Method for monitoring gas outburst in coal mine tunnel

Abstract

The invention discloses a method for monitoring gas outburst in a coal mine tunnel, and the method comprises the steps: constructing a risk assessment model based on multi-factor comprehensive influence through a clean data set, carrying out the feature extraction and classification of a gas outburst risk through a support vector machine algorithm, and obtaining a risk feature vector; calling a dynamic adjustment module to optimize the flexibility of an early warning mechanism according to an abnormal marking result, adjusting an early warning threshold value and response time according to data fluctuation in an abnormal state, and determining a dynamic early warning parameter; according to the dynamic early warning parameters, real-time risk early warning signals are generated, and for potential threats of the gas outburst risk, priority ranking is carried out in combination with the current mine state, and an early warning signal sequence is obtained; and according to the final risk prevention and control instruction, activating a corresponding data recording module, performing log storage for each early warning and processing process, and obtaining a historical data file for subsequent model optimization and parameter adjustment. The efficiency and the accuracy of mine safety management are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of gas outburst monitoring, and in particular relates to a gas outburst monitoring method in a coal mine tunnel. Background Art

[0002] Coal mine safety management is crucial for ensuring the safety of miners and production stability. Its core is the effective prevention of major disasters such as gas outbursts. With increasing mining depths and increasingly complex geological conditions, the risk of gas disasters continues to rise, becoming a major issue that the coal mining industry urgently needs to address, placing higher demands on safety monitoring technologies.

[0003] Current safety monitoring methods have significant shortcomings. Many mines still rely on a single detection method, such as focusing solely on gas concentration while ignoring the combined impact of other key factors. Furthermore, traditional monitoring often relies on manual inspections or fixed standard alarms, which are difficult to adapt to dynamically changing environments. This leads to untimely warnings or frequent misjudgments, seriously compromising the effectiveness of disaster prevention and control.

[0004] Against this backdrop, the field faces several core challenges. First, due to a lack of comprehensive data collection and analysis, monitoring systems are unable to fully reflect the true state of the mine environment, directly leading to inaccurate early warnings. The increasing problem of data uniformity further complicates adaptability to dynamic changes in complex environments. Traditional fixed standards are unable to address the unique circumstances of different mines, resulting in a reduction in the flexibility and reliability of early warning mechanisms. These limitations, from data collection to early warning mechanisms, constitute a major bottleneck in the current technological system.

[0005] Therefore, how to integrate multi-faceted data and build a flexible early warning mechanism to achieve accurate and real-time identification of gas outburst risks has become a key issue in improving the level of coal mine safety management. Summary of the Invention

[0006] To solve the above technical problems, the present invention proposes a method for monitoring gas outbursts in coal mine tunnels, which realizes intelligent real-time monitoring and early warning of mine gas outburst risks, and improves the efficiency and accuracy of mine safety management.

[0007] To achieve the above-mentioned object, the present invention provides a method for monitoring gas outbursts in coal mine tunnels, comprising: collecting multi-dimensional environmental data in the mine, performing real-time monitoring on dynamic environmental changes, and obtaining an initial environmental data set;

[0008] Preprocessing the collected multi-dimensional environmental data according to the initial environmental data set, filtering the noise interference in the preprocessed data to obtain a clean data set;

[0009] Based on the clean data set, a risk assessment model based on the comprehensive influence of multiple factors was constructed to extract and classify the gas outburst risk features and obtain the final risk feature vector;

[0010] Based on the final risk feature vector and a preset threshold range, determine whether the data of each dimension exceeds the safety limit. If the data of a certain dimension exceeds the preset threshold, mark it as an abnormal state and obtain an abnormal marking result;

[0011] By using the anomaly marking results, the flexibility of the early warning mechanism is optimized, and the early warning threshold and response time are adjusted according to the data fluctuations under abnormal conditions to obtain dynamic early warning parameters;

[0012] Generate real-time risk warning signals based on the dynamic warning parameters, prioritize potential threats of gas outburst risks based on the current mine status, and obtain a warning signal sequence;

[0013] Through the early warning signal sequence, the information transmission system in the mine is linked to push the results of real-time risk identification to the control center, and the early warning signals of different priorities are processed in a hierarchical manner to obtain risk prevention and control instructions;

[0014] According to the risk prevention and control instructions, the corresponding data recording module is activated, logs are stored for each warning and processing process, and historical data archives are obtained.

