Comprehensive detection system for real-time early warning of coal and gas outburst

Abstract

The invention relates to the technical field of coal and gas outburst early warning, and discloses a comprehensive detection system for coal and gas outburst real-time early warning, which comprises a data acquisition module, a data processing module, a gas prediction module, an outburst early warning module and a response module, wherein the data acquisition module is used for acquiring gas emission quantity and outburst characteristic indexes; the data processing module is used for processing and coding the gas emission quantity and the outburst characteristic indexes; the gas prediction module is used for predicting the gas emission quantity based on the gas emission quantity and the outburst characteristic indexes; the outburst early warning module performs outburst event early warning based on the prediction result of the gas emission amount; the response module performs feedback adjustment of gas emission quantity and outburst characteristic index acquisition based on the outburst early warning; according to the method, potential outburst events can be found in time by accurately analyzing the change trend of the coal seam ground stress and the gas emission amount, and the possibility of accidents is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of coal and gas outburst warning, and particularly to a comprehensive detection system for real-time warning of coal and gas outburst. Background Art

[0002] At present, for the monitoring and warning of coal and gas outburst events, a variety of technical means have been adopted, mainly including traditional monitoring methods and intelligent monitoring technologies that have emerged in recent years. Traditional monitoring methods include, but are not limited to, geological exploration, borehole measurement, and acoustic emission monitoring. To a certain extent, these methods can provide information on the structure of coal seams and indications of outburst precursors. Although certain progress has been made in existing monitoring and warning technologies, there are still some deficiencies in practical applications. Traditional monitoring methods often require manual data collection at regular intervals and cannot achieve real-time monitoring, which limits the timeliness of warning. The outburst omens vary in different mine environments, and existing warning models are often difficult to adapt to diverse mine conditions. To solve the above problems, an innovative solution is needed that can solve the warning problems in changing environments, provide adaptable and flexible warning strategies for different mines, different coal seams, and different regions of the same mine, and provide multi-level warning strategies and emergency plans to cope with outburst events of different emergency levels and reduce the risk of accidents.

[0003] For example, the Chinese patent application with the authorization announcement number CN110778364B discloses a warning method for coal and gas outburst, including: case representation; retrieval and matching; determination of signal weight values; case reuse; storage and maintenance; and obtaining a danger coefficient based on the signal change trend and case weight values of real-time online monitoring data. This method is more adaptable in terms of multiple factors and large amounts of data, can quickly and accurately obtain warning results, and the warning effect is more accurate as the number of cases increases. Compared with conventional methods, the accuracy and reliability of the warning method are greatly improved.

[0004] As disclosed in the Chinese patent application with the authorization announcement number CN116877203B, a coal and gas outburst monitoring and early warning device is provided, which includes: gas concentration sensors evenly arranged in the coal mine roadway for real-time monitoring of the gas concentration in each area range; temperature and humidity sensors distributed in the coal mine roadway, and a wind direction sensor is also arranged in the coal mine roadway; plug-in anchoring components vertically inserted and distributed on the side wall of the surrounding rock of the coal mine roadway, and the plug-in anchoring components can monitor the stress change data in the side wall of the surrounding rock in real time; a drainage monitoring mechanism erected in the coal mine roadway, and the drainage monitoring mechanism can cooperate with the plug-in anchoring components to monitor the gas drainage of the side wall of the surrounding rock in the goaf; a voice alarm system and an indicator light alarm system are respectively distributed at the wellhead and the working face for issuing strong early warning alarms; a central monitoring system equipped with data storage, real-time display, alarm triggering and data analysis, and an expert system is externally connected to the central monitoring system.

[0005] The above prior arts all have the problems raised in this background art: the early warning scheme for outburst events is difficult to adapt to diverse mine conditions.

[0006] The information disclosed in this background art section is only intended to increase the understanding of the overall background of the present invention, and should not be regarded as an admission or any form of suggestion that this information constitutes the prior art already known to those of ordinary skill in the art. Summary of the Invention

[0007] The technical problem to be solved by the present invention is to overcome the defects of the prior art, provide a comprehensive detection system for real-time early warning of coal and gas outbursts, accurately analyze the change trends of the in-situ stress of the coal seam and the gas emission volume, timely discover potential outburst events, and reduce the possibility of accidents.

