Mining water disaster prediction method and system

Abstract

The invention provides a mining water disaster prediction method and device. The method comprises the following steps: collecting a monitoring parameter set of different monitoring objects in a monitoring area; aimed at the monitoring parameter of each dimension, auxiliary information of the monitoring parameter is acquired, and the auxiliary information comprises a change rate sequence and an average value sequence; inputting the multi-dimensional monitoring parameters and the corresponding auxiliary information into a pre-trained target prediction model, and performing water inflow prediction by the target prediction model according to the multi-dimensional monitoring parameters and the corresponding auxiliary information to obtain a first predicted water inflow; and according to the first predicted water inflow, determining a target water disaster level of the monitoring area, indicating that a water inflow event occurs in the monitoring area in response to the target water disaster level, and sending out alarm information. Water disaster prediction is carried out by adopting an artificial intelligence multi-sensor data fusion mode, deep features contained in monitoring data are also considered, the complexity of coal mine hydrogeological conditions can be better described, and an early warning result is more accurate and stable.

Description

Technical Field

[0001] The present application relates to the field of coal mine safety technology, and in particular to a method and system for predicting water hazards in a mine. Background Art

[0002] Coal mine water damage may cause safety accidents for both production mines and infrastructure mines, which is extremely harmful to the safe production and construction of mines. How to predict the water damage in coal mines has become a hot topic in research. Summary of the invention

[0003] The purpose of this application is to solve one of the technical problems in the related art at least to some extent.

[0004] To this end, the first purpose of this application is to propose a method for predicting water hazards in coal mines to predict water hazards in coal mines in order to avoid coal mine accidents and improve underground operation safety.

[0005] The second objective of this application is to provide a mining water hazard prediction system.

[0006] The third objective of the present application is to provide an electronic device.

[0007] A fourth objective of the present application is to provide a computer-readable storage medium.

[0008] A fifth object of the present application is to provide a computer program product.

[0009] To achieve the above-mentioned purpose, the first embodiment of the present application proposes a method for predicting water damage in a mine, comprising:

[0010] Collecting a set of monitoring parameters of different monitoring objects in the monitoring area, wherein the set of monitoring parameters includes multi-dimensional monitoring parameters related to hydrological monitoring, and the monitoring objects include ground monitoring objects and underground monitoring objects;

[0011] For each dimension of the monitoring parameter, obtaining auxiliary information of the monitoring parameter, wherein the auxiliary information includes a change rate sequence and an average value sequence;

[0012] Inputting the multi-dimensional monitoring parameters and the corresponding auxiliary information into a pre-trained target prediction model, and the target prediction model predicts the water inflow according to the multi-dimensional monitoring parameters and the corresponding auxiliary information to obtain a first predicted water inflow;

[0013] A target water hazard level of the monitoring area is determined based on the first predicted water inflow, and an alarm message is issued in response to the target water hazard level indicating that a water inflow event has occurred in the monitoring area.

[0014] To achieve the above-mentioned purpose, the second embodiment of the present application proposes a mining water hazard prediction system, comprising:

[0015] A downhole sensing unit, used to collect a set of monitoring parameters of downhole monitoring objects in a monitoring area, wherein the set of monitoring parameters includes multi-dimensional monitoring parameters related to hydrological monitoring;

[0016] A ground monitoring unit, used to collect a set of monitoring parameters of ground monitoring objects in the monitoring area;

[0017] A ground server is used to obtain auxiliary information of the monitoring parameters in each dimension, wherein the auxiliary information includes a change rate sequence and an average value sequence, and input the multi-dimensional monitoring parameters and the corresponding auxiliary information into a pre-trained target prediction model, and the target prediction model predicts the water inflow according to the multi-dimensional monitoring parameters and the corresponding auxiliary information to obtain a first predicted water inflow, and determines a target water hazard level of the monitoring area based on the first predicted water inflow, and issues an alarm message in response to the target water hazard level indicating a water inflow event in the monitoring area.

[0018] To achieve the above-mentioned purpose, the third aspect embodiment of the present application proposes an electronic device, comprising: a processor; and a memory communicatively connected to the processor; the memory stores computer-executable instructions; the processor executes the computer-executable instructions stored in the memory, so that the processor can execute the mining water hazard prediction method described in the first aspect embodiment above.

