A coal mine working face roof caving early warning method and system

Abstract

The application discloses a coal mine working face roof collapse early warning method and system, and the method comprises the steps of constructing a collapse key feature extraction model, acquiring historical monitoring data, preprocessing the historical monitoring data, constructing time series data according to the preprocessed data, inputting the time series data into the collapse key feature extraction model to extract key features, and obtaining key feature types of the preprocessed data; based on a swallow-tail mutation model, the key feature types are used to obtain change thresholds of the key feature types; current monitoring data is acquired, feature values of the key feature types are calculated according to the current monitoring data, the feature values are compared with the change thresholds corresponding to the key feature types, comparison results are obtained, early warning levels of the roof collapse are obtained according to the comparison results, and corresponding early warning signals are sent according to the early warning levels. The application improves the precision of the coal mine working face roof collapse early warning, and realizes the advanced early warning of the large-area sudden collapse of the roof.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of coal mine roof disaster prevention, and particularly relates to a coal mine working face roof collapse early warning method, system, terminal and computer readable storage medium. BACKGROUND

[0002] In the process of coal mining, large-area sudden roof collapse is a common and serious disaster type, which has great threat to mine safety. Large-area sudden collapse is usually caused by sudden instability of the roof strata after long-term stress, resulting in the collapse of a large range of rock strata as a whole, which endangers the safety of underground workers and equipment. Although the existing roof monitoring and early warning technology can provide feedback information on the roof state to a certain extent, it still has great limitations in early warning of large-area sudden roof collapse disaster.

[0003] The existing large-area sudden roof collapse disaster monitoring system mainly relies on single or a few monitoring indicators such as support working resistance, roof separation displacement, anchor rod and cable stress, and suspended roof area. These systems usually set simple thresholds, and the system issues an alarm when the monitoring data exceeds the set threshold.

[0004] However, this static alarm mechanism also has serious shortcomings in complex underground environments, mainly manifested as early warning lag, insufficient use of monitoring data, lack of mutation recognition ability, inability to timely capture the critical changes of indicators, and difficulty in accurately judging the stability of the roof strata.

[0005] Therefore, the prior art still needs to be improved and developed. SUMMARY

[0006] The main purpose of the present application is to provide a coal mine working face roof collapse early warning method, system, terminal and computer readable storage medium, which aims to solve the problem of insufficient use of monitoring data and lack of mutation recognition ability in the prior art, resulting in insufficient prediction accuracy of large-area collapse of coal mines.

[0007] To achieve the above purpose, the present application provides a coal mine working face roof collapse early warning method, which comprises the following steps:

[0008] A collapse key feature extraction model is constructed, historical monitoring data is obtained from the monitoring instruments of the coal mine, the historical monitoring data is preprocessed, and preprocessed data is obtained;

[0009] According to the preprocessed data, time series data is constructed, the time series data is input into the collapse key feature extraction model for key feature extraction, and a plurality of key feature types of the preprocessed data are obtained.

[0010] According to the multiple key feature types, a variation threshold of each key feature type is obtained based on a swallowtail mutation model;

[0011] Current monitoring data is obtained, feature values of the multiple key feature types are calculated according to the current monitoring data, the feature values are compared with the variation thresholds corresponding to the key feature types to obtain comparison results, a warning level of roof collapse is obtained according to the comparison results, and a corresponding warning signal is sent according to the warning level.

[0012] Optionally, the coal mine working face roof collapse warning method, wherein the monitoring instrument comprises a support resistance sensor, a roof separation meter, an anchor rod and anchor cable stress sensor, and a laser range finder;

[0013] The support resistance sensor is arranged on the support to monitor the change of support working resistance, the roof separation meter is arranged in the distribution area of the roof rock and the roadway roof to monitor the separation displacement change of the roof rock, the anchor rod and anchor cable stress sensor is arranged on the anchor rod and anchor cable to monitor the stress change of the anchor rod and anchor cable in real time, and the laser range finder is arranged on the guide rail where the return air corner support tail beam is connected with the coal wall to dynamically monitor the suspended roof area to capture the expansion of the suspended roof area.

[0014] Optionally, the coal mine working face roof collapse warning method, wherein the dynamic monitoring of the suspended roof area to capture the expansion of the suspended roof area specifically comprises:

[0015] Three groups of measurement data of the laser range finder at three positions of the return air corner close to the coal wall side, the coal wall and the middle of the support, and the return air corner close to the support side are respectively obtained by a preset measurement method;

[0016] The three groups of measurement data are matched with theoretical data, if the measurement data and the theoretical data are not matched, it indicates that the suspended roof does not exist, and the measurement data is removed, and if the measurement data and the theoretical data are matched, the measurement data matched with the theoretical data is taken as target measurement data, and the suspended roof area is calculated according to the target measurement data;

[0017] The preset measurement method comprises the following steps: the laser range finder is perpendicular to the floor, a first vertical distance from the floor to the roof measured by the laser range finder is obtained, the laser range finder is respectively deflected by 30 degrees and 60 degrees to the goaf direction, a second vertical distance and a third vertical distance from the floor to the roof measured by the laser range finder are obtained, the laser range finder is horizontal to the floor, a first horizontal distance from the floor to the roof measured by the laser range finder is obtained, and the laser range finder is respectively deflected by 30 degrees and 60 degrees to the support direction, a second horizontal distance and a third horizontal distance from the floor to the roof measured by the laser range finder are obtained.