[0015] Technical effect of the present invention: The present invention discloses a method for monitoring gas outbursts in coal mine tunnels, which collects multi-dimensional environmental data in the mine by deploying a multi-sensor network, pre-processes and assesses the risk of the data, and constructs a risk model based on multiple factors. The present invention can monitor the dynamic environmental changes of the mine in real time, determine whether the data exceeds the safety threshold, and flexibly adjust the early warning mechanism according to the abnormal state. At the same time, the present invention can generate a real-time risk early warning signal, and push the identification results to the control center for hierarchical processing, and finally form a risk prevention and control instruction. In addition, the present invention also includes a data recording function, which can store historical data of the early warning and processing process for subsequent optimization. The method realizes the intelligent real-time monitoring and early warning of mine gas outburst risks, and improves the efficiency and accuracy of mine safety management. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying drawings, which constitute part of this application, are intended to provide a further understanding of this application. The exemplary embodiments and descriptions of this application are intended to explain this application and do not constitute an improper limitation on this application. In the accompanying drawings:

[0017] Figure 1 The present invention is a flowchart of a method for monitoring gas outbursts in coal mine tunnels according to an embodiment of the present invention. DETAILED DESCRIPTION

[0018] It should be noted that, in the absence of conflict, the embodiments and features of the embodiments in this application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0019] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0020] like Figure 1 As shown, this embodiment provides a method for monitoring gas outbursts in coal mine tunnels, including: collecting multi-dimensional environmental data in the mine, performing real-time monitoring on dynamic environmental changes, and obtaining an initial environmental data set;

[0021] Preprocessing the collected multi-dimensional environmental data according to the initial environmental data set, filtering the noise interference in the preprocessed data to obtain a clean data set;

[0022] Based on the clean data set, a risk assessment model based on the comprehensive influence of multiple factors was constructed to extract and classify the gas outburst risk features and obtain the final risk feature vector;

[0023] Based on the final risk feature vector and a preset threshold range, determine whether the data of each dimension exceeds the safety limit. If the data of a certain dimension exceeds the preset threshold, mark it as an abnormal state and obtain an abnormal marking result;

[0024] By using the anomaly marking results, the flexibility of the early warning mechanism is optimized, and the early warning threshold and response time are adjusted according to the data fluctuations under abnormal conditions to obtain dynamic early warning parameters;

[0025] Generate real-time risk warning signals based on the dynamic warning parameters, prioritize potential threats of gas outburst risks based on the current mine status, and obtain a warning signal sequence;

[0026] Through the early warning signal sequence, the information transmission system in the mine is linked to push the results of real-time risk identification to the control center, and the early warning signals of different priorities are processed in a hierarchical manner to obtain risk prevention and control instructions;

[0027] According to the risk prevention and control instructions, the corresponding data recording module is activated, logs are stored for each warning and processing process, and historical data archives are obtained.

[0028] Furthermore, the multi-dimensional environmental data includes: gas concentration, temperature, pressure and geological stress.

[0029] Furthermore, obtaining a clean dataset includes:

[0030] Preliminary processing is performed on the collected multi-dimensional environmental data according to the initial environmental data set, and a smoothing operation is performed using a filtering tool to address noise interference in the multi-dimensional environmental data to obtain a preliminary data set;

[0031] For the preliminary data set, unify the format and adjust the value range of the multi-dimensional environmental data to obtain a data set that meets the preset specifications, and determine whether the consistency of the data set meets the requirements;

[0032] If there are still abnormal fluctuations in the data set, the data set is screened again using a data verification tool to obtain abnormal points and make corrections to determine whether the integrity of the data set meets the requirements of environmental monitoring, and obtain a corrected data set;

[0033] The modified data set is archived into a pre-established database to obtain a clean data set according to the needs of environmental monitoring.