[0019] To achieve the above-mentioned purpose, the fourth aspect of the present application proposes a computer-readable storage medium having a computer program stored thereon, wherein the computer instructions are used to enable the computer to execute the mining water hazard prediction method described in the first aspect of the above-mentioned embodiment.

[0020] To achieve the above-mentioned purpose, the fifth embodiment of the present application proposes a computer program product, including a computer program, which, when executed by a processor, implements the mining water hazard prediction method described in the first embodiment above.

[0021] The mine water hazard prediction method and system provided in this application adopts an artificial intelligence multi-sensor data fusion method to predict water hazards. It not only relies on multiple sensors to collect raw data, but also calculates information such as change rate and multi-channel mean, and explores the deep characteristics contained in the coal mine water hazard monitoring data, so as to better describe the complexity of the hydrogeological conditions of the coal mine. Compared with the artificially set empirical threshold alarm and mathematical statistics method, the early warning result is more accurate and stable.

[0022] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0024] Figure 1 A schematic diagram of a flow chart of a method for predicting water hazards in a mine provided in an embodiment of the present application;

[0025] Figure 2 A schematic diagram of the structure of a prediction model provided in an embodiment of the present application;

[0026] Figure 3 A schematic diagram of a flow chart of another method for predicting water hazards in a mine provided in an embodiment of the present application;

[0027] Figure 4 A schematic diagram of the structure of a mine water hazard prediction system provided in an embodiment of the present application;

[0028] Figure 5 A schematic diagram of the structure of another mining water hazard prediction system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0029] Embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0030] The following describes the mining water hazard prediction method and system according to the embodiments of the present application with reference to the accompanying drawings.

[0031] Figure 1 A schematic diagram of a method for predicting water damage in a mine provided in an embodiment of the present application. Figure 1 As shown, the mine water hazard prediction method may include but is not limited to the following steps:

[0032] S101, collecting a set of monitoring parameters of different monitoring objects in a monitoring area, where the set of monitoring parameters includes multi-dimensional monitoring parameters related to hydrological monitoring.

[0033] In some embodiments, the monitoring area of ​​the coal mine may include multiple different monitoring objects. Optionally, the different monitoring objects may include but are not limited to ground monitoring objects and underground monitoring objects.

[0034] In some embodiments, different monitoring objects may include but are not limited to the ground surface, tunnels, working faces, and boreholes.

[0035] In some embodiments, multiple types of sensors are deployed at each monitoring object. For example, sensors can be deployed in a borehole to monitor hydrological parameters.

[0036] In some embodiments, multiple types of sensors include but are not limited to water level sensors, water pressure sensors, hydrological sensors, water quality sensors, meteorological sensors, flow sensors, mined-out sealed wall displacement sensors, and mined-out sealed wall stress sensors.

[0037] In some embodiments, the same type of sensors may be arranged at multiple monitoring points of each monitored object. For example, N monitoring points may be set in the monitored object, and M type sensors may be arranged at the N monitoring points. Optionally, M type sensors may be arranged at each monitoring point. Optionally, M type sensors may be arranged at N monitoring points respectively, that is, there may be some differences in the types of sensors arranged between the monitoring points.

[0038] For example, the M types of sensors include type 1 sensors, type 2 sensors, type 3 sensors, type 4 sensors, and type 5 sensors; the N monitoring points may include monitoring point 1, monitoring point 2, and monitoring point 3. For example, each monitoring point is deployed with type 1 sensors, type 2 sensors, type 3 sensors, type 4 sensors, and type 5 sensors. For another example, monitoring point 1 is deployed with type 1 sensors and type 2 sensors; monitoring point 2 is deployed with type 1 sensors and type 3 sensors. Monitoring point 3 is deployed with type 4 sensors and type 5 sensors.

[0039] In some embodiments, multiple sensors of each type may be deployed on the same monitoring object. For example, sensors of the same type may be deployed at different monitoring points.

[0040] In some embodiments, the monitoring parameter set may include multi-dimensional monitoring parameters such as water level, water pressure, temperature, water quality, meteorological data, displacement of the goaf sealed wall, and stress of the goaf sealed wall monitored at the monitoring object.

[0041] In some embodiments, the monitoring parameters of multiple types of sensors can be collected by a data collector and transmitted to a ground service station through a communication network for storage and management, so as to facilitate data backtracking and statistical analysis. For example, it can be transmitted to a ground server through a downhole ring network. Optionally, a substation is deployed underground, and the substation can obtain the set of monitoring parameters collected by the data collector. Further, the substation is connected to the ground server through the downhole ring network, and then reports the monitoring parameters underground.