[0018] Optionally, the coal mine working face roof collapse early warning method, wherein the historical monitoring data is preprocessed to obtain preprocessed data, specifically comprising:

[0019] The historical monitoring data is subjected to outlier rejection to obtain first intermediate data, and the missing data in the first intermediate data is subjected to mean filling to obtain second intermediate data.

[0020] The second intermediate data is subjected to noise reduction processing by a median filtering or Gaussian filtering method to obtain third intermediate data, and the third intermediate data is subjected to normalization processing to obtain preprocessed data scaled to a specified range.

[0021] Optionally, the coal mine working face roof collapse early warning method, wherein the collapse key feature extraction model is a support vector machine model or a random forest model.

[0022] The time series data is constructed according to the preprocessed data, the time series data is input into the collapse key feature extraction model for key feature extraction to obtain multiple key feature types of the preprocessed data, specifically comprising:

[0023] Key data with records of sudden coal mine roof collapse events is selected from the preprocessed data, and time series data is constructed according to the key data.

[0024] The time series data is input into the support vector machine model or the random forest model for key feature extraction to obtain support bracket working resistance change rate, roof separation displacement increment, anchor rod and anchor cable stress increase, and periodic collapse area of suspended roof of the preprocessed data.

[0025] Optionally, the coal mine working face roof collapse early warning method, wherein based on the swallowtail catastrophe model, the change threshold of each key feature is obtained according to multiple key features, specifically comprising:

[0026] Multiple key features are input into the swallowtail catastrophe model for change threshold setting to obtain the change threshold of each key feature.

[0027] Each change threshold of the key feature includes a normal change threshold, a dangerous change threshold, and a critical change threshold.

[0028] Optionally, the coal mine working face roof collapse early warning method, wherein the current monitoring data includes first target data and second target data.

[0029] The current monitoring data is acquired, feature values of a plurality of key feature types are calculated according to the current monitoring data, the feature values are compared with variation thresholds corresponding to the key feature types to obtain comparison results, and the comparison results are compared with the variation thresholds corresponding to the key feature types.

[0030] At the beginning and end of a preset time interval, the first target data and the second target data monitored by the monitoring instrument in real time are acquired once respectively;

[0031] The first target data includes a first support working resistance, a first roof separation displacement, a first anchor rod and anchor cable stress, and a first hanging roof area, and the second target data includes a second support working resistance, a second roof separation displacement, a second anchor rod and anchor cable stress, and a second hanging roof area.

[0032] A target support working resistance variation rate is calculated according to the first support working resistance and the second support working resistance, a target roof separation displacement increment is calculated according to the first roof separation displacement and the second roof separation displacement, a target anchor rod and anchor cable stress increment is calculated according to the first anchor rod and anchor cable stress and the second anchor rod and anchor cable stress, and a target hanging roof periodic collapse area is calculated according to the first hanging roof area and the second hanging roof area.

[0033] The target support working resistance variation rate, the target roof separation displacement increment, the target anchor rod and anchor cable stress increment, and the target hanging roof periodic collapse area are compared with respective corresponding critical variation thresholds to obtain comparison results.

[0034] Optionally, the coal mine working face roof collapse early warning method, wherein, according to the comparison results, a roof collapse early warning level is obtained, and a corresponding early warning signal is sent according to the early warning level, and the method specifically comprises:

[0035] According to the number of feature values of the key feature types exceeding the corresponding critical variation thresholds, a judgment of the early warning level is made.

[0036] If two or fewer feature values in the comparison results exceed the corresponding critical variation thresholds, the early warning level of the roof collapse is level two, and a preliminary early warning signal is sent.

[0037] If three or more feature values in the comparison results exceed the corresponding critical variation thresholds, the early warning level of the roof collapse is level one, and an emergency early warning signal is sent.

[0038] In addition, in order to achieve the above-mentioned purpose, the present application also provides a coal mine working face roof collapse early warning system, wherein the coal mine working face roof collapse early warning system comprises:

[0039] The data acquisition module is configured to construct a collapse key feature extraction model, acquire historical monitoring data from monitoring instruments of the coal mine, pre-process the historical monitoring data, and obtain pre-processed data.

[0040] The key feature extraction module is configured to construct time series data based on the pre-processed data, input the time series data into the collapse key feature extraction model for key feature extraction, and obtain a plurality of key feature types of the pre-processed data.

[0041] The change threshold generation module is configured to obtain a change threshold of each key feature type based on the plurality of key feature types according to a swallowtail catastrophe model.

[0042] The collapse early warning module is configured to acquire current monitoring data, calculate feature values of the plurality of key feature types based on the current monitoring data, compare the feature values with the change threshold corresponding to the key feature types, obtain a comparison result, obtain a warning level of roof collapse based on the comparison result, and issue a corresponding warning signal based on the warning level.

[0043] In addition, to achieve the above object, the present application also provides a terminal, wherein the terminal comprises a memory, a processor, and a coal mine working face roof collapse early warning program stored in the memory and executable on the processor, and the coal mine working face roof collapse early warning program realizes the steps of the coal mine working face roof collapse early warning method when executed by the processor.

[0044] In addition, to achieve the above object, the present application also provides a computer readable storage medium, wherein the computer readable storage medium stores a coal mine working face roof collapse early warning program, and the coal mine working face roof collapse early warning program realizes the steps of the coal mine working face roof collapse early warning method when executed by a processor.

