Coal and gas outburst hazard warning method and system for machine-dug working face

Abstract

The present invention discloses a method and system for early warning of coal and gas outburst hazards in a machine-dug working face, which belongs to the field of coal mine safety technology. The method and system specifically include: real-time collection of working face data, construction of a comprehensive early warning model to achieve trend prediction and anomaly detection, triggering early warning and evaluating risk levels; combining with mine GIS to accurately locate the early warning area, using an intelligent decision support system of knowledge graph and rule engine to quickly match the optimal emergency plan and generate an emergency response plan; in the present invention, tunneling machine workers can take shelter or escape measures in time when an accident occurs, thereby improving mine operation safety and emergency response efficiency, and improving the operation effect of the tunneling machine.

Description

Technical Field

[0001] The present invention belongs to the technical field of coal mine safety, and in particular relates to a method and system for early warning of coal and gas outburst hazards in a machine-dig working face. Background Art

[0002] Coal and gas outburst is a serious natural disaster faced during coal mining. It refers to the phenomenon that a large amount of gas and coal is suddenly ejected from the coal body into the mining space in a very short period of time. This disaster not only seriously threatens the safe production of coal mines, but also causes significant casualties and property losses. Therefore, researching and developing effective coal and gas outburst hazard warning methods is of great significance to improving the safety and efficiency of coal mine production.

[0003] Coal and gas outbursts are characterized by suddenness, destructiveness, and difficulty in prediction. During coal mining, once a coal and gas outburst occurs, large amounts of gas and coal can instantly flood into the excavation space, causing serious consequences such as tunnel blockage, ventilation system damage, and a sharp increase in gas concentration. This not only severely impacts normal coal mine production but can also trigger secondary disasters such as gas explosions, posing a significant threat to the lives of underground workers. Currently, scholars and engineers at home and abroad have conducted extensive research on the issue of coal and gas outbursts and proposed a variety of prediction and early warning methods. However, these methods still have some shortcomings in practical applications: most existing prediction methods are based on empirical formulas or statistical models, which cannot accurately reflect the complex mechanisms of coal and gas outbursts, resulting in limited prediction accuracy; some early warning systems rely solely on a single monitoring method, such as gas concentration monitoring, and ignore the monitoring of other key parameters, making it difficult to fully reflect the precursor information of coal and gas outbursts; and some early warning systems suffer from data update lags, failing to reflect the dynamic changes of coal and gas outbursts in real time, resulting in poor early warning effectiveness.

[0004] For example, Chinese patent application CN110118103B discloses a coal mine gas early warning method, which includes first obtaining from a gas monitoring system a list of all gas sensor information tables (configList) configured with gas early warning parameters, a list of real-time gas data information tables (realList), and a list of real-time gas early warning data information tables (alarmList). The system then compares and analyzes the real-time gas data information table (realList), the list of gas sensor information tables (configList) configured with gas early warning parameters, and the list of real-time gas early warning data information tables (alarmList) to obtain a list of real-time gas early warning data information tables (alarmList) and a list of historical gas alarm information tables. This technical solution addresses the drawback of existing coal mine gas monitoring systems, which often trigger accidents by only issuing alarms after gas levels exceed a certain limit. This technical solution can predict potential alarms.

[0005] The above existing technologies all have the following problems: lack of real-time and trend prediction capabilities, and lack of spatial positioning capabilities, making it impossible to detect potential dangers in advance and take emergency measures. Summary of the Invention

[0006] In response to the shortcomings of the existing technology, the present invention proposes a method and system for early warning of coal and gas outburst hazards in machine-dug working faces, including: real-time collection of working face data, construction of a comprehensive early warning model to achieve trend prediction and anomaly detection, triggering early warning and evaluating risk levels; combining with mine GIS to accurately locate the early warning area, using the intelligent decision support system of knowledge graph and rule engine to quickly match the optimal emergency plan and generate an emergency response plan; in the present invention, tunneling machine workers can take shelter or escape measures in time when an accident occurs, thereby improving mine operation safety and emergency response efficiency, and improving the operation effect of the tunneling machine.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] Early warning methods for coal and gas outburst hazards in machine-dug working faces include:

[0009] Step S1: Using an early warning component installed on the end face of the roadheader body, real-time working face monitoring information is collected through a wireless data transmitting and receiving component. The early warning component on the end face of the roadheader body includes a mounting shell, a mounting plate provided on the inner wall of the mounting shell, a sensor component provided on the end face of the mounting plate, and a real-time data collection component. The wireless data transmitting and receiving component is used to upload the working status and analysis results of the roadheader and receive instructions and prompt information from the ground central station.