[0042] In some embodiments, the ground server can not only receive monitoring data reported by the underground substation, for example, including but not limited to: underground flow, underground water temperature, water pressure, goaf closed wall transfer, goaf closed wall stress, etc. The ground server can also receive monitoring parameters reported by the ground data collector, for example, including but not limited to: meteorological data, ground water temperature, water pressure, water quality and other parameters.

[0043] S102: For each dimension of monitoring parameters, obtain auxiliary information of the monitoring parameters.

[0044] In some embodiments, auxiliary information of the monitoring parameters in each dimension can be obtained, and the auxiliary information can be used to reflect changes in the monitoring parameters. Comprehensive consideration of the auxiliary information when predicting water disasters can better predict water disasters and improve prediction accuracy.

[0045] In some embodiments, the auxiliary information of the monitoring parameters may include, but is not limited to: a change rate sequence and an average value sequence of the monitoring parameters.

[0046] In some embodiments, the process of acquiring the change rate sequence of the monitoring parameter includes:

[0047] For sampling time t i Monitoring parameters and sampling time t i-1 The difference of the monitoring parameters is calculated, and the calculated difference and the sampling time t are used to calculate the difference. i The monitoring parameters are used to determine the sampling time t i The corresponding change rate is obtained according to the change rate corresponding to all or part of the sampling moments to obtain the change rate sequence of the monitoring parameter, where i is an integer greater than or equal to 1.

[0048] Example description, x i is the sampling time t i The monitoring parameters collected by the sensor, sampling time t i The rate of change of the sensor's monitoring parameter △x i , can be calculated according to the following formula:

[0049] △x i =(x i -x i-1 ) / x i ;

[0050] It can be understood that the change sequence of the change rate sequence of the monitoring parameters is {△x1, △x2, ..., △x i ,……}.

[0051] In some embodiments, the process of determining the variable mean value sequence of the monitoring parameter includes:

[0052] For sampling time ti , sliding time window to determine the sampling time t within the time window i Corresponding to multiple monitoring parameters, the multiple monitoring parameters in the time window are averaged to obtain the sampling time t i The corresponding average value, further, according to the average values ​​corresponding to all or part of the sampling moments, a variable average value sequence of the monitoring parameter is obtained.

[0053] Example description, x i is the sampling time t i The monitoring parameters collected by the sensor, sampling time t i The average value of the monitoring parameter of the sensor ε i , can be calculated according to the following formula:

[0054] ε i =(x i-k +···x i +···x i+k ) / (2k+1);

[0055] Among them, k is the window length of the time window, which can be based on the sampling time t i Set the monitoring parameters of the time length before and after, and calculate t i It can be understood that the variation sequence of the average value sequence of the monitoring parameters is {ε1, ε2, ..., ε i ,……}.

[0056] S103, inputting the multi-dimensional monitoring parameters and the corresponding auxiliary information into a pre-trained target prediction model, and the target prediction model predicts the water inflow according to the multi-dimensional monitoring parameters and the corresponding auxiliary information to obtain a first predicted water inflow.

[0057] In some embodiments, the training process of the target prediction model may include the following process:

[0058] In some embodiments, training samples and labeled water inflows of the training samples are collected, and further, the training samples are input into the prediction model to be trained, and the prediction model predicts the water inflow according to the training samples to obtain a second predicted water inflow. After obtaining the second predicted water inflow of the training samples, the model parameters of the prediction model can be adjusted according to the labeled water inflow and the second predicted water inflow, and the adjusted water inflow model can be continuously trained until the training is completed to obtain the target prediction model.

[0059] In some embodiments, collecting training samples may include multi-dimensional sample monitoring parameters.

[0060] In some embodiments, the model loss of the prediction model can be determined based on the labeled water inflow and the second predicted water inflow, and further, the model parameters of the prediction model can be reversely adjusted and optimized based on the model loss.

[0061] In some embodiments, the conditions for the end of training may include but are not limited to: the training accuracy of the model reaches a set accuracy value, the model converges, the training error of the model is less than or equal to a set error value, the number of iterative training of the model reaches a set number of times, the training time of the model reaches a set time, etc.