[0045] In the present application, a collapse key feature extraction model is constructed, historical monitoring data is acquired from monitoring instruments of the coal mine, the historical monitoring data is pre-processed, and pre-processed data is obtained; time series data is constructed based on the pre-processed data, the time series data is input into the collapse key feature extraction model for key feature extraction, and key feature types of the pre-processed data are obtained; a change threshold of the key feature type is obtained based on the key feature type according to a swallowtail catastrophe model; current monitoring data is acquired, feature values of the key feature type are calculated based on the current monitoring data, the feature values are compared with the change threshold corresponding to the key feature type, a comparison result is obtained, a warning level of roof collapse is obtained based on the comparison result, and a corresponding warning signal is issued based on the warning level. The present application improves the precision of the coal mine working face roof collapse early warning and realizes the advanced early warning of the large-area sudden roof collapse. Attached Figure Description

[0046] Figure 1 This is a flowchart of a preferred embodiment of the coal mine working face roof collapse early warning method of the present invention;

[0047] Figure 2 This is a schematic diagram of a preferred embodiment of the coal mine working face roof collapse early warning system of the present invention;

[0048] Figure 3 This is a schematic diagram of the operating environment of a preferred embodiment of the terminal of the present invention. Detailed Implementation

[0049] This application provides a method, system, terminal, and computer-readable storage medium for early warning of roof collapse in coal mine working faces. To make the objectives, technical solutions, and effects of this application clearer and more explicit, the following detailed description, with reference to the accompanying drawings and embodiments, further illustrates this application. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0050] It will be understood by those skilled in the art that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains. It should also be understood that terms such as those defined in general dictionaries should be understood to have the same meaning as in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless specifically defined as herein.

[0051] Furthermore, if the embodiments of this invention involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. If the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by this invention.

[0052] The preferred embodiment of the coal mine working face roof collapse early warning method of the present invention, such as... Figure 1 As shown, the early warning method for roof collapse at a coal mine working face includes the following steps:

[0053] Step S100: Construct a key feature extraction model for landslides, obtain historical monitoring data from the monitoring instruments in the coal mine, preprocess the historical monitoring data, and obtain preprocessed data.

[0054] It can be understood that a support vector machine (SVM) or a random forest model is used to construct a collapse key feature extraction model, and historical monitoring data used for training the collapse key feature extraction model is obtained from a monitoring instrument of the coal mine.

[0055] In the embodiment, a plurality of monitoring instruments are arranged in a coal mining face and a roadway of the coal mine to monitor key parameters of support working resistance, roof separation displacement, anchor rod and cable stress, and suspended roof area. Specifically, the monitoring instruments include a support resistance sensor, a roof separation instrument, an anchor rod and cable stress sensor, and a laser range finder. The support resistance sensor is arranged on a support to monitor changes in support working resistance. The roof separation instrument is arranged in a distribution area of a roof stratum and a roadway roof to monitor changes in separation displacement of the roof stratum. The anchor rod and cable stress sensor is arranged on an anchor rod and a cable to monitor changes in anchor rod and cable stress in real time. The laser range finder is arranged on a guide rail at which a return air corner support tail beam meets a coal wall to dynamically monitor a suspended roof area to capture expansion of the suspended roof area.

[0056] Further, the laser range finder is used to dynamically monitor the suspended roof area, and specifically includes the following steps.

[0057] Three groups of measurement data of the laser range finder at three positions of a return air corner close to a coal wall side, a coal wall and a support middle, and a return air corner close to a support side are respectively obtained by a preset measurement method.

[0058] The three groups of measurement data are matched with theoretical data. If the measurement data and the theoretical data are not matched, it is indicated that the suspended roof does not exist, and the measurement data is removed. If the measurement data and the theoretical data are matched, the measurement data matched with the theoretical data is taken as target measurement data, and a suspended roof area is calculated according to the target measurement data.

[0059] The preset measurement method includes the following steps. The laser range finder is perpendicular to a floor to obtain a first vertical distance from the floor to a roof measured by the laser range finder. The laser range finder is respectively deflected by 30 degrees and 60 degrees to a goaf direction to obtain a second vertical distance and a third vertical distance from the floor to the roof measured by the laser range finder. The laser range finder is horizontal to the floor to obtain a first horizontal distance from the floor to the roof measured by the laser range finder. The laser range finder is respectively deflected by 30 degrees and 60 degrees to a support direction to obtain a second horizontal distance and a third horizontal distance from the floor to the roof measured by the laser range finder.

[0060] It can be understood that the guide rail connected with the coal wall is arranged at the position of the tail beam of the return air corner support, and the laser range finder is arranged on the guide rail, and the contact position of the laser range finder is arranged as a telescopic rubber soft connection; the laser range finder can be freely adjusted in angle at the position of the guide rail, and is connected with a data collector to monitor and collect data in real time; the automatic measurement method of the hanging roof area is as follows: the laser range finder is located close to the coal wall side of the return air corner, the laser range finder is perpendicular to the floor, the distance from the floor to the roof is measured, then the laser range finder is deflected by 30° to the goaf direction for distance measurement, and then the laser range finder is deflected by 30° (i.e. deflected by 60°) for distance measurement; then the laser range finder is horizontal to the floor, and is deflected by 30° and 60° to the support direction respectively, and then the laser range finder is perpendicular to the floor, and the previous operation is repeated, which is the first group of measurement results. Then the laser range finder is moved to the position between the coal wall and the support, and distance measurement is continued according to the first measurement method, which is the second group of measurement results. Finally, the laser range finder is moved to the side close to the support, the laser range finder is perpendicular to the floor, the distance from the floor to the roof is measured, then the laser range finder is deflected by 30° and 60° to the goaf direction for distance measurement, which is the third group of measurement results. When the measurement data does not match the theoretical data, it means that the hanging roof range has been exceeded. The above measurement results are automatically calculated to obtain the hanging roof area.