[0008] To solve the above technical problems, the present invention provides the following technical solutions: The present invention provides a comprehensive detection system for real-time early warning of coal and gas outbursts, including a data acquisition module, a data processing module, a gas prediction module, an outburst early warning module, and a response module; wherein: The data acquisition module is used to acquire the gas emission volume and outburst characteristic indexes; The data processing module is used to process and encode the gas emission volume and outburst characteristic indexes; The gas prediction module predicts the gas emission volume based on the gas emission volume and outburst characteristic indexes; The outburst early warning module conducts outburst early warning based on the prediction result of the gas emission volume; The response module conducts feedback adjustment of the acquisition of the gas emission volume and outburst characteristic indexes based on the outburst early warning.

[0009] As a preferred embodiment of the comprehensive detection system for real-time early warning of coal and gas outbursts according to the present invention, the outburst characteristic indicators include coal shot frequency, coal shot intensity, microseismic frequency, and microseismic energy; The data acquisition module includes a sensor unit and a feature extraction unit. Among them, the sensor unit includes a gas sensor for measuring the gas concentration, and also includes a wind speed sensor for measuring the wind speed. The feature extraction unit calculates the gas emission volume based on the gas concentration and the wind speed. The sensor unit further includes an acoustic emission sensor for collecting the acoustic wave signals of coal shots. The feature extraction unit, based on the acoustic wave signals, counts the number of occurrences of coal shots within a monitoring period as the coal shot frequency, and calculates the average amplitude of the acoustic wave signals within a monitoring period as the coal shot intensity. The sensor unit further includes a microseismic signal sensor for collecting microseismic signals. The feature extraction unit processes the microseismic signals, counts the number of occurrences of microseismic signals within a monitoring period as the microseismic frequency, and calculates the average energy of the microseismic signals within a monitoring period as the microseismic energy.

[0010] As a preferred embodiment of the comprehensive detection system for real-time early warning of coal and gas outbursts according to the present invention, the outburst characteristic indicators further include support load, support displacement, coal seam thickness, and coal seam dip angle. The sensor unit further includes a pressure sensor for measuring the support load, and also includes a displacement sensor for collecting the support displacement. The sensor unit further includes an inclinometer for measuring the coal seam dip angle, and also includes an electromagnetic sensor for generating an electromagnetic field and receiving the electromagnetic signals reflected from underground. The feature extraction unit processes the electromagnetic signals and calculates the coal seam thickness based on the transient electromagnetic method or the surface electromagnetic method.

[0011] As a preferred embodiment of the comprehensive detection system for real-time early warning of coal and gas outbursts according to the present invention, the data processing module includes a data cleaning unit and a data sorting unit. Among them, the data cleaning unit is used to clean the data of the gas emission volume and the outburst characteristic indicators. The data sorting unit is used to sort the gas emission volume after data cleaning into a gas emission sequence according to time sequence, and respectively perform standardization processing on each outburst characteristic indicator after data cleaning, and respectively encode each standardized outburst characteristic indicator and the gas emission sequence into feature vectors.

[0012] As a preferred embodiment of the comprehensive detection system for real-time early warning of coal and gas outbursts according to the present invention, the gas prediction module is configured with a gas prediction model; the input of the gas prediction model is the feature vector of the gas emission sequence and the feature vector of each outburst feature index, and the output is the predicted values of the gas emission volume at the next N consecutive time points; the gas prediction model includes an input layer, a first LSTM layer, a second LSTM layer, a fully connected layer, and an output layer; where: The input layer is used to receive the feature vectors of the gas emission sequence and the outburst feature index, and splice the feature vectors of the gas emission sequence and the outburst feature index to form a comprehensive feature vector; The first LSTM layer is used to extract features from the comprehensive feature vector; The second LSTM layer is used to further extract the features of the comprehensive feature vector; The fully connected layer is used to map the features of the comprehensive feature vector to the output space; The output layer is used to calculate and output the predicted values of the gas emission volume.

[0013] As a preferred embodiment of the comprehensive detection system for real-time early warning of coal and gas outbursts according to the present invention, the outburst early warning module includes a first early warning unit, and the first early warning unit is configured with a first early warning strategy for performing outburst early warning according to the gas emission sequence and the predicted values of the gas emission volume; specifically as follows: Extract the most recent M gas emission volumes in the gas emission sequence, and form a gas judgment sequence with the predicted values of the N gas emission volumes in chronological order; Calculate the change index of the gas emission volume based on the gas judgment sequence, and the formula is as follows: ; Where V represents the change index of the gas emission volume; represents the value of the i-th gas emission volume in the gas judgment sequence, and the value range of i is 1, 2,..., M + N; represents the value of the (i - 1)-th gas emission volume in the gas judgment sequence; The first early warning unit is further configured with a change threshold for the change index of the gas emission volume; if the change index of the gas emission volume is higher than the change threshold, the first early warning unit sends a first early warning message to the response module.