[0062] In some embodiments, the prediction model is structured as Figure 2 As shown, the prediction model includes an input layer X, a time series calculation processing layer L, a neural network layer A, and an output layer Y. The training sample is input into the input layer X, which performs vector representation and inputs into the time series calculation processing layer L to obtain auxiliary information. The vector representation and the auxiliary information are fused and input into the neural network layer A, which extracts features of the input information and perceives the association relationship to obtain multi-dimensional feature information. Furthermore, the multi-dimensional feature information is input into the output layer Y for prediction. Optionally, the output layer Y may include multiple prediction heads, and different prediction heads may correspond to different prediction information, for example, it may include the predicted water inflow in the monitoring area, the predicted hydrological type in the monitoring area, the predicted water hazard level in the monitoring area, etc.

[0063] Optionally, feed the training samples into Figure 2 In the process of training the prediction model shown, the output value of each node can be calculated forward (from left to right), and then the error of each node can be calculated reversely (from right to left), and the weights can be updated iteratively until the model verification accuracy meets the requirements to obtain the target prediction model. Optionally, after completing the multi-sensor data fusion calculation process, the weight parameters of the target prediction model can be saved to facilitate prediction of the newly collected monitoring data set based on the weight parameters.

[0064] S104, determining a target water hazard level for the monitoring area based on the first predicted water inflow, and issuing an alarm message in response to the target water hazard level indicating that a water inflow event has occurred in the monitoring area.

[0065] In some embodiments, the water damage level can be divided into multiple danger levels. Optionally, the water damage level is divided into three danger levels: low, medium and high.

[0066] In some embodiments, a safe water inflow of the monitoring area is obtained, and a deviation value between the first predicted water inflow and the safe water inflow is determined. Further, a target water hazard level is determined based on the deviation value.

[0067] In some embodiments, the regional location and commissioning start time of the monitoring area are determined, and the historical water inflow of the monitoring area is determined based on the regional location and commissioning start time. The safe water inflow of the monitoring area is determined based on the historical water inflow of the monitoring area.

[0068] In some embodiments, the water inflow of the entire mine in the past three years is counted. If the mine has been in production for less than three years, the statistics are based on the time of production. The water inflow when water inrush occurs is excluded, and the average of the remaining statistical values ​​is used as the safe water inflow of the mine.

[0069] In some embodiments, a mapping relationship between different intervals and water damage levels can be pre-established, and different intervals correspond to different water damage levels. Optionally, a ratio between the first predicted water inflow and the safe water inflow is obtained, and further, a target interval range in which the ratio is located is determined, and the water damage level associated with the target interval range is used as the target water damage level.

[0070] By way of example, when the ratio of the first predicted water inflow to the normal safe water inflow is less than or equal to 1.0, the target water hazard level is determined to be low-level water hazard; when the ratio of the first predicted water inflow to the normal safe water inflow is greater than 1.0 and less than or equal to 1.2 times, the target water hazard level is determined to be medium-level water hazard; and when the ratio of the first predicted water inflow to the normal safe water inflow is greater than 1.2 times and less than or equal to 1.5 times, the target water hazard level is determined to be high-risk water hazard.

[0071] In some embodiments, a mapping relationship between different difference ranges and water hazard levels can be pre-established, and different difference ranges correspond to different water hazard levels. The difference between the first predicted water inflow and the safe water inflow is obtained, and further, a target difference range in which the difference lies is determined, and the water hazard level associated with the target difference range is used as the target water hazard level.

[0072] In some embodiments, in response to the target water hazard level indicating that a water gushing event has occurred in the monitoring area, an alarm message is issued. Optionally, when the target water hazard level exceeds the set water hazard level, it can be determined that a water gushing event has occurred in the monitoring area, and alarm messages of different levels can be issued according to the target water hazard level. For example, a moderate water hazard can be notified to relevant management personnel through various means such as text messages and emails. For another example, a high-risk water hazard can be notified to management personnel of different levels through various means such as text messages, emails, and telephone calls, and the ground server can broadcast the alarm message to ensure that relevant personnel receive the alarm message, so that they can take timely remedial measures to reduce the damage to personnel and facilities when a water gushing event occurs, so as to ensure the safety of personnel and facilities.

[0073] In some embodiments, when the target water hazard level is determined to be a low-risk water hazard based on the first predicted utilization amount and the safe water inflow of the monitoring area, no alarm information may be issued.