[0061] Further, the historical monitoring data is subjected to outlier rejection to obtain first intermediate data, and the missing data in the first intermediate data is subjected to mean filling to obtain second intermediate data.

[0062] The second intermediate data is subjected to noise reduction processing by a median filtering or Gaussian filtering method to obtain third intermediate data, and the third intermediate data is subjected to normalization processing to obtain preprocessed data scaled to a specified range.

[0063] It can be understood that first, the historical monitoring data is subjected to outlier rejection, such as calculating outliers by a standard deviation method, and the outliers are rejected to ensure the rationality of the data, to obtain first intermediate data; then, the missing data in the first intermediate data is subjected to mean filling, which is a commonly used missing value processing method suitable for continuous data. It calculates the mean value of a variable in the data set, and then uses the mean value to fill the missing values to obtain second intermediate data; then, the second intermediate data is subjected to noise reduction processing by a median filtering or Gaussian filtering method to obtain third intermediate data; finally, the third intermediate data is subjected to normalization processing, which is an important step in data preprocessing, and it helps to scale the data to a specific interval (usually [0, 1] or [-1, 1]) to facilitate subsequent data analysis and machine learning algorithm processing, to obtain preprocessed data scaled to a specified range.

[0064] Step S200, constructing time series data according to the preprocessed data, inputting the time series data into the collapse key feature extraction model for key feature extraction, and obtaining multiple key feature types of the preprocessed data.

[0065] Specifically, key data with records of sudden roof collapse events in coal mines is selected from the preprocessed data, and time series data is constructed according to the key data;

[0066] It can be understood that the preprocessed data includes preprocessed support working resistance, roof separation displacement, anchor rod and anchor cable stress, and suspended roof area. These data are obtained by sensors arranged and preprocessed, and real-time records of changes of each monitoring index at different time points. These monitoring data constitute time series, reflecting the dynamic changes of the roof state over time. In particular, the records of the sudden roof collapse events in the history of coal mines contain key information such as the time variation and change rate of each monitoring index before the collapse. The roof stability features in different states can be identified according to these data.

[0067] Further, the time series data is input into the support vector machine model or the random forest model for key feature extraction, and the support working resistance change rate, the roof separation displacement increment, the anchor rod and anchor cable stress increase, and the suspended roof periodic collapse area of the preprocessed data are obtained.

[0068] The time series data is input into the support vector machine model or the random forest model for key feature extraction, and representative monitoring features are generated, including: support working resistance change rate: the time series of support resistance is differentiated to obtain its change rate as an early warning index, and a large increase in general change rate is a precursor feature of roof collapse. Roof separation displacement increment: the instability of the roof is judged by monitoring the displacement change of the roof, and a large increase in the separation amount generally means that the stability of the roof decreases. Large increase in anchor rod and anchor cable stress: the possible failure signs of anchor rod and anchor cable are identified by detecting the stress change of anchor rod and anchor cable, and in general, the greater the stress, the more unstable the roof. Suspended roof periodic collapse area: the periodic expansion of the suspended roof area is monitored by means such as laser ranging. In general, the larger the suspended roof area, the more likely the roof is to collapse.

[0069] Step S300, based on the swallow-tailed mutation model, obtaining a change threshold of each key feature type according to multiple key feature types.

[0070] It can be understood that the swallow-tail mutation model is a powerful tool for understanding and predicting non-continuous sudden change phenomena in nature and social fields. The swallow-tail mutation model reveals the possible drastic state changes of a system under the action of control parameters through folded and intersecting curved surface paths. The three-dimensional curved surface morphology depicts the complex transition process between multiple stable states and unstable states of the system. Especially in the study of roof asymptotic instability, the swallow-tail mutation model provides higher prediction accuracy and more detailed analysis capability than traditional methods. The core of the swallow-tail mutation model is to extract nonlinear features from time series to identify potential precursors of roof asymptotic instability.

[0071] In this embodiment, combined with the data characteristics obtained in step three, the swallow-tail mutation model is used to set the change threshold of the monitoring data for classification to determine whether the current state belongs to normal, dangerous or critical.

[0072] Specifically, the plurality of key features are input into the swallow-tail mutation model for change threshold setting, and the change threshold of each key feature is obtained, that is, the change threshold of each parameter is mainly obtained through a plurality of key features and in combination with the swallow-tail mutation model. Wherein, the change threshold of each key feature includes normal change threshold, dangerous change threshold and critical change threshold. Further, a coupling relationship between the characteristics of support working resistance, roof separation displacement, anchor stress and suspended roof area and the roof state can be established, and the relationship between the plurality of key features can be established through the mapping relationship, and a more accurate prediction result can be obtained.

[0073] It should be noted that the mutation of a certain key feature (index) refers to that the value change exceeds the pre-set change threshold. This threshold is used to determine whether the parameter is in an unstable state, thereby issuing a warning. Here, the change usually refers to the sudden rise of the parameter. In addition, the change here usually refers to reaching the change threshold within a certain period of time, not just reaching this change threshold. In order to identify the instability of the roof, it is usually required that the change amplitude of the parameter reaches or exceeds the threshold within a short period of time to be considered as mutation, which can effectively distinguish normal fluctuations and abnormal states.

[0074] Step S400, obtain the current monitoring data, calculate the feature values of a plurality of key feature types according to the current monitoring data, compare the feature values with the change threshold corresponding to the key feature type, obtain the comparison result, obtain the early warning level of roof collapse according to the comparison result, and issue a corresponding early warning signal according to the early warning level.