[0034] Specifically, in the practical application of mine environmental monitoring, the use of data cleaning tools is a primary step in processing the initial environmental dataset. Data cleaning tools are primarily used to remove invalid data or erroneous records that may occur during the collection process, such as abnormal readings caused by temporary sensor failures. For example, suppose that during a certain collection operation, the gas concentration in a certain area of ​​a mine suddenly spiked to 5.0%, far exceeding the normal range. Using a cleaning tool, we can identify this erroneous data due to a temporary equipment failure and remove it, retaining data that truly reflects the environmental conditions and laying the foundation for subsequent analysis. The use of filtering tools is particularly important for addressing noise interference. Through smoothing, filtering tools can effectively reduce random fluctuations in data, such as frequent small jumps in temperature data caused by minor environmental disturbances. For example, if the temperature data at a monitoring point fluctuates repeatedly from 30.0 to 32.0 degrees Celsius over a short period of time, a filtering tool can smooth these fluctuations to a value closer to the true trend, such as around 31.0 degrees Celsius. This avoids misinterpretations caused by noise and ensures the reliability of the initial dataset. During the standardization phase, standardization tools adjust data of different dimensions to a unified format and value range. For example, gas concentration may be expressed as a percentage, ranging from 0.0% to 2.0%, while pressure data may be expressed in kilopascals, ranging from 100.0 to 200.0. Using normalization tools, these data can be mapped to a uniform interval of 0.0 to 1.0, facilitating subsequent unified analysis. For example, if the pressure data for a region is 150.0 kilopascals, normalization might adjust it to a value of 0.5, making multi-dimensional data more comparable at the same scale.

[0035] Furthermore, obtaining the final risk feature vector includes:

[0036] Based on the clean data set, the multi-dimensional environmental data is preliminarily sorted, key fields in the data dimensions are extracted, and core content related to gas outburst is separated from the preliminarily sorted data set to determine whether the data set matches the environmental monitoring requirements;

[0037] Through the preliminarily sorted data set, the core content is matched with the correlation of multiple factors. If the core content does not meet the preset threshold, the core content is adjusted to obtain a combination of factors with a high correlation degree, and whether the combination of factors meets the requirements of comprehensive analysis is determined;

[0038] Classify the highly correlated factor combinations, group the extracted features, and obtain feature subsets related to gas outbursts. Determine which feature subsets are suitable for hazard identification.

[0039] The feature subset is converted into a risk feature vector. If the deviation of the risk feature vector exceeds a preset range, the risk feature vector is corrected to obtain a final risk feature vector.

[0040] Specifically, in the context of mine environmental monitoring, the application of data filtering tools is a key step in processing clean datasets. Data filtering tools are primarily used to extract core fields relevant to gas outbursts from multi-dimensional information. For example, if a mine's clean dataset contains multiple indicators such as temperature, pressure, and gas concentration, the filtering tool will prioritize gas concentration as the key field and perform preliminary collation with other auxiliary fields, such as pressure data, to form a data set focused on gas outbursts. This approach ensures that the data aligns with environmental monitoring requirements and lays the foundation for subsequent analysis. For this preliminary collated data set, data mapping tools are used to correlate and match multiple factors. Assuming a potential correlation exists between gas concentration and pressure data, the mapping tool will analyze their changing trends. If it finds that the pressure value in a certain area is elevated at a gas concentration of 1.8%, exceeding a preset threshold, the tool will adjust the data weights to generate a highly correlated factor combination. This combination must then be further evaluated to determine if it meets the requirements for comprehensive analysis to ensure the accuracy of the analysis. When extracting features based on highly correlated factor combinations, the feature extraction tool classifies and groups these factor combinations. For example, if a feature subset of high concentration and high pressure is extracted from a combination of gas concentration and pressure, the tool will classify it as highly correlated with gas outbursts. By grouping these features, it is possible to clearly identify which feature subsets are suitable for hazard identification, providing a precise basis for subsequent risk assessment. This classification method effectively improves the targeted nature of data analysis.

[0041] Furthermore, obtaining abnormal marking results includes:

[0042] The final risk feature vectors are tested one by one, and each dimension data is compared with the preset threshold range. If a dimension data exceeds the safety limit, the dimension data is marked as pending inspection to obtain an initial set of pending inspection data;

[0043] Through the initial set of data to be inspected, the dimensional data marked as being in the state to be inspected is deeply checked, and the fluctuation of the set of data to be inspected is compared with the historical records. If the fluctuation range exceeds the boundary detection standard, it is classified as an abnormal state and the abnormal data subset is determined;

[0044] According to the abnormal data subset, each abnormal dimension data is identified and processed, the abnormal state is matched with the data monitoring requirements, and the annotation information related to the abnormal identification is obtained to obtain the abnormal marking result.