[0014] As a preferred embodiment of the comprehensive detection system for real-time early warning of coal and gas outbursts according to the present invention, the outburst early warning module further includes a second early warning unit, and the second early warning unit is configured with a second early warning strategy for performing outburst early warning based on the gas judgment sequence, specifically as follows: Read the gas judgment sequence, and divide the gas judgment sequence into n subsequences with a length of m, where both m and n are positive integers; Calculate the fluctuation index of the gas emission volume of each subsequence. The formula is as follows: ; where, represents the fluctuation index of the gas emission volume of the j-th subsequence, and the value range of j is 1, 2,..., n; represents the value of the k-th gas emission volume of the j-th subsequence, and the value range of k is 1, 2,..., n; The second warning unit is also configured with a threshold range of the fluctuation index; if the fluctuation indices of the gas emission volumes of more than s subsequences exceed the threshold range of the fluctuation index, the second warning unit sends a second warning message to the response module.

[0015] As a preferred solution of the comprehensive detection system for real-time warning of coal and gas outburst according to the present invention, wherein: the response module includes a warning response unit; if the warning response unit receives the first warning message and does not receive the second warning message, it sends a second verification instruction to the data acquisition module; the data acquisition module responds to the second verification instruction, and real-time collects the microseismic frequency and microseismic energy and transmits them to the second warning unit; the second warning unit is also configured with a second verification strategy, and the second warning unit performs anomaly detection on the microseismic frequency and microseismic energy based on the second verification strategy. If the microseismic frequency and microseismic energy do not pass the anomaly detection, the second warning unit sends a second warning message to the response module.

[0016] As a preferred solution of the comprehensive detection system for real-time warning of coal and gas outburst according to the present invention, wherein: if the warning response unit does not receive the first warning message and receives the second warning message, it sends a first verification instruction to the data acquisition module; the data acquisition module responds to the first verification instruction, and real-time collects the coal shot frequency and gas emission volume and transmits them to the first warning unit; the first warning unit is also configured with a first verification strategy, and the first warning unit performs anomaly detection on the coal shot frequency and gas emission volume based on the first verification strategy. If the coal shot frequency and gas emission volume do not pass the anomaly detection, the first warning unit sends a first warning message to the response module.

[0017] As a preferred solution of the comprehensive detection system for real-time warning of coal and gas outburst according to the present invention, wherein: if the warning response unit receives the first warning message and receives the second warning message, it sends an emergency detection instruction to the data acquisition module; The data acquisition module is set with a monitoring period; at the beginning of each monitoring period, the data acquisition module collects the gas emission volume and outburst characteristic indicators once; when the data acquisition module receives the emergency detection instruction, the monitoring period is set as the emergency acquisition period.

[0018] Compared with the prior art, the beneficial effects achieved by the present invention are as follows: Through continuous monitoring and deep learning models, it is possible to more accurately analyze the change trends of coal seam ground stress and gas emission volume in the same time and space, timely discover potential outburst events, and reduce the possibility of accidents.

[0019] Not only paying attention to the gas emission volume, but also incorporating multi-dimensional information such as microseismic signals, coal seam structure, and support status, improving the comprehensiveness and accuracy of the early warning system. Using deep learning technology to process complex time series data, capturing the characteristics of non-linear environmental factors, and achieving accurate prediction.

[0020] Combining the dynamic characteristics of time series and the static threshold of statistics, evaluating the abnormal degree of gas emission volume from different angles, enhancing the robustness and reliability of the early warning system. Based on the outburst risk, setting different emergency plans to achieve a more scientific and efficient outburst response plan, adapting to diverse mine conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained according to these drawings. Among them: Figure 1 is a schematic structural diagram of a comprehensive detection system for real-time early warning of coal and gas outburst provided by the present invention; Figure 2 is a schematic working principle diagram of a comprehensive detection system for real-time early warning of coal and gas outburst provided by the present invention; Figure 3 is a schematic structural diagram of a gas prediction model provided by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0022] The following will specifically describe the technical solutions of the present invention through the drawings and specific embodiments. It should be understood that the specific features in the embodiments of the present invention are detailed descriptions of the technical solutions of the present invention, rather than limitations on the technical solutions of the present invention. Without conflict, the technical features in the embodiments of the present invention and the embodiments can be combined with each other.