[0075] Specifically, the current monitoring data includes first target data and second target data; at the beginning and end of a preset time interval, the first target data and the second target data monitored by the monitoring instrument in real time are acquired once respectively; wherein the first target data includes first support working resistance, first roof separation displacement, first anchor rod and anchor cable stress and first hanging roof area, and the second target data includes second support working resistance, second roof separation displacement, second anchor rod and anchor cable stress and second hanging roof area.

[0076] Further, a target support working resistance change rate is calculated according to the first support working resistance and the second support working resistance, a target roof separation displacement increment is calculated according to the first roof separation displacement and the second roof separation displacement, a target anchor rod and anchor cable stress increment is calculated according to the first anchor rod and anchor cable stress and the second anchor rod and anchor cable stress, and a target hanging roof periodic collapse area is calculated according to the first hanging roof area and the second hanging roof area.

[0077] The target support working resistance change rate, the target roof separation displacement increment, the target anchor rod and anchor cable stress increment and the target hanging roof periodic collapse area are compared with the respective corresponding critical change thresholds respectively to obtain comparison results.

[0078] Further, the number of feature values exceeding the corresponding critical change thresholds is used to determine the warning level of the key feature type; if there are two or fewer feature values exceeding the corresponding critical change thresholds in the comparison results, the warning level of the roof collapse is level two, and a preliminary warning signal is issued; if there are three or more feature values exceeding the corresponding critical change thresholds in the comparison results, the warning level of the roof collapse is level one, and an emergency warning signal is issued.

[0079] It can be understood that when it is detected that a single or two parameters of the system enter an unstable state and show a trend of instability, but have not yet reached the critical state of serious instability, the system will issue an early warning signal. This usually means that the monitoring indicators have deviated from the normal fluctuation range, but there may be a certain time and space to take preventive measures. When multiple indicators, i.e. three or all indicator parameters, change rapidly at the same time and the swallow-tail mutation model identifies the critical mutation point of the system, the system issues an emergency warning signal. This means that there is a great possibility of large-area roof collapse in a short time, and the personnel working in the well need to take emergency measures immediately to prevent disasters.

[0080] Further, each key feature (support working resistance change rate, roof separation displacement increment, anchor stress increase, periodic roof collapse area) has a corresponding change threshold, when some indicators exceed this threshold, the system considers that these parameters may enter an unstable state. However, the over-limit of a single parameter does not necessarily mean that the roof will collapse in a large area. Judgment of roof instability often requires comprehensive analysis of multiple monitoring parameters, especially the joint change pattern of these parameters. When the value of a parameter exceeds the threshold, the system will make a judgment in combination with the change of other parameters. For example, when the roof separation displacement increases, whether the change of support resistance is accompanied by aggravation, whether the anchor stress increases at the same time, etc. Multiple indicators appear abnormal at the same time and have a specific pattern, which usually represents a high risk of collapse.

[0081] The mutation recognition of monitoring data and the judgment of roof instability are based on the swallow tail mutation model, which is a tool that can identify the mutation point in a complex system, and is particularly suitable for identifying the dramatic transition process of the system from a stable state to an unstable state. Specifically, the monitoring data is input into the model, and the model can identify whether the system has reached the critical mutation point by analyzing the nonlinear changes of multiple indicators, i.e. from a stable state to an unstable state.

[0082] Therefore, in another embodiment, the present application inputs the real-time obtained support working resistance change rate, roof separation displacement increment, anchor stress increase, and periodic roof collapse area into the swallow tail mutation model. The swallow tail mutation model analyzes the folding and intersection of the curved surface path of these data to determine whether the system is close to the mutation point. The model compares the change rates of different monitoring indicators and the coupling relationship between them to determine whether the system is on the edge of critical change. If the model detects that the nonlinear changes of all key parameters tend to the critical point of the system, the system will determine that the roof may collapse in a large area and issue a corresponding warning signal.

[0083] Further, as shown in Figure 2 Based on the above coal mining face roof collapse early warning method, the present application also correspondingly provides a coal mining face roof collapse early warning system, wherein the coal mining face roof collapse early warning system comprises:

[0084] A data acquisition module 51 is configured to construct a collapse key feature extraction model, obtain historical monitoring data from a monitoring instrument of a coal mine, pre-process the historical monitoring data to obtain pre-processed data, and input the pre-processed data into the collapse key feature extraction model to extract key features of the pre-processed data.

[0085] A key feature extraction module 52 is configured to construct time series data according to the pre-processed data, input the time series data into the collapse key feature extraction model to extract key features, and obtain multiple key feature types of the pre-processed data.

[0086] a change threshold generation module 53, configured to obtain a change threshold of each of the key feature types based on the swallow-tail mutation model and according to the plurality of key feature types;

[0087] a collapse early warning module 54, configured to obtain current monitoring data, calculate feature values of the plurality of key feature types according to the current monitoring data, compare the feature values with the change thresholds corresponding to the key feature types to obtain comparison results, obtain a warning level of roof collapse according to the comparison results, and send a corresponding warning signal according to the warning level.

[0088] Further, as shown in the accompanying drawings, Figure 3 Based on the coal mining face roof collapse early warning method and system, the application further provides a terminal, which comprises a processor 10, a memory 20 and a display 30. Figure 3 Only some components of the terminal are shown, but it should be understood that all the shown components are not required to be implemented, and more or fewer components can be alternatively implemented.