[0074] In the embodiment of the present application, an artificial intelligence multi-sensor data fusion method is used to predict water hazards. It not only relies on multiple sensors to collect raw data, but also calculates information such as the rate of change and multi-channel mean, and explores the deep characteristics contained in the coal mine water hazard warning data, so as to better describe the complexity of the hydrogeological conditions of the coal mine. Compared with the artificially set empirical threshold alarm and mathematical statistics method, the warning result is more accurate and stable.

[0075] Figure 3 A schematic diagram of another method for predicting flood damage provided in an embodiment of the present application. Figure 3 As shown, the water damage prediction method may include but is not limited to the following steps:

[0076] S301, collecting training samples and labeled water inflow of the training samples, and training a prediction model based on the training samples and labeled water inflow to obtain a target prediction model.

[0077] S302: Collect a set of monitoring parameters of different monitoring objects in the monitoring area.

[0078] S303, obtaining auxiliary information of the monitoring parameters of each dimension, wherein the auxiliary information includes a change rate sequence and an average value sequence.

[0079] In some embodiments, after obtaining the monitoring parameter sets of different monitoring objects, the multi-dimensional monitoring data in the monitoring parameter sets can be preprocessed, classified and sorted by scores, abnormal data are kicked out, normalized, etc. Further, based on the preprocessed monitoring parameters of each dimension, auxiliary information of the monitoring parameters is calculated.

[0080] S304, inputting the multi-dimensional monitoring parameters and the corresponding auxiliary information into the target prediction model, and the target prediction model predicts the water inflow according to the multi-dimensional monitoring parameters and the corresponding auxiliary information to obtain a first predicted water inflow.

[0081] S305: Determine a target flood hazard level in the monitoring area according to the first predicted water inflow.

[0082] S306: In response to the target water hazard level indicating that a water inrush event occurs in the monitoring area, an alarm message is issued.

[0083] S307, in response to the target water hazard level indicating that no water inrush event occurs in the monitoring area, continue monitoring.

[0084] S308, collecting a set of monitoring parameters corresponding to a water inrush event in the monitoring area as a model optimization sample.

[0085] S309, optimizing and fine-tuning the target prediction model based on the model optimization samples.

[0086] In some embodiments, the set of monitoring parameters corresponding to a water gushing event in the monitoring area can be used as a model optimization sample. After the set conditions are met, the target prediction model can be optimized and fine-tuned based on the model optimization sample to improve the accuracy of the prediction.

[0087] In some embodiments, after the number of model optimization samples reaches a set number of samples, the target prediction model can be optimized and fine-tuned based on the model optimization samples.

[0088] In some embodiments, after the set optimization and fine-tuning period is reached at the current moment, the target prediction model can be optimized and fine-tuned based on the model optimization samples.

[0089] In some embodiments, after receiving the set optimization and fine-tuning instruction at the current moment, the target prediction model can be optimized and fine-tuned based on the model optimization sample.

[0090] In the embodiment of the present application, artificial intelligence multi-sensor data fusion is used to predict water hazards. It not only relies on multiple sensors to collect raw data, but also calculates information such as the rate of change and multi-channel mean, and explores the deep characteristics contained in the coal mine water hazard warning data, so as to better describe the complexity of the hydrogeological conditions of the coal mine. Compared with the artificially set empirical threshold alarm and mathematical statistics method, the warning result is more accurate and stable. Furthermore, the target prediction model can be continuously optimized and fine-tuned based on the model optimization sample, so as to continuously improve the prediction accuracy.

[0091] Figure 4 This is a schematic diagram of the structure of a water disaster prediction system provided in an embodiment of the present application. Figure 4 The water hazard prediction system shown may include, but is not limited to: a downhole sensing unit 10 , a ground monitoring unit 20 and a ground server 30 .

[0092] A downhole sensing unit 10, used to collect a set of monitoring parameters of downhole monitoring objects in a monitoring area, wherein the set of monitoring parameters includes multi-dimensional monitoring parameters related to hydrological monitoring;

[0093] A ground monitoring unit 20, used to collect a set of monitoring parameters of ground monitoring objects in a monitoring area;

[0094] The ground server 30 is used to obtain auxiliary information of the monitoring parameters in each dimension, wherein the auxiliary information includes a change rate sequence and an average value sequence, and input the multi-dimensional monitoring parameters and the corresponding auxiliary information into a pre-trained target prediction model, and the target prediction model predicts the water inflow according to the multi-dimensional monitoring parameters and the corresponding auxiliary information to obtain a first predicted water inflow, and determines the target water hazard level of the monitoring area according to the first predicted water inflow, and issues an alarm message in response to the target water hazard level indicating that a water inflow event has occurred in the monitoring area.