[0023] This embodiment introduces a comprehensive detection system for real-time early warning of coal and gas outburst. Refer toFigure 1 , the system includes a data acquisition module, a data processing module, a gas prediction module, a outburst warning module, a response module, and a monitoring and display module; among which: the data acquisition module is used to collect the gas emission volume and outburst characteristic indicators; The outburst characteristic indicators include coal seam thickness, coal seam dip angle, support load, support displacement, coal shot frequency, coal shot intensity, microseismic frequency, and microseismic energy; The data acquisition module includes a sensor unit and a feature extraction unit; among which, the sensor unit includes a gas sensor for measuring the gas concentration; it also includes an air velocity sensor for measuring the air velocity; the feature extraction unit calculates the gas emission volume based on the gas concentration and the air velocity; there is a direct correlation between the gas emission volume and the coal and gas outburst event. The outburst event is usually caused by the accumulation of gas pressure in the coal seam to a certain extent, exceeding the structural strength of the coal body, resulting in the sudden release of coal and gas. By monitoring the gas emission volume, the precursor of the outburst event can be detected in time.

[0024] The sensor unit also includes an inclinometer for measuring the coal seam dip angle; it also includes an electromagnetic sensor for generating an electromagnetic field and receiving the electromagnetic signal reflected from underground; the feature extraction unit processes the electromagnetic signal and calculates the coal seam thickness based on the transient electromagnetic method or the surface electromagnetic method; the thickness of the coal seam affects the gas storage, and the dip angle of the coal seam affects the distribution of in-situ stress and the gas flow condition, both of which have an important impact on the outburst event.

[0025] The sensor unit also includes a pressure sensor for measuring the support load; it also includes a displacement sensor for collecting the support displacement. The pressure sensor can be installed on the column or crossbeam of the support. The sudden increase in the support load may be an indication of the increase in coal seam pressure; the abnormal displacement of the support may be a sign of coal seam movement, and the abnormal support load and support displacement are both precursors of the outburst event.

[0026] The sensor unit also includes an acoustic emission sensor for collecting the acoustic wave signal of the coal shot; the feature extraction unit counts the number of occurrences of the coal shot in a monitoring period based on the acoustic wave signal as the coal shot frequency, and calculates the average amplitude of the acoustic wave signal in a monitoring period as the coal shot intensity. The coal shot refers to the sound generated by the sudden fracture and energy release of the coal body due to the in-situ stress concentration in the coal seam during the coal mining process. The amplitude of the coal shot can be used to evaluate the stability of the coal seam. Too high a coal shot frequency may indicate the accumulation of coal seam pressure and the risk of coal and gas outburst.

[0027] The sensor unit also includes a microseismic signal sensor for collecting microseismic signals; the feature extraction unit processes the microseismic signals, counts the number of occurrences of the microseismic signals within a monitoring period as the microseismic frequency, and calculates the average energy of the microseismic signals within a monitoring period as the microseismic energy. Microseismic signals generally refer to seismic wave signals generated when underground rocks rupture or stress is released. When the microseismic signals increase and large-energy microseismic signals appear, it indicates a greater risk of coal and gas outburst.

[0028] The data processing module is used to process and encode the gas emission volume and outburst characteristic indicators; it includes a data cleaning unit and a data sorting unit; among them, the data cleaning unit is used to clean the data of the gas emission volume and outburst characteristic indicators; it includes outlier detection and replacement, missing value difference filling, and duplicate value deletion; the data sorting unit is used to sort the gas emission volume after data cleaning into a gas emission sequence in chronological order, and respectively perform standardization processing on each outburst characteristic indicator after data cleaning, and respectively encode each standardized outburst characteristic indicator and the gas emission sequence into feature vectors.

[0029] The gas prediction module predicts the gas emission volume based on the gas emission volume and outburst characteristic indicators; The gas prediction module is configured with a gas prediction model; the input of the gas prediction model is the feature vectors of the gas emission sequence and each outburst characteristic indicator, and the output is the predicted values of the gas emission volume at the next consecutive P time points, where P is a positive integer; refer to Figure 3 , the gas prediction model includes an input layer, a hidden layer, and an output layer; among them, the hidden layer includes a first LSTM layer, a second LSTM layer, and a fully connected layer.