[0089] The memory 20 can be an internal storage unit of the terminal in some embodiments, such as a hard disk or a memory of the terminal. The memory 20 can also be an external storage device of the terminal in other embodiments, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. Further, the memory 20 can include both the internal storage unit and the external storage device of the terminal. The memory 20 is used to store application software and various data installed on the terminal, such as program codes of the terminal, etc. The memory 20 can also be used to temporarily store data that has been output or will be output. In an embodiment, the memory 20 stores a coal mining face roof collapse early warning program 40, which can be executed by the processor 10 to implement the coal mining face roof collapse early warning method of the application.

[0090] The processor 10 can be a central processing unit (CPU), a microprocessor or other data processing chip in some embodiments, which is used to run program codes or process data stored in the memory 20, such as to execute the coal mining face roof collapse early warning method, etc.

[0091] The display 30 can be, in some embodiments, an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch, etc. The display 30 is used to display information at the terminal and to display a visualized user interface. The components of the terminal communicate with each other via a system bus.

[0092] In an embodiment, the following steps are implemented when the processor 10 executes the coal mine working face roof collapse early warning program 40 in the memory 20:

[0093] A collapse key feature extraction model is constructed, historical monitoring data is obtained from a monitoring instrument of a coal mine, and the historical monitoring data is preprocessed to obtain preprocessed data;

[0094] Time series data is constructed according to the preprocessed data, the time series data is input into the collapse key feature extraction model for key feature extraction, and a plurality of key feature types of the preprocessed data are obtained;

[0095] Based on a swallow-tail mutation model, a change threshold of each key feature type is obtained according to a plurality of key feature types;

[0096] Current monitoring data is obtained, feature values of a plurality of key feature types are calculated according to the current monitoring data, the feature values are compared with change thresholds corresponding to the key feature types to obtain comparison results, an early warning level of roof collapse is obtained according to the comparison results, and a corresponding early warning signal is sent according to the early warning level.

[0097] The monitoring instrument includes a support resistance sensor, a roof separation instrument, an anchor rod and anchor cable stress sensor, and a laser range finder.

[0098] The support resistance sensor is arranged on a support to monitor changes in support working resistance, the roof separation instrument is arranged in a distribution area of a roof rock layer and a roadway roof to monitor changes in separation displacement of the roof rock layer, the anchor rod and anchor cable stress sensor is arranged on an anchor rod and an anchor cable to monitor changes in anchor rod and anchor cable stress in real time, and the laser range finder is arranged on a guide rail where a return air corner support tail beam meets a coal wall to dynamically monitor a suspended roof area to capture expansion of the suspended roof area.

[0099] The dynamic monitoring of the suspended roof area to capture the expansion of the suspended roof area specifically includes:

[0100] Three groups of measurement data of the laser range finder at three positions of a return air corner close to a coal wall side, a coal wall and a support middle, and a return air corner close to a support side are respectively obtained by a preset measurement method.

[0101] Match the three groups of measurement data with the theoretical data, if the measurement data does not match the theoretical data, it means that the hanging roof does not exist, remove the measurement data, if the measurement data matches the theoretical data, the measurement data matched with the theoretical data is taken as the target measurement data, and the hanging roof area is calculated according to the target measurement data;

[0102] The preset measurement method comprises the following steps: vertically arranging the laser range finder relative to the bottom plate to obtain a first vertical distance from the bottom plate to the top plate measured by the laser range finder; deflecting the laser range finder by 30 degrees and 60 degrees respectively in the direction of the goaf to obtain a second vertical distance and a third vertical distance from the bottom plate to the top plate measured by the laser range finder; horizontally arranging the laser range finder relative to the bottom plate to obtain a first horizontal distance from the bottom plate to the top plate measured by the laser range finder; and deflecting the laser range finder by 30 degrees and 60 degrees respectively in the direction of the support to obtain a second horizontal distance and a third horizontal distance from the bottom plate to the top plate measured by the laser range finder.

[0103] The pre-processing of the historical monitoring data comprises the following steps:

[0104] The historical monitoring data is subjected to outlier rejection to obtain first intermediate data, and the missing data in the first intermediate data is subjected to mean filling to obtain second intermediate data.

[0105] The second intermediate data is subjected to noise reduction processing by a median filtering method or a Gaussian filtering method to obtain third intermediate data, and the third intermediate data is subjected to normalization processing to obtain pre-processed data scaled to a specified range.

[0106] The key feature extraction model is a support vector machine model or a random forest model.

[0107] The time series data is constructed according to the pre-processed data, the time series data is input into the key feature extraction model to extract key features, and a plurality of key feature types of the pre-processed data are obtained.

[0108] Key data with records of sudden roof collapse events of coal mines is selected from the pre-processed data, and time series data is constructed according to the key data.

[0109] The time series data is input into the support vector machine model or the random forest model to extract key features, and the support working resistance change rate of the pre-processed data, the roof separation displacement increment, the anchor rod and anchor cable stress increment, and the periodic collapse area of the hanging roof are obtained.

[0110] The change threshold of each key feature is obtained based on the swallow-tail mutation model according to the plurality of key features, and specifically includes:

[0111] The plurality of key features are input into the swallow-tail mutation model for change threshold setting to obtain the change threshold of each key feature.

[0112] The change threshold of each key feature includes a normal change threshold, a dangerous change threshold and a critical change threshold.

[0113] The current monitoring data includes first target data and second target data.

[0114] The current monitoring data is obtained, the feature values of the plurality of key feature types are calculated according to the current monitoring data, the feature values are compared with the change thresholds corresponding to the key feature types to obtain comparison results, and specifically includes:

[0115] The first target data and the second target data monitored by the monitoring instrument in real time are obtained once at the beginning and the end of the preset time interval, respectively.