[0095] In some embodiments, the ground server 30 is further configured to:

[0096] Collecting training samples and labeled water inflow of the training samples;

[0097] Inputting the training samples into a prediction model to be trained, and having the prediction model predict water inflow according to the training samples to obtain a second predicted water inflow;

[0098] According to the labeled water inflow and the second predicted water inflow, the model parameters of the prediction model are adjusted, and the adjusted water inflow model is continuously trained until the training is completed to obtain the target prediction model.

[0099] In some embodiments, the ground server 30 is further configured to:

[0100] For sampling time t i Monitoring parameters and sampling time t i-1 The difference calculation is performed on the monitoring parameters, and the difference and the sampling time t are obtained. i The ratio of the monitoring parameters to determine the sampling time t i The corresponding rate of change, where i is an integer greater than or equal to 1;

[0101] The change rate sequence is obtained according to the change rate corresponding to the sampling time.

[0102] In some embodiments, the ground server 30 is further configured to:

[0103] For sampling time t i , sliding the time window to determine the sampling time t within the time window i Corresponding to multiple monitoring parameters, the multiple monitoring parameters in the time window are averaged to obtain the sampling time t i The corresponding average value;

[0104] The variable average value sequence is obtained according to the average values ​​corresponding to the sampling moments.

[0105] In some embodiments, the ground server 30 is further configured to:

[0106] Obtaining a safe water inflow in the monitoring area, and determining a deviation value between the first predicted water inflow and the safe water inflow;

[0107] The target water damage level is determined according to the deviation value.

[0108] In some embodiments, the ground server 30 is further configured to:

[0109] Obtaining a ratio between the first predicted water inflow and the safe water inflow;

[0110] Determining the target water damage level according to the deviation value includes:

[0111] Determining a target interval range of the ratio, wherein different ranges correspond to different water damage levels;

[0112] A water hazard level associated with the target interval range is determined as the target water hazard level.

[0113] In some embodiments, the ground server 30 is further configured to:

[0114] Obtaining a difference between the first predicted water inflow and the safe water inflow;

[0115] Determining the target water damage level according to the deviation value includes:

[0116] Determining a target difference range within which the difference lies, wherein different difference ranges correspond to different water damage levels;

[0117] The water damage level associated with the target difference range is used as the target water damage level.

[0118] In some embodiments, the ground server 30 is further configured to:

[0119] Determine the regional location and production start time of the monitoring area;

[0120] Determine the historical water inflow of the monitoring area according to the location of the area and the start time of production;

[0121] The safe water inflow volume of the monitoring area is determined based on the historical water inflow volume of the monitoring area.

[0122] In some embodiments, the ground server 30 is further configured to:

[0123] A set of monitoring parameters corresponding to a water inrush event in the monitoring area is determined as a model optimization sample, and based on the model optimization sample, the target prediction model is optimized and fine-tuned.

[0124] In the embodiment of the present application, artificial intelligence multi-sensor data fusion is used to predict water hazards. It not only relies on multiple sensors to collect raw data, but also calculates information such as the rate of change and multi-channel mean, and explores the deep characteristics contained in the coal mine water hazard warning data, so as to better describe the complexity of the hydrogeological conditions of the coal mine. Compared with the artificially set empirical threshold alarm and mathematical statistics method, the warning result is more accurate and stable. Furthermore, the target prediction model can be continuously optimized and fine-tuned based on the model optimization sample, so as to continuously improve the prediction accuracy.

[0125] Figure 5 This is a schematic diagram of the structure of a water disaster prediction system provided in an embodiment of the present application. Figure 5 The water hazard prediction system shown may include, but is not limited to: a downhole sensing unit 10 , a ground monitoring unit 20 , a ground server 30 , a downhole monitoring unit 40 , a ground monitoring host 50 and a terminal device 60 .

[0126] In some embodiments, the downhole sensing unit 10 may include but is not limited to: a flow monitoring component 11, a water level monitoring component 12, a water pressure and temperature monitoring component 13, a goaf closed wall displacement monitoring component 14 and a goaf closed wall stress monitoring component 15.

[0127] In some embodiments, the ground monitoring unit 20 may include, but is not limited to: a wireless intelligent water level and water temperature monitor 21 and a meteorological multifunctional sensor 22 .