[0010] Step S2: Based on the working face monitoring information, a comprehensive early warning model is established to perform real-time trend prediction and anomaly detection on the working face monitoring information. If any data anomalies or potential hazards are found, the electronic control early warning platform immediately triggers the early warning mechanism and automatically assesses the risk level through the built-in risk assessment algorithm;

[0011] Step S3: Based on the warning information, the warning information is located in conjunction with the mine geographic information system, and an intelligent emergency decision support system based on the knowledge graph and rule engine is constructed. The optimal emergency plan is automatically matched according to the urgency and risk level of the warning information. At the same time, based on the emergency plan, the emergency response plan is executed through the eight-way control loop in the power outage control component;

[0012] Step S4: After receiving the warning information, the tunnel boring machine staff enters the refuge or escape mechanism according to the warning information and the instructions of the emergency plan.

[0013] Specifically, the specific steps of step S2 include:

[0014] S2.1: Receive working face monitoring information and perform preprocessing, extract features from the preprocessed working face monitoring information, and generate working face feature data ,in, Indicates the N Working surface characteristic data, N Indicates the number of working face characteristic data; the working face characteristic data includes working face gas concentration data, coal seam stress state data, and working face wind speed data;

[0015] S2.2: Build a comprehensive early warning model based on machine learning methods and anomaly detection algorithms, train the comprehensive early warning model using historical working face monitoring information, and use confusion matrix to evaluate the model;

[0016] The confusion matrix is ​​used to show the comparison between the model prediction results and the actual categories, and the elements in the confusion matrix are filled based on the prediction results of the comprehensive early warning model.

[0017] Specifically, the specific steps of step S2 also include:

[0018] S2.3: Work surface feature data Input the trained comprehensive early warning model and use Perform trend prediction and anomaly detection to obtain anomaly detection results ,in, represents the anomaly detection result at the current time point t, g represents the constant term, r represents the order of the autoregressive part, represents the coefficient of the autoregressive part, Indicates a time point The anomaly detection result when u represents the order of the moving average part, represents the coefficient of the moving average part, Represents the moving average part The error term of order , v represents the number of external variables, represents the coefficient of external variables, represents the external variable, z represents the number of periodic patterns, represents the periodic function coefficient, represents a periodic function, represents the coefficient of the dummy variable, represents a dummy variable, represents the time trend term coefficient, represents the time trend term, represents the error term, Indicates the anomaly detection result of the Nth working surface feature data.

[0019] Specifically, the specific steps of step S2 also include:

[0020] S2.4: Setting anomaly detection thresholds ,in, , represents the mean of Y, b represents the slope coefficient, represents the standard deviation of Y;

[0021] like , it means that the working surface feature data is normal data;

[0022] like , it means there is abnormal data and the early warning mechanism is triggered through the electronic control early warning platform;

[0023] S2.5: Set the warning signals as blue, yellow, orange, and red, and automatically assess the risk level based on the severity of the warning signal.

[0024] Specifically, the specific steps of step S3 include:

[0025] S3.1: Receive the warning information in step S2 and pre-process the warning information;

[0026] S3.2: Use the mine geographic information system to locate the warning information on the mine map and use the GIS map display function to display the warning location and its surrounding environment;

[0027] S3.3: Build a mine safety knowledge graph, extract knowledge from text and images using natural language processing, and store the extracted knowledge in a structured manner in the knowledge graph;

[0028] S3.4: Design a rule engine to automatically match the optimal emergency plan based on the urgency and risk level of warning information;

[0029] S3.5: Develop decision rules. When receiving warning information, the intelligent emergency decision support system automatically triggers the rule engine to perform decision analysis. The rule engine matches the optimal emergency plan based on the attributes of the warning information and the knowledge in the knowledge graph, and generates an emergency response plan based on the emergency plan.

[0030] If the emergency response plan is triggered, the corresponding power-off and shutdown safety measures will be automatically executed through the control loop;

[0031] S3.6: Send the generated emergency response plan to staff. Simultaneously, monitor the emergency response process, collect feedback information in real time, and dynamically adjust the emergency response plan based on the feedback information.

[0032] S3.7: Continuously optimize and iterate the intelligent emergency decision support system based on the actual effects and feedback information of the emergency response.

[0033] Specifically, the abnormality detection result in S2.3 includes the time, location, type and severity of the abnormality; the emergency plan in S3 includes evacuating personnel, cutting off power, starting ventilation equipment, and allocating resources.

[0034] The coal and gas outburst hazard early warning system for machine-dug working faces includes: data acquisition module, comprehensive early warning module, emergency response module, and risk avoidance module;

[0035] The data acquisition module is used to collect working face monitoring information in real time, including gas concentration, coal seam stress state, and wind speed parameters, using the early warning component and wireless data sending and receiving component on the end face of the roadheader body;

[0036] The comprehensive early warning module is used to evaluate and predict the risk of coal and gas outburst using machine learning methods based on the working face monitoring information provided by the data acquisition module;

[0037] The emergency response module is used to activate the emergency plan after receiving the early warning signal and guide the on-site staff to take appropriate emergency measures;

[0038] The risk avoidance module is used to send information about safe shelters or escape routes after the early warning information is triggered, and guide staff to enter the shelter or escape mechanism according to the instructions of the emergency plan.