[0116] The first target data includes a first support working resistance, a first roof separation displacement, a first anchor rod and anchor cable stress and a first hanging roof area, and the second target data includes a second support working resistance, a second roof separation displacement, a second anchor rod and anchor cable stress and a second hanging roof area.

[0117] The target support working resistance change rate is calculated according to the first support working resistance and the second support working resistance, the target roof separation displacement increment is calculated according to the first roof separation displacement and the second roof separation displacement, the target anchor rod and anchor cable stress increment is calculated according to the first anchor rod and anchor cable stress and the second anchor rod and anchor cable stress, and the target hanging roof periodic collapse area is calculated according to the first hanging roof area and the second hanging roof area.

[0118] The target support working resistance change rate, the target roof separation displacement increment, the target anchor rod and anchor cable stress increment and the target hanging roof periodic collapse area are compared with the respective corresponding critical change thresholds to obtain comparison results.

[0119] The comparison results are obtained, the warning level of the roof collapse is obtained according to the comparison results, and a corresponding warning signal is sent according to the warning level, and specifically includes:

[0120] The number of feature values of the key feature type exceeding the corresponding critical change threshold is determined to determine the warning level.

[0121] If there are two or less characteristic values exceeding the corresponding critical change threshold in the comparison result, the early warning level of the roof collapse is level two, and a preliminary early warning signal is sent out;

[0122] If there are three or more characteristic values exceeding the corresponding critical change threshold in the comparison result, the early warning level of the roof collapse is level one, and an emergency early warning signal is sent out.

[0123] The application further provides a computer readable storage medium, wherein the computer readable storage medium stores a coal mine working face roof collapse early warning program, and the coal mine working face roof collapse early warning program realizes the steps of the coal mine working face roof collapse early warning method when executed by a processor.

[0124] To sum up, the application provides a coal mine working face roof collapse early warning method and system, the method comprising: constructing a collapse key feature extraction model, obtaining historical monitoring data, preprocessing the historical monitoring data, constructing time series data according to the preprocessed data, inputting the time series data into the collapse key feature extraction model for key feature extraction to obtain the key feature types of the preprocessed data; obtaining the change threshold of the key feature types according to the key feature types based on the swallow-tail mutation model; obtaining current monitoring data, calculating the feature values of the key feature types according to the current monitoring data, comparing the feature values with the change threshold corresponding to the key feature types to obtain a comparison result, obtaining the early warning level of the roof collapse according to the comparison result, and sending out a corresponding early warning signal according to the early warning level. The application comprehensively considers the key indicators of support working resistance, roof separation displacement, anchor rod and anchor cable stress, and suspended roof area, combines machine learning and mutation theory, can accurately identify the precursor features of large-area sudden roof collapse, thereby sending out an early warning signal in advance, improves the accuracy of coal mine working face roof collapse early warning, and realizes the early warning of large-area sudden roof collapse.

[0125] It should be noted that in this document, the terms "comprising", "including", or any other variant thereof are intended to cover non-exclusive inclusions, so that processes, methods, articles, or terminals including a series of elements not only include those elements, but also include other elements not explicitly listed, or include elements inherent to such processes, methods, articles, or terminals. Without more limitations, the element defined by the statement "including a" does not exclude the presence of another identical element in the process, method, article, or terminal including the element.

[0126] Of course, those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when the computer program is executed, the computer program can include the processes of the above-mentioned embodiments of the method. Any reference to memory, storage, database or other medium used in the embodiments provided by the present application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration but not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0127] It should be understood that the application is not limited to the above examples, and those skilled in the art can make improvements or changes according to the above description, and all these improvements and changes shall fall within the protection scope of the appended claims of the present application.

Claims

1. A coal mine working face roof caving early warning method, characterized in that, The coal mine working face roof collapse early warning method comprises: A collapse key feature extraction model is constructed, historical monitoring data is obtained from a monitoring instrument of a coal mine, and the historical monitoring data is preprocessed to obtain preprocessed data; According to the preprocessed data, time series data is constructed, the time series data is input into the collapse key feature extraction model for key feature extraction, and a plurality of key feature types of the preprocessed data are obtained; Based on a swallow tail mutation model, a change threshold of each key feature type is obtained according to a plurality of key feature types; Current monitoring data is obtained, feature values of a plurality of key feature types are calculated according to the current monitoring data, the feature values are compared with change thresholds corresponding to the key feature types, a comparison result is obtained, a roof collapse early warning level is obtained according to the comparison result, and a corresponding early warning signal is sent according to the early warning level; The collapse key feature extraction model is a support vector machine model or a random forest model; The preprocessed data is constructed into time series data, the time series data is input into the collapse key feature extraction model for key feature extraction, and a plurality of key feature types of the preprocessed data are obtained, specifically comprising: Key data with a record of a sudden coal mine roof collapse event is selected from the preprocessed data, and time series data is constructed according to the key data; The time series data is input into the support vector machine model or the random forest model for key feature extraction, and support bracket working resistance change rate, roof separation displacement increment, anchor rod and anchor cable stress increase, and periodic collapse area of a suspended roof are obtained from the preprocessed data; Based on the swallow tail mutation model, a change threshold of each key feature is obtained according to a plurality of key features, specifically comprising: A plurality of key features are input into the swallow tail mutation model for change threshold setting, and a change threshold of each key feature is obtained; The change threshold of each key feature comprises a normal change threshold, a dangerous change threshold and a critical change threshold.