[0045] Specifically, in the context of mine gas outburst risk assessment, the use of data comparison tools is a key step in the detection and processing of risk feature vectors. Data comparison tools primarily compare data from each vector dimension against pre-set safety thresholds. Suppose, for example, that the data vectors for a mine monitoring point contain gas concentration, pressure, and temperature. The pre-set gas concentration safety range is 0.5% to 1.5%. However, during actual testing, the gas concentration at a particular monitoring point reaches 1.8%, exceeding the safety limit. The data comparison tool automatically marks it as pending for inspection, forming a preliminary data set for inspection. This approach quickly identifies potential risk points and provides a foundation for subsequent verification. In-depth verification of this preliminary data set using classification and screening tools is particularly important. Suppose, within the aforementioned data set, data from a monitoring point with a gas concentration of 1.8% is flagged. Its historical records show that the concentration has remained below 1.0% over the past week, significantly exceeding the pre-set threshold. The classification and screening tool classifies this data as abnormal and includes it in the abnormal data subset. By comparing data with historical data, we can more accurately determine whether data anomalies are isolated incidents or persistent risks, providing a reliable basis for subsequent processing. When processing a subset of abnormal data, the status annotation tool matches the abnormal dimension data with monitoring requirements and identifies them. For example, if an abnormal data point shows a gas concentration of 1.8% and persists for more than 2 hours, the status annotation tool will mark it as a high-risk anomaly based on monitoring requirements and attach relevant information such as the monitoring point location and time, forming a result set with an anomaly label. This annotation method facilitates the rapid identification of key risk points and improves processing efficiency.

[0046] Furthermore, dynamic warning parameters are obtained including:

[0047] Based on the results of the anomaly marking, the data fluctuations and fluctuation ranges are checked one by one, and the corresponding data of the fluctuation range is obtained from the historical records. If the fluctuation range exceeds the preset threshold range, it is classified as a data set that needs to be adjusted, and a list of fluctuation data to be processed is obtained;

[0048] By using the list of fluctuation data to be processed, correlation processing is performed on the early warning mechanism and the threshold adjustment, parameter configuration information related to the response time is obtained, and a preliminary parameter set for dynamic adjustment is determined;

[0049] Based on the preliminary parameter set, matching adjustments are made to the response time. If a significant deviation between the fluctuation range and historical data is detected, the adjustment frequency is recalibrated to determine the appropriate response time interval.

[0050] Through the adapted response time interval, a final check is performed on the dynamic parameters to obtain an adjustment plan that matches the early warning mechanism and obtain dynamic early warning parameters.

[0051] Specifically, in the context of mine gas monitoring, data comparison tools are crucial for the subsequent processing of anomaly flagged results. Data comparison tools are primarily used to verify that data fluctuations match the preset fluctuation range. For example, suppose the gas concentration fluctuation range at a mine monitoring point is preset to 0.2% to 0.5%. During actual monitoring, the fluctuation at a particular point reaches 0.7%. Comparing historical records reveals that the fluctuations over the past week have been below 0.3%, significantly exceeding the threshold. Therefore, this data set is classified as requiring adjustment, creating a list of pending fluctuation data. This approach quickly identifies anomalous fluctuation points, providing a basis for subsequent adjustments. Within this list of pending fluctuation data, the parameter mapping tool links the early warning mechanism with threshold adjustments. For example, if the fluctuation data at a particular monitoring point frequently exceeds the range, the parameter mapping tool extracts relevant parameter configuration information based on historical response time data. For example, it can reduce the warning trigger time from 10 minutes to 5 minutes, thereby forming a preliminary parameter set. This correlation helps ensure that the early warning mechanism matches the actual fluctuations. Using this preliminary parameter set, the time calibration tool adjusts the response time to accommodate the fluctuations. If the fluctuation range of a monitoring point shows significant deviations from historical data, such as an increase in fluctuation frequency from once a day to three times a day, the time calibration tool will recalibrate the frequency, adjusting the response time interval from 5 minutes to 3 minutes, ensuring that the warning can more promptly reflect the risk change. This calibration method can improve the sensitivity of the warning.