[0039] Specifically, the comprehensive early warning module includes: an early warning model construction unit and a risk assessment unit;

[0040] The early warning model construction unit is used to construct a comprehensive early warning model based on historical working face monitoring information and expert knowledge, including a trend prediction model and an anomaly detection model;

[0041] The risk assessment unit is used to use the comprehensive early warning model to perform real-time analysis on the working face monitoring information, assess the risk level of the hazard, and trigger the corresponding early warning mechanism.

[0042] Specifically, the emergency response module includes: a positioning unit, an emergency response execution unit;

[0043] The positioning unit is used to combine the warning information with the mine spatial data to locate the warning information;

[0044] The emergency response execution unit is used to automatically select and start the corresponding emergency plan and execute emergency response measures according to the early warning information and risk assessment results.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] 1. The present invention proposes a coal and gas outburst hazard early warning system for machine-dug working faces, and optimizes and improves the structure, operation steps and processes. The system has the advantages of simple process, low investment and operation costs, and low production work costs.

[0047] 2. The present invention proposes an early warning method for the dangers of coal and gas outburst in machine-dug working faces. Through the acquisition of working face monitoring information, a comprehensive early warning model and an intelligent emergency decision support system, a comprehensive and real-time early warning method for the dangers of coal and gas outburst in machine-dug working faces is achieved. It can not only detect potential dangers in advance and trigger the early warning mechanism, but also automatically evaluate the risk level and match the optimal emergency plan, thereby significantly improving the accuracy and timeliness of the early warning and accelerating the emergency response speed. At the same time, it reduces the risk of coal and gas outburst accidents, ensures personnel safety, and improves the safety and efficiency of coal mine production.

[0048] 3. The present invention proposes a method for early warning of the dangers of coal and gas outbursts in machine-dug working faces. Through real-time monitoring and early warning, the roadheader can more accurately judge the geological conditions of the working face, thereby adjusting the tunneling parameters and avoiding high-risk operations in coal and gas outburst danger areas. This not only reduces the failure rate and maintenance costs of the roadheader, but also improves tunneling efficiency and coal production. At the same time, the early warning system can also provide real-time safety guidance for the roadheader, ensuring that tunneling operations are carried out under safe conditions, further protecting the lives of roadheader workers. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 This is a schematic diagram of the method for early warning of coal and gas outburst danger in a machine-dug working face according to the present invention;

[0050] Figure 2 This is a flow chart showing the principle of the method for early warning of coal and gas outburst danger in a machine-dug working face according to the present invention;

[0051] Figure 3 This is a risk level assessment flow chart of the method for early warning of coal and gas outburst hazards in a machine-dug working face according to the present invention;

[0052] Figure 4 This is the architecture diagram of the coal and gas outburst hazard warning system for the mechanized excavation working face of the present invention. DETAILED DESCRIPTION

[0053] Example 1

[0054] See also Figure 1-Figure 3 The present invention provides an embodiment of a method for early warning of coal and gas outburst danger in a machine-dug working face, comprising the following steps:

[0055] Step S1: Using an early warning component installed on the end face of the roadheader body, real-time working face monitoring information is collected through a wireless data transmitting and receiving component. The early warning component on the end face of the roadheader body includes a mounting shell, a mounting plate provided on the inner wall of the mounting shell, a sensor component provided on the end face of the mounting plate, and a real-time data collection component. The wireless data transmitting and receiving component is used to upload the working status and analysis results of the roadheader and receive instructions and prompt information from the ground central station.

[0056] Among them, the working face monitoring information includes the environmental information of the machine excavation working face, early warning information, the location of the workers, the working status and distribution of the workers, the analysis of the ground central station, early warning prompts, and control instructions;

[0057] The process of obtaining working face monitoring information includes:

[0058] (1) Deploy microseismic sensors, gas sensors, and wind speed sensors in the working area of ​​the tunnel boring machine and connect them to the monitoring system via wired or wireless means;

[0059] (2) Install cameras that can monitor the changes in physical parameters of the coal seam at the working face in real time, as well as an accurate personnel positioning system to achieve real-time tracking of personnel location, work status and distribution.

[0060] Step S2: Based on the working face monitoring information, a comprehensive early warning model is established to perform real-time trend prediction and anomaly detection on the working face monitoring information. If any data anomalies or potential hazards are found, the electronic control early warning platform immediately triggers the early warning mechanism and automatically assesses the risk level through the built-in risk assessment algorithm;

[0061] Step S3: Based on the warning information, the warning information is located in conjunction with the mine geographic information system, and an intelligent emergency decision support system based on the knowledge graph and rule engine is constructed. The optimal emergency plan is automatically matched according to the urgency and risk level of the warning information. At the same time, based on the emergency plan, the emergency response plan is executed through the eight-way control loop in the power outage control component;

[0062] Furthermore, the emergency plan in S3 includes but is not limited to evacuating personnel, cutting off power, activating ventilation equipment, and deploying resources;

[0063] Step S4: After receiving the warning information, the tunnel boring machine staff enters the refuge or escape mechanism according to the warning information and the instructions of the emergency plan.

