corresponding to each risk type is calculated to obtain a total risk index, including: The risk index corresponding to each risk type is added together to obtain a total risk index.

[0012] Optionally, work data is collected in real time, the work data including work surface top plate pressure data, support support stress state data and bottom plate condition data, including: The working surface top plate pressure data, the bracket support stress state data and the bottom plate condition data are acquired through sensors.

[0013] The second aspect of the present application provides a risk early warning system for fully-mechanized mining in a coal mine, the system comprising: An acquisition unit, used to acquire historical data, wherein the historical data includes historical support force data, historical working surface top plate pressure data and historical bottom plate condition data; A construction unit for constructing a risk warning model based on a deep learning network; An extraction unit, used for extracting target features according to the historical data; A training unit, used for training the risk warning model according to the target feature to obtain a target risk warning model, wherein the target risk warning model is used for predicting risk information according to the stress state of the support; A collection unit, used to collect working data in real time, wherein the working data includes working surface top plate pressure data, support support stress state data and bottom plate condition data; An input unit, used for inputting the working data into the target risk warning model to obtain the stress state information of the support, and predicting the risk information according to the stress state information; A determination unit, configured to determine a warning level according to the risk information, wherein the warning level includes a primary warning, an intermediate warning, and a high warning; A first processing unit is used to, when determining that the risk information is a primary warning or an intermediate warning, give an audible and visual alarm, and reinforce the bracket or adjust the position of the bracket according to the stress state of the bracket; The second processing unit is used to, when it is determined that the risk information is a high-level warning, determine a maintenance plan through sound and light alarm and voice playback, and according to the stress state information of the bracket, so that the bracket is processed according to the maintenance plan.

[0014] A third aspect of the present application provides a risk warning device for fully-mechanized mining in a coal mine, the device comprising: Processor, memory, input-output unit, and bus; The processor is connected to the memory, the input and output unit, and the bus; The memory stores a program, and the processor calls the program to execute the method as described in the first aspect and any one of the first aspects.

[0015] A fourth aspect of the present application provides a computer-readable storage medium, on which a program is stored. When the program is executed on a computer, the method described in the first aspect and any one of the first aspects is performed.

[0016] It can be seen from the above technical solutions that this application has the following advantages: 1. Building a risk warning model based on a deep learning network can automatically learn and extract complex features in the data. Compared with traditional methods, deep learning models have significant advantages in processing nonlinear and high-dimensional data, thereby improving the accuracy of risk warnings.

[0017] 2. Real-time collection of work data ensures the timeliness of risk warning. By inputting real-time work data into the target risk warning model, the stress state information of the support can be obtained, and risk information can be predicted based on the information, thus realizing real-time monitoring and early warning of the coal mining process.

[0018] 3. Determine the warning level based on risk information and provide clear classification standards for different levels of risks. For primary or intermediate warnings, use sound and light alarms, and reinforce the bracket or adjust the position of the bracket according to the stress state of the bracket, and take countermeasures to ensure safety and avoid unnecessary production interruptions.

[0019] For advanced warnings, stronger warnings are provided through sound and light alarms and voice playback, and the maintenance plan is determined based on the stress status information of the bracket, so that the bracket can be processed through the maintenance plan, ensuring rapid response and effective handling in high-risk situations.

[0020] 4. Through real-time monitoring and early warning, potential risks in the coal mining process can be discovered and dealt with in a timely manner, effectively avoiding safety accidents such as roof collapse caused by uneven support force or overload.

[0021] 5. Through intelligent monitoring, early warning and processing mechanisms, real-time monitoring and precise control of the stent status can be achieved, effectively avoiding or solving the problem of stent "crushing", thereby effectively protecting the life, health and safety of the staff. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solution in the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0023] In order to more clearly illustrate the technical solution in the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0024] Figure 1 This is a schematic diagram of an embodiment of a risk early warning method for fully mechanized mining of a coal mine in this application; Figure 2 This is another embodiment schematic diagram of the risk early warning method for fully mechanized mining of coal mines of the present application; Figure 3 This is another embodiment schematic diagram of the risk early warning method for fully mechanized mining of coal mines of the present application; Figure 4 This is another embodiment schematic diagram of the risk early warning method for fully mechanized mining of coal mines of the present application; Figure 5 This is a schematic diagram of an embodiment of a risk early warning system for fully mechanized mining of a coal mine in this application; Figure 6This is a schematic diagram of an embodiment of a risk warning device for fully mechanized mining in a coal mine according to the present application. DETAILED DESCRIPTION

[0025] It should be noted that the risk warning method for comprehensive mining of coal mines provided in this application can be applied to a terminal, a system, or a server. For example, the terminal can be a smart phone or a computer, a tablet computer, a smart TV, a smart watch, a portable computer terminal, or a fixed terminal such as a desktop computer. For the convenience of explanation, this application uses the terminal as an example for explanation.