[0052] Furthermore, obtaining the early warning signal sequence includes:

[0053] Based on the dynamic warning parameters, environmental monitoring indicators are extracted from the mine status data, and preliminary comparisons are performed against the potential threats of the gas outburst risk. If the indicators exceed the preset threshold range, they are classified as a high-risk data group, and a preliminary risk signal set is obtained;

[0054] Prioritize the high-risk data group based on the preliminary risk signal set, determine the urgency of each signal based on the data collection frequency and risk level classification, and determine a ranked risk signal list;

[0055] According to the ranked risk signal list, real-time risk signals are matched one by one. In combination with the signal generation rules, if a signal is detected to have similar characteristics to historical high-risk data, it is preferentially marked to obtain a marked signal combination;

[0056] The early warning signal sequence is finally checked by combining the marked signals, and the signal sequence matching the gas outburst risk is determined in combination with the environmental monitoring index to obtain the early warning signal sequence.

[0057] Specifically, in the context of mine gas monitoring, for the application of dynamic early warning parameters, the implementation of a data integration tool can begin with the extraction of environmental monitoring indicators. Data integration tools are primarily used to filter out key indicators related to gas outburst risk, such as gas concentration and air velocity, from a large amount of mine status data. For example, suppose the gas concentration data at a mine monitoring point is 0.8%, while the preset threshold range is 0.2% to 0.5%. A preliminary comparison can determine that this indicator is out of range and classify it as a high-risk data group, forming a preliminary risk signal set. This approach can quickly identify potential threats and provide a basis for subsequent action. For this preliminary risk signal set, a ranking algorithm can be implemented to prioritize data collection frequency and risk level. For example, suppose a mine has multiple monitoring points. One point collects gas concentration data every hour, while another collects data every 10 minutes. The former has a medium risk level, while the latter has a high risk level. The ranking algorithm will prioritize signals with higher collection frequency and higher risk level, forming a ranked risk signal list. This ranking method helps ensure that signals with the highest urgency receive priority attention. Using the sorted risk signal list, the Signal Mapping tool matches the characteristics of real-time risk signals with historical high-risk data. For example, if the gas concentration fluctuation pattern at a monitoring point resembles the fluctuation trend preceding a major accident, the Signal Mapping tool will prioritize these patterns based on signal generation rules, forming a labeled signal combination. This matching mechanism effectively identifies potential high-risk patterns and provides a more accurate basis for early warning.

[0058] Furthermore, obtaining risk prevention and control instructions includes:

[0059] Extract key risk indicators from the mine environmental monitoring data based on the warning signal sequence, compare the warning signals with preset threshold ranges, and if an indicator exceeds the threshold range, classify the indicator as a high-risk signal group, thereby obtaining a preliminary classified signal list;

[0060] Using the signal list after the preliminary classification, the high-risk signal group is prioritized according to the urgency level. If the urgency level of a signal is higher than the preset standard, the signal is marked as a priority object to obtain a ranked signal combination;

[0061] According to the sorted signal combination, the marked signals are pushed to the control center in real time, and signals of different priority levels are hierarchically classified. If a signal has the highest priority, it is placed at the front of the transmission queue to obtain a hierarchical signal set;

[0062] Through the layered signal set, the corresponding risk prevention and control rules are matched for each type of signal, the final adapted instruction content is determined, and the risk prevention and control instruction is obtained.

[0063] Specifically, in the context of mine environmental monitoring, data integration tools can be used to extract key risk indicators from multiple dimensions. For indicators such as gas concentration and tunnel temperature, the data integration tool compares real-time data against pre-set thresholds. For example, suppose the gas concentration at a mine monitoring point is 0.8%, while the threshold range is 0.2% to 0.5%. If this exceeds the upper limit, the point is immediately classified as a high-risk signal. This approach quickly identifies potential risk points and creates a preliminary categorized list of signals, laying the foundation for subsequent action. Based on this preliminary categorized list, a ranking tool prioritizes the high-risk signal groups based on urgency. Urgency is determined based on the magnitude of the indicator exceedance and the importance of the monitoring point location. For example, if a monitoring point in a main tunnel has a gas concentration exceeding the limit by 0.3%, while another monitoring point in a secondary tunnel has a 0.1% exceedance, the ranking tool will prioritize the former. This categorization ensures that risk signals in critical areas receive priority attention, improving resource allocation. After the signal combinations are sorted, the information transmission tool pushes the marked signals to the control center in real time and categorizes them by priority. For example, if a signal is marked as the highest priority, the transmission tool places it at the top of the transmission queue, ensuring that the control center receives it first. This hierarchical transmission mechanism effectively avoids information congestion, ensuring that urgent signals are not delayed, and buying time for a rapid response.