[0064] It should be noted that the overall implementation process of the present invention includes:

[0065] (1) The tunnel boring machine is started and each system is initialized and checked to ensure that the equipment is in good working condition. At the same time, the built-in sensors of the tunnel boring machine, such as vibration sensors and gas concentration sensors, start working and prepare to receive data;

[0066] (2) Using sensors and cameras on the early warning component on the end face of the tunnel boring machine, real-time monitoring information such as the geological structure, coal seam thickness, gas concentration, temperature, and humidity of the working face is collected to form a comprehensive working face monitoring data set;

[0067] (3) The computer processing system built into the tunnel boring machine receives and analyzes the monitoring data, and performs real-time trend prediction and anomaly detection based on a pre-set comprehensive early warning model. If any data anomaly is found, such as a sharp increase in gas concentration, a decrease in coal seam stability, or the presence of potential danger, the early warning mechanism is immediately triggered, and the risk level is automatically assessed through the risk assessment algorithm. At the same time, the early warning information is sent to the tunnel boring machine staff and the ground control center through the tunnel boring machine display screen or wireless communication equipment. The early warning information includes the risk level, warning type, and recommended measures;

[0068] (4) The tunnel boring machine uses the mine geographic information system to accurately locate the warning information, determine the specific location and scope of the warning area, and build an intelligent emergency decision support system based on knowledge graph and rule engine to automatically match the optimal emergency plan according to the urgency and risk level of the warning information;

[0069] (5) Generate a detailed emergency response plan, including evacuation routes, TBM parking locations, and emergency equipment instructions, and send it to TBM personnel via the TBM display or wireless communication equipment;

[0070] (6) Upon receiving the early warning information and emergency response plan, the tunnel boring machine crew shall immediately stop the tunneling operation, shut down the main systems of the tunnel boring machine, and enter the refuge or escape mechanism according to the early warning information and the instructions of the emergency plan, including wearing protective equipment, evacuating to a safe area along the designated route, and activating the emergency ventilation system. At the same time, the tunnel boring machine crew shall maintain contact with the ground control center through wireless communication equipment to report their location and status and receive further guidance and support;

[0071] (7) On the premise of ensuring safety, the ground control center organizes a professional team to investigate and analyze the cause of the accident, and formulates a recovery plan based on the investigation results, carries out necessary repairs and reinforcements on the tunnel boring machine and working face, and restarts the tunneling operation after confirming that all safety hazards have been eliminated.

[0072] The specific steps of step S2 include:

[0073] S2.1: Receive working face monitoring information and perform preprocessing, extract features from the preprocessed working face monitoring information, and generate working face feature data ,in, Indicates the N Working surface characteristic data, N Indicates the number of working face characteristic data; the working face characteristic data includes working face gas concentration data, coal seam stress state data, and working face wind speed data;

[0074] S2.2: Build a comprehensive early warning model based on machine learning methods and anomaly detection algorithms, train the comprehensive early warning model using historical working face monitoring information, and use confusion matrix to evaluate the model;

[0075] The confusion matrix is ​​used to show the comparison between the model prediction results and the actual categories, and the elements in the confusion matrix are filled based on the prediction results of the comprehensive early warning model;

[0076] Furthermore, the specific steps of building a comprehensive early warning model include:

[0077] S2.21: Clarify the application scenarios and specific requirements of the early warning model, and set the expected goals of the model, such as early warning accuracy, early warning speed, and system stability;

[0078] S2.22: Collect historical working face monitoring information from the data source, including data in normal and abnormal states, and preprocess the historical working face monitoring information, including: 1) data cleaning, such as removing noise, removing duplicate data, and filling missing values; 2) data transformation, such as normalization and standardization; 3) data partitioning, such as dividing the data set into training, validation, and test sets;

[0079] S2.23: Extract features from pre-processed historical working face monitoring information, where these features should be able to reflect the characteristics and changes of the target state, and perform feature selection to remove redundant or irrelevant features, reduce model complexity and improve prediction performance;

[0080] S2.24: Select a decision tree-based machine learning method and an anomaly detection algorithm based on local anomaly factors based on the characteristics of the problem and the data, and design the architecture and parameters of the comprehensive early warning model, such as the number of layers, number of nodes, and activation function of the neural network. The decision tree-based machine learning method and the anomaly detection algorithm based on local anomaly factors are both prior art in this field and do not constitute the inventive solution of this application, and are not described in detail here.

[0081] S2.25: Use the training set data to train the comprehensive early warning model, allowing the comprehensive early warning model to learn the patterns and regularities in the data. During the training process, use the validation set data to evaluate the comprehensive early warning model, and adjust the comprehensive early warning model parameters and architecture based on the evaluation results to optimize the comprehensive early warning model performance.