[0026] See also Figure 1 The present application first provides an embodiment of a risk early warning method for fully mechanized mining in a coal mine, the embodiment comprising: 101. Acquire historical data, wherein the historical data includes historical support force data, historical working surface top plate pressure data, and historical bottom plate condition data; 102. Construct risk warning model based on deep learning network; 103. Extracting target features according to the historical data; 104. Training the risk warning model according to the target feature to obtain a target risk warning model, wherein the target risk warning model is used to predict risk information according to the stress state of the bracket; 105. Collect working data in real time, wherein the working data includes working surface top plate pressure data, support support stress state data and bottom plate condition data; 106. Inputting the working data into the target risk warning model to obtain stress state information of the bracket, and predicting risk information according to the stress state information; 107. Determine a warning level according to the risk information, wherein the warning level includes primary warning, intermediate warning and advanced warning; 108. When it is determined that the risk information is a primary warning or a mid-level warning, an audible and visual alarm is sounded, and the bracket is reinforced or the position of the bracket is adjusted according to the stress state of the bracket; 109. When it is determined that the risk information is a high-level warning, a maintenance plan is determined through sound and light alarms and voice playback, and according to the stress state information of the bracket, so that the bracket is processed according to the maintenance plan.

[0027] In an embodiment of the present application, historical data is first obtained, and the historical data includes historical support force data, historical working surface top plate pressure data, and historical bottom plate condition data. Then, a risk warning model is constructed based on a deep learning network, and after extracting target features based on historical data; the risk warning model is trained according to the target features to obtain a target risk warning model, and the target risk warning model is used to predict risk information based on the force state of the support; after the target risk warning model is trained, working data is collected in real time, and the working data includes working surface top plate pressure data, support support force state data, and bottom plate condition data; and the working data is input into the target risk warning model to obtain the force state information of the support, and the risk information is predicted based on the force state information, and the warning level is determined based on the risk information, and the warning level includes primary warning, intermediate warning, and advanced warning; when the risk information is determined to be a primary warning or an intermediate warning, an audible and visual alarm is used, and the support is reinforced or the position of the support is adjusted according to the force state of the support; when the risk information is determined to be an advanced warning, an audible and visual alarm and voice playback are used, and a maintenance plan is determined based on the force state information of the support, so that the support is processed through the maintenance plan.

[0028] In step 101, historical data must be obtained first. This part of historical data is collected in real time by sensors or strain gauges installed on the bracket. The historical data should include the stress conditions of the bracket in different time periods and different working conditions, such as pressure, tension, shear force, etc., as well as the changing trends and peak values ​​of these forces.

[0029] It should be noted that in order to fully train the model in the future, outliers in the historical data must be eliminated.

[0030] Among them, the working face roof pressure data is usually collected through pressure sensors installed on the roof. The working face roof pressure data includes the pressure distribution of the roof under different geological conditions, mining stages and support methods; the bottom plate condition data can be obtained through geological exploration, drilling or field testing. The data includes the physical and mechanical properties of the bottom plate, such as rock type, thickness, bedding structure, hardness, etc. For the bottom plates under different geological conditions, data can be collected separately, as well as the force data of the support, which is also obtained through sensors.

[0031] In step 102, a risk warning model is constructed based on a deep learning network. Specifically, a convolutional neural network (CNN) is used to construct a risk warning model for the support stress data. After the risk warning model is constructed, it is necessary to use the acquired data to train the risk warning model. During the training process, the model is trained using a loss function and an optimization algorithm. The loss function is used to measure the difference between the predicted result of the risk warning model and the actual result. The optimization algorithm is used to update the parameters of the model to minimize the loss function.