[0030] The input layer is used to receive the feature vectors of the gas emission sequence and outburst characteristic indicators, and splice the feature vectors of the gas emission sequence and outburst characteristic indicators to form a comprehensive feature vector; The first LSTM layer is used to extract features from the comprehensive feature vector; the number of neurons is 128, and the return sequence is set to True so as to transfer the comprehensive feature sequence to the second LSTM layer; dropout regularization is adopted, and the proportion of discarded neurons is 0.2 to prevent overfitting; The second LSTM layer is used to further extract the features of the comprehensive feature vector; the number of neurons is 64, and the return sequence is set to False; dropout regularization is adopted, and the proportion of discarded neurons is 0.2; The fully connected layer is used to map the features of the comprehensive feature vector to the output space; the ReLU function is used as the activation function, and the number of neurons is 32; The output layer is used to calculate and output the predicted value of gas emission. A linear activation function is adopted, and the number of neurons is N, where N is the number of predicted values of gas emission; each neuron corresponds to the predicted value of gas emission at a time point.

[0031] Different mines have different environmental characteristics and different outburst risks. Therefore, it is difficult to formulate a general outburst risk criterion for different mine environments. The key to warning of coal and gas outburst lies in capturing precursor information, including changes in ground stress and gas emission. The present invention selects a plurality of outburst characteristic indexes reflecting changes in ground stress, combines with the time series of gas emission, and constructs a prediction model through a deep learning algorithm. On the one hand, it captures the time series characteristics of gas emission, and on the other hand, it learns the characteristics of other related outburst characteristic indexes, realizing the accurate prediction of gas emission under different mine environments, and providing a data basis for subsequent warning of coal and gas outburst.

[0032] The outburst warning module conducts outburst warning based on the prediction result of gas emission; The outburst warning module includes a first warning unit and a second warning unit. The first warning unit is configured with a first warning strategy for conducting outburst warning according to the gas emission sequence and the predicted value of gas emission; specifically as follows: Extract the most recent M gas emissions in the gas emission sequence, where M is a positive integer, and form a gas judgment sequence G in chronological order with the N predicted values of gas emission, in the following form: ; where represents the value of the i-th gas emission in the gas judgment sequence, and the value range of i is 1, 2,..., M + N; Calculate the change index of gas emission based on the gas warning sequence, and the formula is as follows: ; where V represents the change index of gas emission; represents the value of the (i - 1)-th gas emission in the gas judgment sequence; The first warning unit is also configured with a change threshold of the change index of gas emission; if the change index of gas emission is higher than the change threshold, the first warning unit sends a first warning message to the response module; The second warning unit is configured with a second warning strategy for conducting outburst warning based on the gas judgment sequence, specifically as follows: Read the gas judgment sequence and divide the gas judgment sequence into n subsequences with a length of m, where both m and n are positive integers; Calculate the fluctuation index of gas emission for each subsequence, and the formula is as follows: ; wherein, represents the fluctuation index of the gas emission volume of the j-th subsequence, and the value range of j is 1, 2,..., n; represents the value of the k-th gas emission volume of the j-th subsequence, and the value range of k is 1, 2,..., n; The second warning unit is also configured with a threshold range of the fluctuation index; if the fluctuation indexes of the gas emission volumes of more than s subsequences exceed the threshold range of the fluctuation index, the second warning unit sends a second warning message to the response module.

[0033] The first warning strategy is based on the dynamic characteristics of the gas judgment sequence to judge the risk of outburst events. Specifically, it evaluates the abnormality degree of the gas emission volume through the cumulative change rate. The second warning strategy is a static threshold criterion of statistics. Specifically, it performs abnormal detection by analyzing the fluctuation of the gas emission volume in each time period. Among them, if the first warning message is sent, it means that the change range of the gas emission volume is large and there is a certain outburst risk; if the second warning message is sent, it means that the fluctuation of the gas emission volume is obvious and the risk of outburst events is high; if the gas emission volume has the characteristics of large change range and obvious fluctuation at the same time, it means that the risk of outburst events is extremely high. Combining the dynamic characteristics of the time series and the static threshold of statistics can provide a reasonable and comprehensive method to judge the abnormality of the gas emission volume. By introducing multi-dimensional analysis, potential abnormal situations can be captured more comprehensively. This dual judgment mechanism combining dynamic and static analysis can not only improve the sensitivity and specificity of abnormal detection of gas emission volume, but also enhance the robustness and adaptability of the system.

[0034] The response module performs feedback regulation on the gas emission volume and the collection of outburst characteristic indicators based on the outburst warning.

[0035] The response module includes a warning response unit; if the warning response unit receives the first warning message and does not receive the second warning message, it sends a second verification instruction to the data collection module; the data collection module responds to the second verification instruction, and real-time collects the microseismic frequency and microseismic energy and transmits them to the second warning unit; the second warning unit is also configured with a second verification strategy, and the second warning unit performs abnormal detection on the microseismic frequency and microseismic energy based on the second verification strategy. If the microseismic frequency and microseismic energy do not pass the abnormal detection, the second warning unit sends a second warning message to the response module.