[0082] S2.26: Use the test set data to conduct a final evaluation of the trained comprehensive early warning model, verify the warning accuracy of the comprehensive early warning model, and analyze the confusion matrix evaluation indicators;

[0083] S2.27: Deploy the trained comprehensive early warning model into the actual working environment and integrate it with the monitoring system;

[0084] S2.28: Configure the interfaces for real-time input data and output of warning results of the comprehensive early warning model to ensure that the comprehensive early warning model can work normally and respond to abnormal situations in a timely manner, and regularly maintain and update the comprehensive early warning model.

[0085] Furthermore, the specific steps for using the confusion matrix to evaluate the model include:

[0086] (1) Define positive and negative categories, such as prominent danger as positive category and normal state as negative category;

[0087] (2) Obtain the prediction results of the comprehensive early warning model;

[0088] (3) Construct an empty 2×2 matrix and fill the matrix;

[0089] If the predicted category of the comprehensive early warning model and the actual category of the sample are both positive, a true positive is obtained. For example, the actual gas concentration at the monitoring point of the working face exceeds the standard, and the comprehensive early warning model also predicts that the gas concentration at the monitoring point exceeds the standard;

[0090] If the predicted category of the comprehensive early warning model is positive, but the actual category of the sample is negative, a false positive is obtained. For example, the actual gas concentration at the working face monitoring point is within the normal range, but the comprehensive early warning model predicts that the gas concentration at the monitoring point exceeds the normal range.

[0091] If the predicted category of the comprehensive early warning model is negative, but the actual category of the sample is positive, a false negative example is obtained;

[0092] If the predicted category of the comprehensive early warning model and the actual category of the sample are both negative, a true negative example is obtained;

[0093] (4) Obtain the confusion matrix by taking the true positives and false negatives as the elements of the first row of the matrix and the false positives and true negatives as the elements of the second row of the matrix;

[0094] (5) Using a confusion matrix to calculate evaluation indicators, wherein the evaluation indicators include accuracy, precision, and recall, and the calculation formulas for accuracy, precision, and recall are prior art content in this field and are not the inventive solution of this application, and are not described in detail here.

[0095] S2.3: Work surface feature data Input the trained comprehensive early warning model and use Perform trend prediction and anomaly detection to obtain anomaly detection results ,in, represents the anomaly detection result at the current time point t, g represents the constant term, r represents the order of the autoregressive part, represents the coefficient of the autoregressive part, Indicates a time point The anomaly detection result when u represents the order of the moving average part, represents the coefficient of the moving average part, Represents the moving average part The error term of order , v represents the number of external variables, represents the coefficient of external variables, represents the external variable, z represents the number of periodic patterns, represents the periodic function coefficient, represents a periodic function, represents the coefficient of the dummy variable, represents a dummy variable, represents the time trend term coefficient, represents the time trend term, represents the error term, Indicates the abnormality detection result of the Nth working surface feature data;

[0096] Furthermore, the anomaly detection result includes information about the time, location, type, and severity of the anomaly.

[0097] It should be noted that the present invention incorporates external variables, cyclical patterns, dummy variables, and time trend terms. These additional parameters and variables enable the model to consider more diverse dimensions of information during forecasting, more accurately reflecting the impact of seasonal fluctuations on forecasted values. Furthermore, the model considers the impact of unexpected events on time series data and incorporates the underlying development trends of the data during forecasting. By comprehensively considering these various factors, the modified model can more comprehensively reflect the complexity and dynamics of time series data during forecasting, enhancing the model's flexibility, accuracy, and adaptability while reducing forecast errors.

[0098] S2.4: Setting anomaly detection thresholds ,in, , represents the mean of Y, b represents the slope coefficient, represents the standard deviation of Y;

[0099] like , it means that the working surface feature data is normal data;

[0100] like , it means there is abnormal data and the early warning mechanism is triggered through the electronic control early warning platform;

[0101] S2.5: Set the warning signals as blue, yellow, orange, and red, and automatically assess the risk level based on the severity of the warning signal.

[0102] Furthermore, when the sound and light alarm is green, it means there is no danger of a sudden outburst and the working face is under normal construction. When the sound and light alarm is blue, it means that the danger of a sudden outburst has reached a threat value. At this time, it is necessary to slow down the cutting section speed and adopt intermittent operation mode to reduce the concentration of concentrated stress and reduce the danger of a sudden outburst. When the sound and light alarm is orange, it means that the danger of a sudden outburst has reached a dangerous value. At this time, it is necessary to stop the cutting section operation and cut off the power supply of the tunnel boring machine cutting head motor. By stopping the tunneling, the already very tense concentrated stress is transferred to the deep part of the working face, reducing the danger of a sudden outburst. When the sound and light alarm is red, it means that the danger of a sudden outburst has reached a critical value. At this time, it is necessary to turn on the sound and light voice prompt, issue a stop operation and evacuation order to each worker on the working face, and cut off the power supply of all equipment on the working face except the early warning platform.