[0032] In step 103, target features are extracted based on historical data, wherein statistical features such as the maximum value, minimum value, average value, standard deviation, and rate of change of the force on the bracket can be extracted from the bracket force data of the historical data based on the historical data. It should be noted that in the process of feature extraction, data preprocessing, such as denoising, normalization, and standardization, is required to improve the validity and comparability of the features.

[0033] In step 104, the risk warning model is trained according to the target features to obtain a target risk warning model. Specifically, after the target features are extracted in step 103, the extracted target features are used as input, and the risk information in the historical data is used as output to construct a training data set. The risk warning model is trained using the training data set. During the training process, the loss function and accuracy of the model are monitored to evaluate the learning effect of the risk warning model. During the training process, the performance of the risk warning model is continuously improved by adjusting the parameters of the risk warning model and the optimization algorithm. When the performance of the risk warning model reaches the qualified level, it can be considered that the risk warning model training is completed and the target risk warning model is obtained. The target risk warning model is used to predict risk information according to the stress state of the bracket.

[0034] In step 105, after the target risk warning model is trained, the target risk warning model is used to provide real-time risk warning for the support. First, the working data must be collected in real time. The real-time working data includes the working face top plate pressure data, the support stress state data and the bottom plate condition data. These real-time collection of working data is achieved through various sensors. It should be noted that after the working data is obtained in real time, the abnormal or missing data in the working data will be eliminated to ensure the integrity and availability of the working data.

[0035] In step 106, the working data is input into the target risk warning model to obtain the stress state information of the support, and the risk information is predicted based on the stress state information. Specifically, the working data collected in real time is input into the target risk warning model. The target risk warning model will calculate the stress state information of the support based on these data, such as the distribution of force, change trend, etc., and the target risk warning model will predict possible risk information based on the stress state information, such as the support is about to be damaged, the working surface is about to collapse, etc. The prediction result can be expressed in the form of probability or classification, which is not specifically limited here.

[0036] In step 107, after inputting the work data into the target risk warning model to obtain the corresponding risk information, the warning level is determined according to the risk information, wherein the determination of the warning level needs to be based on the severity and urgency of the risk information, and different warning levels can be set, such as primary warning, intermediate warning and advanced warning. Primary warning means that there is a potential risk, but it has not yet reached a critical level; intermediate warning means that the risk is already obvious and measures need to be taken to prevent it; advanced warning means that the risk is already very serious and measures need to be taken immediately for emergency treatment.

[0037] In step 108, when the risk information is determined to be a primary warning or an intermediate warning, the sound and light alarm device will be triggered to sound an alarm or flash a light to alert the staff. At the same time, the system will give suggestions for strengthening the bracket or adjusting the bracket position based on the stress state information of the bracket. For example, the support force of the bracket can be increased or the angle of the bracket can be changed to reduce the risk. The staff can take corresponding measures according to the suggestions of the system and feedback the processing results in time so that the system can evaluate and adjust the warning effect.

[0038] In step 109, when the risk information is determined to be a high-level warning, the system will simultaneously trigger the sound and light alarm and voice playback device. The voice playback can issue a clear alarm message, such as "The bracket is about to be damaged, please take measures immediately!", to remind the staff in a stronger way. At the same time, the system will give a specific maintenance plan based on the stress state information of the bracket. For example, it can be recommended to replace the damaged bracket parts, strengthen the support structure of the bracket, or take other emergency measures to ensure the safety and stability of the working surface.

[0039] Staff can process the bracket according to the maintenance plan and provide timely feedback on the results. The system will evaluate the treatment effect to ensure that risks are effectively controlled.

[0040] Therefore, this application realizes real-time monitoring and precise control of the stent status through intelligent monitoring, early warning and processing mechanisms, effectively avoiding or solving the problem of stent "crushing", thereby effectively protecting the life, health and safety of workers.

[0041] Please refer to Figure 2 According to some embodiments of the present invention, extracting target features according to the historical data in step 103 may specifically include, but is not limited to, the following: 201. Separating a support support force data sequence and a top plate pressure data sequence from the historical data; 202. Perform outlier removal processing and noise interference removal processing on the bracket support force data sequence and the top plate pressure data sequence to obtain processed data; 203. Normalizing the processed data to obtain normalized data; 204. Obtaining support force variation trend characteristics and top plate pressure fluctuation range characteristics according to the normalized data; 205. Integrate the support force variation trend characteristics of the bracket and the top plate pressure fluctuation range characteristics to obtain the target characteristics.