[0128] In some embodiments, the downhole monitoring unit 40 may include but is not limited to: multiple substations 41, a power source 42 and a downhole ring network 43. The power source 42 is used to supply power to the substation 41. The substation 41 communicates with the ground monitoring center 30 through the downhole ring network 43.

[0129] In such Figure 5 In the monitoring system shown, the underground sensing unit 10 can collect the monitoring data set of the tunnel and the borehole, and the ground monitoring unit 30 can collect the monitoring data set of the surface. Further, the underground sensing unit 10 can send the monitoring data set of the tunnel and the borehole to the ground server 30 through the underground monitoring unit 40, and the ground monitoring unit 20 sends the monitoring data set of the surface to the ground server 30 through the wireless network.

[0130] The ground monitoring server 30 can predict the water inflow in the monitoring area based on the monitoring data set received from different monitoring objects, and further, identify and determine the water hazard level in the monitoring area based on the predicted water inflow in the monitoring area. After determining that the water hazard level is higher than the set water hazard level, an alarm can be sent to the ground monitoring host 50 and the terminal device 60.

[0131] Optionally, the terminal device 60 includes at least one of a mobile phone, a wearable device, an Internet of Things device, a tablet computer (Pad), a computer with wireless transceiver function, a virtual reality (VR) terminal device, an augmented reality (AR) terminal device, a wireless terminal device in industrial control, and a wireless terminal device in a smart grid, but is not limited thereto.

[0132] In some embodiments, the ground server 30 may send relevant monitoring parameters, monitoring results and alarm information to the ground monitoring host 50 and the terminal device 60. The ground server 30 may also store the monitoring data, monitoring results and alarm information for subsequent data backtracking and analysis.

[0133] In the embodiment of the present application, artificial intelligence multi-sensor data fusion is used to predict water hazards. It not only relies on multiple sensors to collect raw data, but also calculates information such as the rate of change and multi-channel mean, and explores the deep characteristics contained in the coal mine water hazard warning data, so as to better describe the complexity of the hydrogeological conditions of the coal mine. Compared with the artificially set empirical threshold alarm and mathematical statistics method, the warning result is more accurate and stable. Furthermore, the target prediction model can be continuously optimized and fine-tuned based on the model optimization sample, so as to continuously improve the prediction accuracy.

[0134] It should be noted that the explanation of the aforementioned mining water hazard prediction method embodiment is also applicable to a mining water hazard prediction device of this embodiment, which will not be repeated here.

[0135] In order to implement the above embodiments, the present application also proposes an electronic device, comprising: a processor, and a memory communicatively connected to the processor; the memory stores computer-executable instructions; the processor executes the computer-executable instructions stored in the memory to implement the method provided by the above embodiments.

[0136] In order to implement the above embodiments, the present application also proposes a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the methods provided by the above embodiments.

[0137] In order to implement the above embodiments, the present application also proposes a computer program product, including a computer program, which implements the methods provided by the above embodiments when executed by a processor.

[0138] The collection, storage, use, processing, transmission, provision and application of user personal information involved in this application are in compliance with relevant laws and regulations and do not violate public order and good morals.

[0139] It should be noted that personal information from users should be collected for legitimate and reasonable purposes and should not be shared or sold outside of these legitimate uses. In addition, such collection / sharing should be carried out after receiving the user's informed consent, including but not limited to notifying the user to read the user agreement / user notice and sign the agreement / authorization including authorization of relevant user information before the user uses the function. In addition, any necessary steps should be taken to protect and safeguard access to such personal information data and ensure that others who have access to personal information data comply with its privacy policy and procedures.

[0140] This application is expected to provide an implementation scheme for users to selectively block the use or access of personal information data. That is, this application is expected to provide hardware and / or software to prevent or block access to such personal information data. Once the personal information data is no longer needed, risks can be minimized by limiting data collection and deleting data. In addition, when applicable, personal identifiers are removed from such personal information to protect the privacy of users.

[0141] In the description of the aforementioned embodiments, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0142] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0143] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code comprising one or more executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present application belong.

[0144] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute the instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purpose of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing in other suitable ways if necessary, and then stored in a computer memory.