[0103] The specific steps of step S3 include:

[0104] S3.1: Receive the warning information in step S2 and pre-process the warning information;

[0105] S3.2: Use the mine geographic information system to locate the warning information on the mine map, and use the GIS map display function to visually display the warning location and its surrounding environment;

[0106] S3.3: Build a mine safety knowledge graph, extract knowledge from text and images using natural language processing, and store the extracted knowledge in a structured manner in the knowledge graph;

[0107] Furthermore, the specific steps of S3.3 include:

[0108] S3.31: Clarify the application scenarios of the knowledge graph, such as mine safety monitoring, accident prevention, and emergency response, and define the model layer of the knowledge graph, including entities, relationships, attributes, such as mine, equipment, accident type, safety regulation entities, and the relationships between them;

[0109] S3.32: Acquire, clean, remove noise, and format mine safety data. Mine safety data includes: 1) structured data, such as database records; 2) semi-structured data, such as web pages and XML files; and 3) unstructured data, such as text and images.

[0110] S3.33: (1) Text knowledge extraction: Use NLP technology to perform text segmentation, part-of-speech tagging, named entity recognition, and relationship extraction to identify mine safety entities, attributes, and relationships in the text. Entities include: equipment name, accident type, personnel distribution, emergency plan, and historical cases; attributes include: equipment model, accident level; and relationships include: equipment-fault, accident-cause.

[0111] (2) Image knowledge extraction: Apply image recognition methods to identify mine safety elements in images, such as equipment status and safety hazards, and extract text information from images, such as safety signs and equipment labels;

[0112] S3.34: Fuse the knowledge extracted from text and images to solve entity alignment and relationship conflict problems, and store the fused knowledge in a graph database in Neo4j format for efficient query and reasoning.

[0113] S3.35: Continuously update and maintain the knowledge graph to ensure the timeliness and accuracy of the data.

[0114] S3.4: Design a rules engine to automatically match the optimal emergency response plan based on the urgency and risk level of the warning information. The rules engine is implemented based on conditional judgment. Emergency response plans are divided into four categories: A, B, C, and D in descending order of the urgency of the warning information.

[0115] For example, if the urgency of the warning information is "high", the risk level is 4 and it matches emergency plan A;

[0116] If the urgency of the warning information is "high", the risk level is 3 and the emergency plan B is matched;

[0117] If the urgency of the warning information is "medium", the risk level is 2 and the emergency plan C is matched;

[0118] If the urgency of the warning information is "low", the risk level is 1 and matches the default emergency plan D.

[0119] S3.5: Develop decision rules. When receiving warning information, the intelligent emergency decision support system automatically triggers the rule engine to perform decision analysis. The rule engine matches the optimal emergency plan based on the attributes of the warning information and the knowledge in the knowledge graph, and generates an emergency response plan based on the emergency plan.

[0120] If the emergency response plan is triggered, the corresponding power-off and shutdown safety measures will be automatically executed through the control loop;

[0121] Furthermore, the specific steps of S3.5 include:

[0122] S3.51: Develop decision rules to clarify under what conditions which emergency plans should be triggered. Decision rules can be based on the attributes of warning information, entity relationships, and attributes in the knowledge graph.

[0123] S3.52: Obtain a rule engine and integrate it with the early warning system to ensure that when the early warning system receives new warning information, it can automatically trigger the rule engine to perform decision analysis;

[0124] S3.53: The rule engine matches the optimal emergency plan based on the attributes of the warning information and the knowledge in the knowledge graph. Based on the matched emergency plan, it automatically generates a detailed emergency response plan, including response measures, responsibility allocation, time nodes, and rescue routes.

[0125] S3.54: Test the entire system to ensure that the rule engine can be correctly triggered, the emergency plan can be matched, and an effective emergency response plan can be generated under various circumstances. Optimize and adjust the system based on the test results.

[0126] S3.55: Deploy the system into a production environment and provide ongoing operations and management, and regularly maintain and update the system to adapt to new warning information and emergency response plans.

[0127] S3.6: Send the generated emergency response plan to staff for execution. Simultaneously, monitor the emergency response process, collect feedback information in real time, and dynamically adjust the emergency response plan based on the feedback information.

[0128] S3.7: Continuously optimize and iterate the intelligent emergency decision support system based on the actual effects and feedback information of the emergency response.

[0129] Example 2

[0130] See also Figure 4 Another embodiment provided by the present invention is a coal and gas outburst hazard early warning system for a machine-dug working face, comprising:

[0131] Data acquisition module, comprehensive early warning module, emergency response module, and risk avoidance module;

[0132] The data acquisition module is used to collect working face monitoring information in real time, including gas concentration, coal seam stress state, and wind speed parameters, using the early warning component and wireless data transmission and reception component on the end face of the roadheader body;

[0133] The comprehensive early warning module is used to evaluate and predict the risk of coal and gas outbursts using machine learning methods based on the working face monitoring information provided by the data acquisition module;

[0134] The emergency response module is used to activate the emergency plan after receiving the early warning signal and guide the on-site staff to take appropriate emergency measures;

[0135] The risk avoidance module is used to send information about safe shelters or escape routes after the early warning information is triggered, and guide staff to enter the shelter or escape mechanism according to the instructions of the emergency plan to ensure that personnel can quickly reach a safe area when danger occurs.