[0042] In an embodiment of the present application, data related to the support force of the bracket are separated from the historical data through data screening and extraction. These data usually include the force values ​​of the bracket at different time points or working conditions and exist in the form of time series. The integrity and continuity of the data are ensured during extraction to facilitate subsequent analysis of the changing trend of the bracket force.

[0043] Similarly, data related to roof pressure are separated from historical data. The roof pressure data include roof pressure values ​​under different geological conditions and mining stages, and also exist in the form of time series. Similarly, attention is paid to the accuracy and consistency of the data during extraction in order to accurately reflect the changes in roof pressure.

[0044] After extracting the support force data series and the top plate pressure data series, outlier detection is required. Outliers may be caused by sensor failure, data recording errors, or extreme working conditions. These outliers are identified and removed through statistical methods or machine learning algorithms to ensure the accuracy and reliability of the data. In the process of data collection, it may be interfered by various noises, such as environmental noise, sensor noise, etc. These noises will affect the authenticity and accuracy of the data, so denoising is required. Therefore, filtering algorithms, signal processing methods or machine learning techniques are required to remove noise and retain the effective information of the data.

[0045] The processed data has different dimensions and ranges. In order to facilitate subsequent feature extraction and model training, normalization is required. Normalization is to scale the data to a specific range (such as between 0 and 1) so that the values ​​of different features are comparable. Z-score normalization and other methods can be used for data processing.

[0046] The normalized bracket support force data series is analyzed to extract its change trend characteristics. The change trend characteristics can reflect the change of bracket force over time or working conditions, such as gradual increase, gradual decrease, fluctuating changes, etc. The change trend of bracket force can also be described by calculating indicators such as slope, curvature, and rate of change.

[0047] Similarly, the normalized roof pressure data series is analyzed to extract its fluctuation range characteristics. The fluctuation range characteristics can reflect the variation amplitude and stability of the roof pressure within a certain period of time. The fluctuation range of the roof pressure can be described by calculating indicators such as standard deviation, range, and fluctuation coefficient.

[0048] Finally, the extracted support force change trend characteristics and roof pressure fluctuation range characteristics are integrated to form target features. The target features are the key input for subsequent model training and risk warning. The final target features will be used to train the risk warning model to predict the stress state of the support and the changes in roof pressure, providing strong support for safety management in the coal mining process.

[0049] Please refer to Figure 3 According to some embodiments of the present invention, in step 106, the working data is input into the target risk warning model to obtain the stress state information of the bracket, and the risk information predicted based on the stress state information may specifically include, but is not limited to, the following: 301. Preprocessing the working data to obtain preprocessed data; 302. Inputting the preprocessed data into the target risk warning model to obtain stress state characteristics of the stent; 303. Determine the stress level according to the stress state characteristics; 304. Determine whether the force level meets the preset requirement level; 305. If yes, generate first stress state information of the bracket according to the stress level; if no, generate second stress state information of the bracket according to the stress level; 306. Predict risk information according to the first stress state information or the second stress state information.

[0050] In an embodiment of the present application, first, the working data is cleaned to remove or correct erroneous, abnormal or missing data points to ensure the accuracy and completeness of the data. The preprocessed data is then input into a trained target risk warning model. The target risk warning model extracts the stress state characteristics of the stent based on the input data. These stress state characteristics include pattern information such as the distribution, trend, and periodicity of the stress on the stent. The extracted stress state characteristics are used to determine the stress level and predict risk information.

[0051] Specifically, the force level of the bracket is determined based on the extracted stress state characteristics and the preset stress level classification standard. The force level is a discrete value or category, such as "low stress", "medium stress", "high stress", etc., which is used to reflect the current stress state of the bracket. The determined force level is then compared with the preset required level to determine whether the stress state of the bracket is within a safe range.

[0052] If the force level meets the preset requirement level, the first force state information of the bracket is generated according to the force level. The first force state information may include positive information such as the description of the force state of the bracket, the force level, and the safety state, which is used to indicate that the bracket is currently in a safe state. If the force level does not meet the preset requirement level, the second force state information of the bracket is generated according to the force level. The second force state information may include negative information such as the description of the abnormal force of the bracket, the force level, and the risk level, which is used to warn the staff that the bracket currently has a force risk.