[0036] The second verification strategy is specifically as follows: calculate the second abnormal index, and the formula is as follows: ; wherein, represents the second anomaly index; f represents the microseismic frequency collected in real time; represents the mean value of the microseismic frequency in the most recent R monitoring periods, represents the variance of the microseismic frequency in the most recent R monitoring periods; E represents the microseismic energy collected in real time; represents the mean value of the microseismic energy in the most recent R monitoring periods, represents the variance of the microseismic energy in the most recent R monitoring periods; 、 are both weight coefficients and can be set according to actual requirements. R is a positive integer.

[0037] If the second anomaly index is greater than the preset second anomaly threshold , then the microseismic frequency and microseismic energy fail the anomaly detection; otherwise, the microseismic frequency and microseismic energy pass the anomaly detection.

[0038] Microseismic signals are usually generated by underground rock fractures or stress releases. An increase in microseismic frequency and the appearance of microseismic signals with large energy mean that the stability of the rock around the coal seam is affected, reflecting the unstable state of the coal seam. If the early warning response unit receives the first early warning information and does not receive the second early warning information, this indicates that the change range of gas emission is large, but the fluctuation situation is still within the relatively normal range. At this time, the coal seam structure may already be in the early stage of instability. Implementing the second verification strategy can improve the sensitivity of gas emission fluctuation detection, help to timely discover potential coal seam change trends, and prevent outburst risks.

[0039] Preferably, the response module further includes a plan unit; the plan unit is configured with emergency plans, and the emergency plans include an outburst threat plan, an outburst danger plan, and an outburst critical plan; When the early warning response unit receives the first early warning information, does not receive the second early warning information, and the microseismic frequency and microseismic energy pass the anomaly detection, the plan unit sends the outburst threat plan to relevant management personnel; If the early warning response unit does not receive the first early warning information and receives the second early warning information, it sends a first verification instruction to the data acquisition module; the data acquisition module responds to the first verification instruction, collects the coal shot frequency and gas emission in real time and transmits them to the first early warning unit; the first early warning unit is also configured with a first verification strategy, and the first early warning unit performs anomaly detection on the coal shot frequency and gas emission based on the first verification strategy. If the coal shot frequency and gas emission do not pass the anomaly detection, the first early warning unit sends the first early warning information to the response module.

[0040] The specific content of the first verification strategy is as follows: If the frequency of the coal bump sound collected in real time is greater than the preset frequency threshold, or the difference between the gas emission volume collected in real time and the average value of the gas emission volumes in the most recent R monitoring cycles exceeds the preset threshold range, then the frequency of the coal bump sound and the gas emission volume fail the anomaly detection; otherwise, the frequency of the coal bump sound and the gas emission volume pass the anomaly detection.

[0041] A too high frequency of coal bump sounds may indicate the accumulation of coal seam pressure. The change in coal seam pressure will affect the gas pressure, which in turn reflects the risk of gas outburst. The gas emission volume is directly related to coal and gas outburst events. The change in the gas emission volume can reflect the dynamic change of gas pressure. When the gas pressure accumulates to a certain extent, the gas emission volume will increase significantly. If the early warning response unit does not receive the first early warning information but receives the second early warning information, it means that the fluctuation of the gas emission volume is obvious, but the change amplitude has not reached the warning standard for the time being. This may indicate a change in gas pressure at this time. By implementing the first verification strategy, the abnormal change of gas pressure can be captured more timely, avoiding outburst accidents caused by sudden changes in gas pressure.

[0042] When the early warning response unit receives the second early warning information, and does not receive the first early warning information, and the frequency of the coal bump sound and the gas emission volume pass the anomaly detection, the plan unit sends the outburst danger plan to relevant management personnel; If the early warning response unit receives the first early warning information and receives the second early warning information, it sends an emergency detection instruction to the data acquisition module; The data acquisition module is set with a monitoring cycle; at the beginning of each monitoring cycle, the data acquisition module collects the gas emission volume and outburst characteristic indicators once; when the data acquisition module receives the emergency detection instruction, the monitoring cycle is set to an emergency acquisition cycle. At the same time, the plan unit sends the outburst critical plan to relevant management personnel.

[0043] The emergency acquisition cycle is shorter than the monitoring cycle during normal operation. For example, the monitoring cycle is shortened by 50% as the emergency acquisition cycle. If the early warning response unit receives the first early warning information and the second early warning information at the same time, it indicates that the risk of outburst events is extremely high. At this time, it is necessary to monitor various indicators in an all-round, high-precision and high-frequency manner in order to obtain the most accurate information in a timely manner, provide reliable data support for subsequent decision-making, and minimize the possibility of outburst accidents.