[0136] The comprehensive early warning module includes: data processing unit, early warning model construction unit, and risk assessment unit;

[0137] Data processing unit, used to clean, convert and store the collected working face monitoring information to ensure data quality;

[0138] An early warning model building unit is used to build a comprehensive early warning model based on historical working face monitoring information and expert knowledge, including a trend prediction model and an anomaly detection model;

[0139] The risk assessment unit is used to use the comprehensive early warning model to conduct real-time analysis of the working face monitoring information, evaluate the risk level of the hazard, and trigger the corresponding early warning mechanism.

[0140] The emergency response module includes: positioning unit, emergency plan library unit, and emergency response execution unit;

[0141] Positioning unit, used to combine early warning information with mine spatial data to achieve accurate positioning of early warning information;

[0142] The emergency plan library unit is used to store various emergency plans, including response measures under different risk levels and urgency levels;

[0143] The emergency response execution unit is used to automatically select and activate the corresponding emergency plan and implement emergency response measures based on the early warning information and risk assessment results. The emergency response measures include: shutting down the ventilation system, stopping mechanical equipment, and turning on emergency lighting and communication equipment.

[0144] The risk avoidance module includes: early warning information release unit, risk avoidance guidance unit, and safety monitoring unit;

[0145] The early warning information release unit is used to release early warning information to the tunnel boring machine workers in a timely manner through sound and light alarms and communication equipment;

[0146] The hazard guidance unit is used to provide workers with clear refuge or escape routes and operation methods according to the instructions of the emergency plan;

[0147] The safety monitoring unit is used to continuously monitor the safety of personnel during the risk avoidance process to ensure the effective implementation of risk avoidance measures.

[0148] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific embodiments. The above-mentioned specific embodiments are merely illustrative and not restrictive. Under the guidance of the present invention, ordinary technicians in this field can also change, modify, replace and modify the above-mentioned embodiments without departing from the purpose and scope of protection of the present invention. These are all protected by the present invention.

[0149] If the technical solution disclosed herein involves personal information, the product using the technical solution disclosed herein has clearly informed the individual of the personal information processing rules and obtained the individual's voluntary consent before processing the personal information. If the technical solution disclosed herein involves sensitive personal information, the product using the technical solution disclosed herein has obtained the individual's separate consent before processing the sensitive personal information and at the same time meets the "explicit consent" requirement. For example, on personal information collection devices such as cameras, a clear and prominent sign is set to inform that the individual has entered the personal information collection scope and that personal information will be collected. If the individual voluntarily enters the collection scope, it is deemed that they agree to the collection of their personal information; or on the personal information processing device, when the personal information processing rules are notified by obvious signs / information, the individual's authorization is obtained through pop-up information or by asking the individual to upload their personal information. The personal information processing rules may include information such as the personal information processor, the purpose of personal information processing, the processing method, and the type of personal information processed.

Claims

1. A method for early warning of coal and gas outburst hazards in a machine-dug working face, characterized in that: include: Step S1: Using an early warning component installed on the end face of the roadheader body, real-time working face monitoring information is collected through a wireless data transmitting and receiving component. The early warning component on the end face of the roadheader body includes a mounting shell, a mounting plate provided on the inner wall of the mounting shell, a sensor component provided on the end face of the mounting plate, and a real-time data collection component. The wireless data transmitting and receiving component is used to upload the working status and analysis results of the roadheader and receive instructions and prompt information from the ground central station. Step S2: Based on the working face monitoring information, a comprehensive early warning model is established to perform real-time trend prediction and anomaly detection on the working face monitoring information. If any data anomalies or potential hazards are found, the electronic control early warning platform immediately triggers the early warning mechanism and automatically assesses the risk level through the built-in risk assessment algorithm; Step S3: Based on the warning information, the warning information is located in conjunction with the mine geographic information system, and an intelligent emergency decision support system based on the knowledge graph and rule engine is constructed. The optimal emergency plan is automatically matched according to the urgency and risk level of the warning information. At the same time, based on the emergency plan, the emergency response plan is executed through the eight-way control loop in the power outage control component; Step S4: After receiving the warning information, the tunnel boring machine staff enters the refuge or escape mechanism according to the warning information and the instructions of the emergency plan; The specific steps of step S2 include: S2.1: Receive working face monitoring information and perform preprocessing, extract features from the preprocessed working face monitoring information, and generate working face feature data ,in, represents the Nth working face characteristic data, where N represents the number of working face characteristic data; the working face characteristic data includes working face gas concentration data, coal seam stress state data, and working face wind speed data; S2.2: Build a comprehensive early warning model based on machine learning methods and anomaly detection algorithms, train the comprehensive early warning model using historical working face monitoring information, and use confusion matrix to evaluate the model; The confusion matrix is ​​used to show the comparison between the model prediction results and the actual categories, and the elements in the confusion matrix are filled based on the prediction results of the comprehensive early warning model; The specific steps of step S2 also include: S2.3: Work surface feature data Input the trained comprehensive early warning model and use Perform trend prediction and anomaly detection to obtain anomaly detection results ,in, represents the anomaly detection result at the current time point t, g represents the constant term, r represents the order of the autoregressive part, represents the coefficient of the autoregressive part, Indicates a time point The anomaly detection result when u represents the order of the moving average part, represents the coefficient of the moving average part, Represents the moving average part The error term of order , v represents the number of external variables, represents the coefficient of external variables, represents the external variable, z represents the number of periodic patterns, represents the periodic function coefficient, represents a periodic function, represents the coefficient of the dummy variable, represents a dummy variable, represents the time trend term coefficient, represents the time trend term, represents the error term, Indicates the abnormality detection result of the Nth working surface feature data; The specific steps of step S2 also include: S2.4: Setting anomaly detection thresholds ,in, , represents the mean of Y, b represents the slope coefficient, represents the standard deviation of Y; like , it means that the working surface feature data is normal data; like , it means there is abnormal data and the early warning mechanism is triggered through the electronic control early warning platform; S2.5: Set the warning signals as blue, yellow, orange, and red, and automatically assess the risk level based on the severity of the warning signal.