[0053] Finally, the risk information that the support may face is predicted based on the generated first stress state information or the second stress state information. The risk information includes the probability of support damage, the risk level of working face collapse, the emergency measures that need to be taken, etc. The predicted risk information will be transmitted to the staff in a timely manner so that they can take corresponding measures according to the severity and urgency of the risk information to ensure the safety and stability of the working face.

[0054] Please refer to Figure 4 According to some embodiments of the present invention, in step 107, the warning level is determined according to the risk information, and the warning level includes primary warning, intermediate warning and advanced warning, and specifically may include, but is not limited to, the following: 401. Determine the risk type in the risk information, where the risk type includes support instability, roof collapse, or floor uplift; 402. Calculate a risk index corresponding to each risk type in the risk information; 403. Calculate the risk index corresponding to each risk type to obtain a total risk index; 404. Setting primary threshold, intermediate threshold and advanced threshold; 405. When the total risk index is less than the initial threshold, a primary warning is determined according to the risk information; 406. When the total risk index is greater than the initial threshold but less than the intermediate threshold, a medium-level warning is determined according to the risk information; 407. When the total risk index is greater than or equal to the advanced threshold, a high-level warning is determined according to the risk information.

[0055] In the embodiment of the present application, the risk type is determined by comprehensive judgment based on multiple factors such as the stress state of the support, the pressure on the top plate of the working face and the conditions of the bottom plate. The specific risk types include support instability, top plate collapse and bottom plate uplift. Support instability refers to the instability of the support due to uneven force, insufficient support or damage to the support during the support process; top plate collapse refers to the collapse of the top plate of the working face due to excessive pressure, insufficient support or changes in geological conditions, and bottom plate uplift refers to the uplift of the bottom plate due to changes in geological conditions, mining activities or groundwater action.

[0056] Then, the risk index corresponding to each risk type in the risk information is calculated. The risk index is an indicator used to quantify the size of the risk. For each risk type, its corresponding risk index needs to be calculated. Specifically, the calculation of the risk index can refer to the following method: define the initial risk index of each type in the risk type, and then calculate the risk index corresponding to each risk type in the risk information according to the target formula; the target formula is: Ri = Ki × Pi × Ii, where Ri is the risk index corresponding to the risk type, Ki is the initial risk index of the risk type, Pi and Ii are exponential coefficients respectively, and the exponential coefficient takes the value of [0,1].

[0057] The risk index corresponding to each risk type is then added together to obtain the total risk index. The total risk index is an indicator used to comprehensively reflect the size of the entire work risk. By calculating the total risk index, the risk status of the work face can be more comprehensively assessed, providing a basis for the subsequent determination of the warning level.

[0058] The threshold is used to divide the warning level. According to the size of the total risk index, the primary threshold, intermediate threshold and advanced threshold can be set. The primary threshold indicates that the risk is low and needs to be paid attention to but no immediate measures are required; the intermediate threshold indicates that the risk is high and measures need to be taken for prevention; the advanced threshold indicates that the risk is very high and emergency measures need to be taken immediately.

[0059] When the total risk index is less than the primary threshold, it means that the risk of the working face is low, but there is still a certain potential risk. At this time, the system will issue a primary warning based on the risk information to remind the staff to pay attention to the risk situation, but no immediate measures are required. The staff needs to pay attention to the changes in the working face and promptly discover and deal with potential risk problems.

[0060] When the total risk index is greater than the primary threshold but less than the intermediate threshold, it indicates that the risk of the working face is high and preventive measures need to be taken. At this time, the system will issue an intermediate warning based on the risk information, reminding the staff to take necessary measures to reduce the risk. The staff can promptly reinforce the support, adjust the support structure or take other preventive measures according to the system's suggestions to ensure the stability and safety of the working face.

[0061] When the total risk index is greater than or equal to the advanced threshold, it means that the risk of the working face is very high and immediate emergency measures need to be taken. At this time, the system will issue an advanced warning based on the risk information, reminding the staff to take emergency measures immediately to deal with the risk. The staff needs to respond quickly to the system's warning information and handle it according to the emergency plan to ensure the safety of personnel and equipment, and restore the stability of the working face as soon as possible.