[0044] During mine operation, an emergency plan is established according to the actual conditions of the mine, and outburst early warning is continuously carried out; according to the outburst early warning results, the corresponding emergency plan is extracted to deal with the threat of potential outburst events and minimize the harm of outburst accidents.

[0045] An example of a prominent threat plan is as follows: Strengthen the emergency training of on-site operators to ensure that they understand the latest emergency measures; Prepare necessary emergency supplies and equipment, such as rescue equipment, personal protective equipment, etc.; Organize evacuation drills to ensure that all personnel are familiar with the emergency evacuation routes.

[0046] An example of a prominent danger plan is as follows: Invite experts in the fields of geology and safety for on-site assessment and guidance; Release the latest situation to all relevant personnel to ensure information transparency; Prepare to activate the emergency evacuation procedure at any time to ensure that personnel can evacuate quickly.

[0047] An example of a prominent critical plan is as follows: Immediately stop all operation activities to ensure the rapid evacuation of all personnel; Issue emergency notifications through various channels, including notifying local management and the media; Isolate the mine and prohibit unauthorized personnel from entering; Organize an expert team for post-event assessment, analyze the cause of the incident, and formulate rectification measures.

[0048] The monitoring and display module includes a visualization unit and a human-machine interaction unit; Among them, the visualization unit is used to visually display system parameters, including gas emission volume and prominent feature indicators, predicted values of gas emission volume, early warning information, and corresponding triggered emergency plans; The human-machine interaction unit provides human-machine interaction functions and supports managers to manually add or modify emergency plans. In this embodiment, the mechanism of the system modules working together is as Figure 2 shown.

[0049] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0050] The embodiments of the present invention have been described above in conjunction with the accompanying drawings, but the present invention is not limited to the above specific embodiments. The above specific embodiments are merely illustrative and not restrictive. Under the inspiration of the present invention, those of ordinary skill in the art can also make many forms without departing from the purpose and scope of the present invention, and these all fall within the protection scope of the present invention.

Claims

1. A comprehensive detection system for real-time early warning of coal and gas outbursts, characterized by: It includes data acquisition module, data processing module, gas prediction module, outburst warning module and response module; among which: The data acquisition module is used to collect gas outflow and prominent characteristic indicators; The data processing module is used to process and encode the gas outburst volume and prominent characteristic index; The gas prediction module predicts the gas emission amount based on the gas emission amount and the prominent characteristic index; The outburst warning module provides outburst warning based on the prediction results of gas outburst volume; The response module performs feedback adjustment on the gas outburst volume and the collection of outburst characteristic indicators based on the outburst warning.

2. A comprehensive detection system for real-time early warning of coal and gas outburst according to claim 1, characterized in that: The prominent characteristic indicators include coal cannon sound frequency, coal cannon sound intensity, microseismic frequency, and microseismic energy; The data acquisition module includes a sensor unit and a feature extraction unit; wherein the sensor unit includes a gas sensor for measuring gas concentration; and also includes a wind speed sensor for measuring wind speed; the feature extraction unit calculates the gas outflow amount based on the gas concentration and wind speed; The sensor unit also includes an acoustic emission sensor for collecting the acoustic wave signal of the coal cannon sound; the feature extraction unit counts the number of occurrences of the coal cannon sound in a monitoring period based on the acoustic wave signal as the coal cannon sound frequency, and calculates the average amplitude of the acoustic wave signal in a monitoring period as the coal cannon sound intensity; The sensor unit also includes a microseismic signal sensor for collecting microseismic signals; the feature extraction unit processes the microseismic signals, counts the number of occurrences of the microseismic signals within a monitoring period as the microseismic frequency, and calculates the average energy of the microseismic signals within a monitoring period as the microseismic energy.

3. A comprehensive detection system for real-time early warning of coal and gas outburst as claimed in claim 2, characterized in that: The prominent characteristic indicators also include support load, support displacement, coal seam thickness, and coal seam inclination; the sensor unit also includes a pressure sensor for measuring the support load; and a displacement sensor for collecting the support displacement; The sensor unit also includes an inclinometer for measuring the inclination of the coal seam; and an electromagnetic sensor for generating an electromagnetic field and receiving electromagnetic signals reflected from underground; the feature extraction unit processes the electromagnetic signals and calculates the thickness of the coal seam based on transient electromagnetic method or ground electromagnetic method.