2. The method for early warning of coal and gas outburst danger in a machine-dug working face according to claim 1, characterized in that: The specific steps of step S3 include: S3.1: Receive the warning information in step S2 and pre-process the warning information; S3.2: Use the mine geographic information system to locate the warning information on the mine map and use the GIS map display function to display the warning location and its surrounding environment; S3.3: Build a mine safety knowledge graph, extract knowledge from text and images using natural language processing, and store the extracted knowledge in a structured manner in the knowledge graph; S3.4: Design a rule engine to automatically match the optimal emergency plan based on the urgency and risk level of warning information; S3.5: Develop decision rules. When receiving warning information, the intelligent emergency decision support system automatically triggers the rule engine to perform decision analysis. The rule engine matches the optimal emergency plan based on the attributes of the warning information and the knowledge in the knowledge graph, and generates an emergency response plan based on the emergency plan. If the emergency response plan is triggered, the corresponding power-off and shutdown safety measures will be automatically executed through the control loop; S3.6: Send the generated emergency response plan to staff. Simultaneously, monitor the emergency response process, collect feedback information in real time, and dynamically adjust the emergency response plan based on the feedback information. S3.7: Continuously optimize and iterate the intelligent emergency decision support system based on the actual effects and feedback information of the emergency response.

3. The method for early warning of coal and gas outburst danger in a machine-dug working face according to claim 2, characterized in that: The abnormality detection result in S2.3 includes the time, location, type and severity of the abnormality; the emergency plan in S3 includes evacuating personnel, cutting off power, starting ventilation equipment, and allocating resources.

4. A coal and gas outburst hazard early warning system for a machine-dug working face, which is used to implement the coal and gas outburst hazard early warning method for a machine-dug working face according to any one of claims 1 to 3, characterized in that: include: Data acquisition module, comprehensive early warning module, emergency response module, and risk avoidance module; The data acquisition module is used to collect working face monitoring information in real time, including gas concentration, coal seam stress state, and wind speed parameters, using the early warning component and wireless data sending and receiving component on the end face of the roadheader body; The comprehensive early warning module is used to evaluate and predict the risk of coal and gas outburst using machine learning methods based on the working face monitoring information provided by the data acquisition module; The emergency response module is used to activate the emergency plan after receiving the early warning signal and guide the on-site staff to take appropriate emergency measures; The risk avoidance module is used to send information about safe shelters or escape routes after the early warning information is triggered, and guide staff to enter the shelter or escape mechanism according to the instructions of the emergency plan.

5. The coal and gas outburst hazard warning system for a machine-dug working face according to claim 4, characterized in that: The comprehensive early warning module includes: an early warning model construction unit and a risk assessment unit; The early warning model construction unit is used to construct a comprehensive early warning model based on historical working face monitoring information and expert knowledge, including a trend prediction model and an anomaly detection model; The risk assessment unit is used to use the comprehensive early warning model to perform real-time analysis on the working face monitoring information, assess the risk level of the hazard, and trigger the corresponding early warning mechanism.

6. The coal and gas outburst hazard warning system for a machine-dug working face according to claim 5, characterized in that: The emergency response module includes: a positioning unit and an emergency response execution unit; The positioning unit is used to combine the warning information with the mine spatial data to locate the warning information; The emergency response execution unit is used to automatically select and start the corresponding emergency plan and execute emergency response measures according to the early warning information and risk assessment results.

Images