[0064] The present invention discloses a method for monitoring gas outbursts in coal mine tunnels. By deploying a multi-sensor network to collect multi-dimensional environmental data in the mine, the data is pre-processed and risk assessed, and a risk model based on multiple factors is constructed. The present invention can monitor the dynamic environmental changes of the mine in real time, determine whether the data exceeds the safety threshold, and flexibly adjust the early warning mechanism according to the abnormal state. At the same time, the present invention can generate a real-time risk early warning signal, and push the identification results to the control center for hierarchical processing, and finally form a risk prevention and control instruction. In addition, the present invention also includes a data recording function, which can store historical data of the early warning and processing process for subsequent optimization. The method realizes the intelligent real-time monitoring and early warning of mine gas outburst risks, and improves the efficiency and accuracy of mine safety management.

[0065] The above are merely preferred embodiments of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A method for monitoring gas outburst in coal mine tunnels, characterized in that: include: Collect multi-dimensional environmental data within the mine, conduct real-time monitoring of dynamic environmental changes, and obtain the initial environmental data set; Preprocessing the collected multi-dimensional environmental data according to the initial environmental data set, filtering the noise interference in the preprocessed data to obtain a clean data set; Based on the clean data set, a risk assessment model based on the comprehensive influence of multiple factors was constructed to extract and classify the gas outburst risk features and obtain the final risk feature vector; Based on the final risk feature vector and a preset threshold range, determine whether the data of each dimension exceeds the safety limit. If the data of a certain dimension exceeds the preset threshold, mark it as an abnormal state and obtain an abnormal marking result; By using the anomaly marking results, the flexibility of the early warning mechanism is optimized, and the early warning threshold and response time are adjusted according to the data fluctuations under abnormal conditions to obtain dynamic early warning parameters; Generate real-time risk warning signals based on the dynamic warning parameters, prioritize potential threats of gas outburst risks based on the current mine status, and obtain a warning signal sequence; Through the early warning signal sequence, the information transmission system in the mine is linked to push the results of real-time risk identification to the control center, and the early warning signals of different priorities are processed in a hierarchical manner to obtain risk prevention and control instructions; According to the risk prevention and control instructions, the corresponding data recording module is activated, logs are stored for each warning and processing process, and historical data archives are obtained.

2. The method for monitoring gas outburst in a coal mine tunnel according to claim 1, wherein: The multi-dimensional environmental data includes: gas concentration, temperature, pressure and geological stress.

3. The method for monitoring gas outburst in coal mine tunnels according to claim 1, characterized in that: Obtaining a clean dataset involves: Preliminary processing is performed on the collected multi-dimensional environmental data according to the initial environmental data set, and a smoothing operation is performed using a filtering tool to address noise interference in the multi-dimensional environmental data to obtain a preliminary data set; For the preliminary data set, unify the format and adjust the value range of the multi-dimensional environmental data to obtain a data set that meets the preset specifications, and determine whether the consistency of the data set meets the requirements; If there are still abnormal fluctuations in the data set, the data set is screened again using a data verification tool to obtain abnormal points and make corrections to determine whether the integrity of the data set meets the requirements of environmental monitoring, and obtain a corrected data set; The modified data set is archived into a pre-established database to obtain a clean data set according to the needs of environmental monitoring.