[0062] See also Figure 5 The second aspect of the present application provides a risk early warning system for fully-mechanized mining in coal mines, the system comprising: An acquisition unit 501 is used to acquire historical data, wherein the historical data includes historical support force data, historical working surface top plate pressure data and historical bottom plate condition data; A construction unit 502 is used to construct a risk warning model based on a deep learning network; An extraction unit 503, configured to extract target features according to the historical data; A training unit 504 is used to train the risk warning model according to the target feature to obtain a target risk warning model, wherein the target risk warning model is used to predict risk information according to the stress state of the support; A collecting unit 505 is used to collect working data in real time, wherein the working data includes working surface top plate pressure data, support support stress state data and bottom plate condition data; An input unit 506, used for inputting the working data into the target risk warning model to obtain the stress state information of the support, and predicting the risk information according to the stress state information; A determination unit 507 is used to determine a warning level according to the risk information, where the warning level includes a primary warning, an intermediate warning, and a high warning; The first processing unit 508 is configured to, when determining that the risk information is a primary warning or a mid-level warning, give an audible and visual alarm, and reinforce the bracket or adjust the position of the bracket according to the stress state of the bracket; The second processing unit 509 is used to, when it is determined that the risk information is a high-level warning, determine a maintenance plan through sound and light alarm and voice playback, and according to the stress state information of the bracket, so that the bracket is processed according to the maintenance plan.

[0063] See also Figure 6 The present application also provides a risk warning device for fully mechanized mining in coal mines, comprising: Processor 601, memory 602, input and output unit 603, bus 604; The processor 601 is connected to the memory 602, the input and output unit 603 and the bus 604; The memory 602 stores a program, and the processor 601 calls the program to execute any of the above methods.

[0064] The present application also relates to a computer-readable storage medium on which a program is stored, wherein when the program is run on a computer, the computer is caused to execute any of the above methods.

[0065] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0066] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0067] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0068] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0069] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, read-only memory), random access memory (RAM, random access memory), disk or optical disk and other media that can store program code.

Claims

1. A risk early warning method for fully mechanized mining in coal mines, characterized in that: The method comprises: Acquire historical data, the historical data including historical support force data, historical working face roof pressure data and historical floor condition data, the historical working face roof pressure data is collected by a pressure sensor installed on the roof, the historical support force data and the historical floor condition data are both acquired by the sensor; the historical working face roof pressure data includes pressure distribution data of the roof under different geological conditions, mining stages and support methods; the historical floor condition data includes lithology, thickness, bedding structure and hardness physical and mechanical property data of the floor; Build a risk warning model based on deep learning network; extracting target features according to the historical data; Training the risk warning model according to the target feature to obtain a target risk warning model, wherein the target risk warning model is used to predict risk information according to the stress state of the stent; Collecting working data in real time, including working surface top plate pressure data, support support stress state data and bottom plate condition data; Inputting the working data into the target risk warning model to obtain the stress state information of the bracket, and predicting the risk information according to the stress state information; Determine a warning level according to the risk information, wherein the warning level includes primary warning, intermediate warning and advanced warning; When the risk information is determined to be a primary warning or a mid-level warning, an audible and visual alarm is sounded, and the bracket is reinforced or the position of the bracket is adjusted according to the stress state of the bracket; When it is determined that the risk information is a high-level warning, an audible and visual alarm and voice broadcast are used, and a maintenance plan is determined according to the stress state information of the bracket, so that the bracket is processed according to the maintenance plan.

2. The risk early warning method for fully mechanized mining in coal mines according to claim 1 is characterized in that: Extracting target features according to the historical data includes: Separating a support support force data sequence and a top plate pressure data sequence from the historical data; Performing outlier removal processing and noise interference removal processing on the bracket support force data sequence and the top plate pressure data sequence to obtain processed data; Normalizing the processed data to obtain normalized data; Obtaining the support force variation trend characteristics and the top plate pressure fluctuation range characteristics according to the normalized data; The support force variation trend characteristics of the bracket and the top plate pressure fluctuation range characteristics are integrated to obtain the target characteristics.