2. The coal face roof collapse warning method of claim 1, wherein, The monitoring instrument comprises a support resistance sensor, a roof separation instrument, an anchor rod and anchor cable stress sensor and a laser range finder; The support resistance sensor is arranged on the support for monitoring the support working resistance change, the roof separation instrument is arranged in the roof rock and the roadway top distribution area for monitoring the separation displacement change of the roof rock, the anchor rod and anchor cable stress sensor is arranged on the anchor rod and anchor cable for real-time monitoring of the anchor rod and anchor cable stress change, and the laser range finder is arranged on the guide rail where the return air corner support tail beam is connected with the coal wall for dynamic monitoring of the suspended roof area to capture the expansion of the suspended roof area.

3. The coal face roof collapse warning method of claim 2, wherein, The dynamic monitoring of the suspended roof area to capture the expansion of the suspended roof area specifically comprises: Three groups of measurement data of the laser range finder at three positions of the return air corner close to the coal wall side, the coal wall and the middle of the support and the return air corner close to the support side are obtained respectively through a preset measurement method; Match the three groups of measurement data with the theoretical data. If the measurement data does not match the theoretical data, it indicates that the hanging roof does not exist, and the measurement data is removed. If the measurement data matches the theoretical data, the measurement data that matches the theoretical data is taken as target measurement data, and the target measurement data is used to calculate the area of the hanging roof; The preset measurement method comprises the following steps: vertically arranging the laser range finder relative to the floor to obtain a first vertical distance from the floor to the roof measured by the laser range finder; deflecting the laser range finder by 30 degrees and 60 degrees, respectively, towards the goaf to obtain a second vertical distance and a third vertical distance from the floor to the roof measured by the laser range finder; horizontally arranging the laser range finder relative to the floor to obtain a first horizontal distance from the floor to the roof measured by the laser range finder; and deflecting the laser range finder by 30 degrees and 60 degrees, respectively, towards the support to obtain a second horizontal distance and a third horizontal distance from the floor to the roof measured by the laser range finder.

4. The coal face roof collapse warning method of claim 1, wherein, The pre-processing of the historical monitoring data comprises the following steps: The historical monitoring data is subjected to outlier rejection to obtain first intermediate data, and the missing data in the first intermediate data is subjected to mean filling to obtain second intermediate data. The second intermediate data is subjected to noise reduction processing by a median filtering method or a Gaussian filtering method to obtain third intermediate data, and the third intermediate data is subjected to normalization processing to obtain pre-processed data scaled to a specified range.

5. The coal face roof fall early warning method of claim 1, wherein, The current monitoring data comprises first target data and second target data. The current monitoring data is obtained, and the feature values of the plurality of key feature types are calculated based on the current monitoring data. The feature values are compared with the change threshold values corresponding to the key feature types to obtain comparison results. The comparison results are obtained by comparing the feature values with the change threshold values corresponding to the key feature types. The first target data and the second target data are obtained at the beginning and the end of the preset time interval. The first target data comprises a first support working resistance, a first roof separation displacement, a first anchor rod and anchor cable stress, and a first hanging roof area. The second target data comprises a second support working resistance, a second roof separation displacement, a second anchor rod and anchor cable stress, and a second hanging roof area. A target support working resistance change rate is calculated based on the first support working resistance and the second support working resistance. A target roof separation displacement increment is calculated based on the first roof separation displacement and the second roof separation displacement. A target anchor rod and anchor cable stress increase is calculated based on the first anchor rod and anchor cable stress and the second anchor rod and anchor cable stress. A target hanging roof periodic collapse area is calculated based on the first hanging roof area and the second hanging roof area. The target support working resistance change rate, the target roof separation displacement increment, the target anchor rod and anchor cable stress increase, and the target hanging roof periodic collapse area are compared with the respective critical change threshold values to obtain comparison results.

6. The coal face roof fall early warning method of claim 5, wherein, The comparison results are used to obtain a roof collapse warning level, and a corresponding warning signal is sent based on the warning level. The comparison results are used to obtain a roof collapse warning level, and a corresponding warning signal is sent based on the warning level. The number of feature values of the key feature types that exceed the corresponding critical change threshold values is used to determine the warning level. If there are two or less characteristic values exceeding the corresponding critical change threshold in the comparison result, the early warning level of the roof collapse is level two, and a preliminary early warning signal is sent out; If there are three or more characteristic values exceeding the corresponding critical change threshold in the comparison result, the early warning level of the roof collapse is level one, and an urgent early warning signal is sent out.

7. A coal mine working face roof collapse early warning system, characterized in that, The coal mine working face roof collapse early warning system is applied to the coal mine working face roof collapse early warning method of any one of claims 1-6, and comprises: A data acquisition module configured to construct a collapse key feature extraction model, acquire historical monitoring data from a monitoring instrument of a coal mine, pre-process the historical monitoring data, and obtain pre-processed data; A key feature extraction module configured to construct time series data based on the pre-processed data, input the time series data into the collapse key feature extraction model for key feature extraction, and obtain a plurality of key feature types of the pre-processed data; A change threshold generation module configured to obtain a change threshold of each key feature type based on a swallow tail mutation model and the plurality of key feature types; A collapse early warning module configured to acquire current monitoring data, calculate characteristic values of the plurality of key feature types based on the current monitoring data, compare the characteristic values with the change threshold corresponding to the key feature types, obtain a comparison result, obtain an early warning level of the roof collapse based on the comparison result, and send out a corresponding early warning signal based on the early warning level.

8. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a coal mine working face roof collapse early warning program, and the coal mine working face roof collapse early warning program, when executed by the processor, implements the steps of the coal mine working face roof collapse early warning method of any one of claims 1-6.

Images