4. A comprehensive detection system for real-time early warning of coal and gas outburst as claimed in claim 3, characterized in that: The data processing module includes a data cleaning unit and a data sorting unit; wherein the data cleaning unit is used to perform data cleaning on the gas outflow volume and prominent characteristic indicators; the data sorting unit is used to sort the gas outflow volume after data cleaning into a gas outflow sequence in chronological order, and to perform standardization processing on each prominent characteristic indicator after data cleaning, and to encode each prominent characteristic indicator after standardization and the gas outflow sequence into feature vectors.

5. A comprehensive detection system for real-time early warning of coal and gas outburst as claimed in claim 4, characterized in that: The gas prediction module is configured with a gas prediction model; the input of the gas prediction model is the feature vector of the gas emission sequence and the feature vector of each prominent feature indicator, and the output is the predicted value of the gas emission at N consecutive time points in the future; the gas prediction model includes an input layer, a first LSTM layer, a second LSTM layer, a fully connected layer, and an output layer; wherein: The input layer is used to receive the feature vectors of the gas emission sequence and the prominent feature index, and concatenate the feature vectors of the gas emission sequence and the prominent feature index to form a comprehensive feature vector; The first LSTM layer is used to extract features from the comprehensive feature vector; The second LSTM layer is used to extract the features of the comprehensive feature vector; The fully connected layer is used to map the features of the comprehensive feature vector to the output space; The output layer is used to calculate and output the predicted value of gas emission.

6. A comprehensive detection system for real-time early warning of coal and gas outburst according to claim 5, characterized in that: The outburst warning module includes a first warning unit, which is configured with a first warning strategy for performing outburst warning according to the gas outburst sequence and the predicted value of the gas outburst amount; specifically, as follows: Extracting the most recent M gas emission amounts in the gas emission sequence and combining them with the predicted values ​​of the N gas emission amounts in chronological order to form a gas judgment sequence; Calculating a change index of gas emission based on the gas judgment sequence; The first warning unit is also configured with a change threshold of the change index of the gas outflow volume; if the change index of the gas outflow volume is higher than the change threshold, the first warning unit sends a first warning message to the response module.

7. A comprehensive detection system for real-time early warning of coal and gas outburst according to claim 6, characterized in that: The outburst warning module further includes a second warning unit, which is configured with a second warning strategy and performs an outburst warning based on the gas judgment sequence, specifically as follows: Read the gas judgment sequence, and divide the gas judgment sequence into n subsequences of length m, where m and n are both positive integers; Calculate the fluctuation index of gas emission of each subsequence; The second warning unit is also configured with a threshold range of the fluctuation index; If the fluctuation index of the gas emission amount of more than s subsequences exceeds the threshold range of the fluctuation index, the second warning unit sends a second warning message to the response module.

8. A comprehensive detection system for real-time early warning of coal and gas outburst according to claim 7, characterized in that: The response module includes an early warning response unit; if the early warning response unit receives the first early warning information and does not receive the second early warning information, a second verification instruction is sent to the data acquisition module; The data acquisition module responds to the second verification instruction, collects the microseismic frequency and microseismic energy in real time and transmits them to the second early warning unit; the second early warning unit is also configured with a second verification strategy, and the second early warning unit performs abnormal detection on the microseismic frequency and microseismic energy based on the second verification strategy. If the microseismic frequency and microseismic energy fail to pass the abnormal detection, the second early warning unit sends a second early warning information to the response module.

9. A comprehensive detection system for real-time early warning of coal and gas outburst according to claim 8, characterized in that: If the early warning response unit does not receive the first early warning information, but receives the second early warning information, then sending a first verification instruction to the data acquisition module; The data acquisition module responds to the first verification instruction, collects the coal cannon sound frequency and gas outflow in real time and transmits them to the first early warning unit; the first early warning unit is also configured with a first verification strategy, and the first early warning unit performs abnormal detection on the coal cannon sound frequency and gas outflow based on the first verification strategy. If the coal cannon sound frequency and gas outflow fail to pass the abnormal detection, the first early warning unit sends a first early warning information to the response module.

10. A comprehensive detection system for real-time early warning of coal and gas outburst according to claim 9, characterized in that: If the warning response unit receives the first warning information and the second warning information, it sends an emergency detection instruction to the data acquisition module; The data acquisition module is provided with a monitoring cycle; at the beginning of each monitoring cycle, the data acquisition module collects gas outflow volume and prominent characteristic indicators once; When the data acquisition module receives the emergency detection instruction, the monitoring period is set as the emergency acquisition period.

Images