3. The risk early warning method for fully mechanized mining in coal mines according to claim 1 is characterized in that: Inputting the working data into the target risk warning model to obtain the stress state information of the bracket, and predicting the risk information according to the stress state information, including: Preprocessing the working data to obtain preprocessed data; Inputting the preprocessed data into the target risk warning model to obtain the stress state characteristics of the stent; Determining the stress level according to the stress state characteristics; Determining whether the force level meets the preset requirement level; If yes, then generating first stress state information of the bracket according to the stress level; if no, then generating second stress state information of the bracket according to the stress level; Predict risk information according to the first stress state information or the second stress state information.

4. The risk early warning method for fully mechanized mining in coal mines according to claim 3 is characterized in that: The warning level is determined according to the risk information, and the warning level includes primary warning, intermediate warning and advanced warning, including: Determining the risk type in the risk information, wherein the risk type includes support instability, roof collapse or floor uplift; Calculating a risk index corresponding to each risk type in the risk information; Calculate the risk index corresponding to each risk type to obtain a total risk index; Set primary, intermediate and advanced thresholds; When the total risk index is less than the initial threshold, a primary warning is determined based on the risk information; When the total risk index is greater than the initial threshold but less than the intermediate threshold, it is determined as an intermediate warning according to the risk information; When the total risk index is greater than or equal to the advanced threshold, it is determined as an advanced warning according to the risk information.

5. The risk early warning method for fully mechanized mining in coal mines according to claim 4 is characterized in that: Calculating the risk index corresponding to each risk type in the risk information includes: defining an initial risk index for each of said risk types; Calculate the risk index corresponding to each risk type in the risk information according to the target formula; The target formula is: Ri = Ki × Pi × Ii, where Ri is the risk index corresponding to the risk type, Ki is the initial risk index of the risk type, Pi and Ii are index coefficients, and the index coefficients are in the range of [0,1].

6. The risk early warning method for fully mechanized mining in coal mines according to claim 4 is characterized in that: The risk index corresponding to each risk type is calculated to obtain a total risk index, including: The risk index corresponding to each risk type is added together to obtain a total risk index.

7. The risk early warning method for fully mechanized mining in coal mines according to claim 1 is characterized in that: Collect working data in real time, including working surface top plate pressure data, support support stress state data and bottom plate condition data, including: The working surface top plate pressure data, the bracket support stress state data and the bottom plate condition data are acquired through sensors.

8. A risk early warning system for fully mechanized mining in coal mines, characterized in that: The system comprises: An acquisition unit is used to acquire historical data, wherein the historical data includes historical support force data, historical working face roof pressure data and historical floor condition data, wherein the historical working face roof pressure data is collected by a pressure sensor installed on the roof, and the historical support force data and the historical floor condition data are both acquired by the sensor; the historical working face roof pressure data includes pressure distribution data of the roof under different geological conditions, mining stages and support methods; the historical floor condition data includes lithology, thickness, bedding structure and hardness physical and mechanical property data of the floor; A construction unit for constructing a risk warning model based on a deep learning network; An extraction unit, used for extracting target features according to the historical data; A training unit, used for training the risk warning model according to the target feature to obtain a target risk warning model, wherein the target risk warning model is used for predicting risk information according to the stress state of the support; A collection unit, used to collect working data in real time, wherein the working data includes working surface top plate pressure data, support support stress state data and bottom plate condition data; An input unit, used for inputting the working data into the target risk warning model to obtain the stress state information of the support, and predicting the risk information according to the stress state information; A determination unit, configured to determine a warning level according to the risk information, wherein the warning level includes a primary warning, an intermediate warning, and a high warning; A first processing unit is used to, when determining that the risk information is a primary warning or an intermediate warning, give an audible and visual alarm, and reinforce the bracket or adjust the position of the bracket according to the stress state of the bracket; The second processing unit is used to, when it is determined that the risk information is a high-level warning, determine a maintenance plan through sound and light alarm and voice playback, and according to the stress state information of the bracket, so that the bracket is processed according to the maintenance plan.

9. A risk warning device for fully mechanized mining in coal mines, characterized in that: The device comprises: Processor, memory, input-output unit, and bus; The processor is connected to the memory, the input and output unit, and the bus; The memory stores a program, and the processor calls the program to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a program, and when the program is executed on a computer, the method according to any one of claims 1 to 7 is performed.

Images