4. The method for monitoring gas outburst in coal mine tunnels according to claim 1, wherein: Obtaining the final risk feature vector includes: Based on the clean data set, the multi-dimensional environmental data is preliminarily sorted, key fields in the data dimensions are extracted, and core content related to gas outburst is separated from the preliminarily sorted data set to determine whether the data set matches the environmental monitoring requirements; Through the preliminarily sorted data set, the core content is matched with the correlation of multiple factors. If the core content does not meet the preset threshold, the core content is adjusted to obtain a combination of factors with a high correlation degree, and whether the combination of factors meets the requirements of comprehensive analysis is determined; Classify the highly correlated factor combinations, group the extracted features, and obtain feature subsets related to gas outbursts. Determine which feature subsets are suitable for hazard identification. The feature subset is converted into a risk feature vector. If the deviation of the risk feature vector exceeds a preset range, the risk feature vector is corrected to obtain a final risk feature vector.

5. The method for monitoring gas outburst in coal mine tunnels according to claim 1, characterized in that: Obtaining abnormal marking results includes: The final risk feature vectors are tested one by one, and each dimension data is compared with the preset threshold range. If a dimension data exceeds the safety limit, the dimension data is marked as pending inspection to obtain an initial set of pending inspection data; Through the initial set of data to be inspected, the dimensional data marked as being in the state to be inspected is deeply checked, and the fluctuation of the set of data to be inspected is compared with the historical records. If the fluctuation range exceeds the boundary detection standard, it is classified as an abnormal state and the abnormal data subset is determined; According to the abnormal data subset, each abnormal dimension data is identified and processed, the abnormal state is matched with the data monitoring requirements, and the annotation information related to the abnormal identification is obtained to obtain the abnormal marking result.

6. The method for monitoring gas outburst in coal mine tunnels according to claim 1, characterized in that: Dynamic warning parameters include: Based on the results of the anomaly marking, the data fluctuations and fluctuation ranges are checked one by one, and the corresponding data of the fluctuation range is obtained from the historical records. If the fluctuation range exceeds the preset threshold range, it is classified as a data set that needs to be adjusted, and a list of fluctuation data to be processed is obtained; By using the list of fluctuation data to be processed, correlation processing is performed on the early warning mechanism and the threshold adjustment, parameter configuration information related to the response time is obtained, and a preliminary parameter set for dynamic adjustment is determined; Based on the preliminary parameter set, matching adjustments are made to the response time. If a significant deviation between the fluctuation range and historical data is detected, the adjustment frequency is recalibrated to determine the appropriate response time interval. Through the adapted response time interval, a final check is performed on the dynamic parameters to obtain an adjustment plan that matches the early warning mechanism and obtain dynamic early warning parameters.

7. The method for monitoring gas outburst in coal mine tunnels according to claim 1, characterized in that: The sequence of obtaining early warning signals includes: Based on the dynamic warning parameters, environmental monitoring indicators are extracted from the mine status data, and preliminary comparisons are performed against the potential threats of the gas outburst risk. If the indicators exceed the preset threshold range, they are classified as a high-risk data group, and a preliminary risk signal set is obtained; Prioritize the high-risk data group based on the preliminary risk signal set, determine the urgency of each signal based on the data collection frequency and risk level classification, and determine a ranked risk signal list; According to the ranked risk signal list, real-time risk signals are matched one by one. In combination with the signal generation rules, if a signal is detected to have similar characteristics to historical high-risk data, it is preferentially marked to obtain a marked signal combination; The marked signal combination is used to perform a final check on the early warning signal sequence, and the signal sequence that matches the gas outburst risk is determined in combination with environmental monitoring indicators to obtain the early warning signal sequence.

8. The method for monitoring gas outburst in coal mine tunnels according to claim 1, characterized in that: Obtaining risk prevention and control instructions includes: Extract key risk indicators from the mine environmental monitoring data based on the warning signal sequence, compare the warning signals with preset threshold ranges, and if an indicator exceeds the threshold range, classify the indicator as a high-risk signal group, thereby obtaining a preliminary classified signal list; Using the signal list after the preliminary classification, the high-risk signal group is prioritized according to the urgency level. If the urgency level of a signal is higher than the preset standard, the signal is marked as a priority object to obtain a ranked signal combination; According to the sorted signal combination, the marked signals are pushed to the control center in real time, and signals of different priority levels are hierarchically classified. If a signal has the highest priority, it is placed at the front of the transmission queue to obtain a hierarchical signal set; Through the layered signal set, the corresponding risk prevention and control rules are matched for each type of signal, the final adapted instruction content is determined, and the risk prevention and control instruction is obtained.

Images