[0145] It should be understood that the various parts of the present application can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or a combination thereof: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0146] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0147] In addition, each functional unit in each embodiment of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0148] The storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, etc. Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present application. A person of ordinary skill in the art may change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A method for predicting water damage in a mine, characterized in that: The method comprises: Collecting a set of monitoring parameters of different monitoring objects in the monitoring area, wherein the set of monitoring parameters includes multi-dimensional monitoring parameters related to hydrological monitoring, and the monitoring objects include ground monitoring objects and underground monitoring objects; For each dimension of the monitoring parameter, obtaining auxiliary information of the monitoring parameter, wherein the auxiliary information includes a change rate sequence and an average value sequence; Inputting the multi-dimensional monitoring parameters and the corresponding auxiliary information into a pre-trained target prediction model, and the target prediction model predicts the water inflow according to the multi-dimensional monitoring parameters and the corresponding auxiliary information to obtain a first predicted water inflow; A target water hazard level of the monitoring area is determined based on the first predicted water inflow, and an alarm message is issued in response to the target water hazard level indicating that a water inflow event has occurred in the monitoring area.

2. The method according to claim 1, characterized in that The training process of the target prediction model includes: Collecting training samples and labeled water inflow of the training samples; Inputting the training samples into a prediction model to be trained, and having the prediction model predict water inflow according to the training samples to obtain a second predicted water inflow; According to the labeled water inflow and the second predicted water inflow, the model parameters of the prediction model are adjusted, and the adjusted water inflow model is continuously trained until the training is completed to obtain the target prediction model.

3. The method according to claim 1 or 2, characterized in that: The process of acquiring the change rate sequence includes: For sampling time t i Monitoring parameters and sampling time t i-1 The difference calculation is performed on the monitoring parameters, and the difference and the sampling time t are obtained. i The ratio of the monitoring parameters to determine the sampling time t i The corresponding rate of change, where i is an integer greater than or equal to 1; The change rate sequence is obtained according to the change rate corresponding to the sampling time.

4. The method according to claim 1 or 2, characterized in that: The process of determining the average value sequence includes: For sampling time t i , sliding the time window to determine the sampling time t within the time window i Corresponding to multiple monitoring parameters, the multiple monitoring parameters in the time window are averaged to obtain the sampling time t i The corresponding average value; The variable average value sequence is obtained according to the average values ​​corresponding to the sampling moments.

5. The method according to claim 1, characterized in that Determining the target water hazard level of the monitoring area according to the first predicted water inflow includes: Obtaining a safe water inflow in the monitoring area, and determining a deviation value between the first predicted water inflow and the safe water inflow; The target water damage level is determined according to the deviation value.

6. The method according to claim 5, characterized in that The determining of the deviation value between the first predicted water inflow and the safe water inflow comprises: Obtaining a ratio between the first predicted water inflow and the safe water inflow; Determining the target water damage level according to the deviation value includes: Determining a target interval range of the ratio, wherein different ranges correspond to different water damage levels; A water hazard level associated with the target interval range is determined as the target water hazard level.

7. The method according to claim 5, characterized in that The determining of the deviation value between the first predicted water inflow and the safe water inflow comprises: Obtaining a difference between the first predicted water inflow and the safe water inflow; Determining the target water damage level according to the deviation value includes: Determining a target difference range within which the difference lies, wherein different difference ranges correspond to different water damage levels; The water damage level associated with the target difference range is used as the target water damage level.

8. The method according to any one of claims 5 to 7, characterized in that: The obtaining of the safe water inflow in the monitoring area comprises: Determine the regional location and production start time of the monitoring area; Determine the historical water inflow of the monitoring area according to the location of the area and the start time of production; The safe water inflow volume of the monitoring area is determined based on the historical water inflow volume of the monitoring area.

9. The method according to any one of claims 1 or 2 or 5-7, characterized in that: The method further comprises: A set of monitoring parameters corresponding to a water inrush event in the monitoring area is determined as a model optimization sample, and based on the model optimization sample, the target prediction model is optimized and fine-tuned.

10. A mining water hazard prediction system, characterized in that: The system comprises: A downhole sensing unit, used to collect a set of monitoring parameters of downhole monitoring objects in a monitoring area, wherein the set of monitoring parameters includes multi-dimensional monitoring parameters related to hydrological monitoring; A ground monitoring unit, used to collect a set of monitoring parameters of ground monitoring objects in the monitoring area; A ground server is used to obtain auxiliary information of the monitoring parameters in each dimension, wherein the auxiliary information includes a change rate sequence and an average value sequence, and input the multi-dimensional monitoring parameters and the corresponding auxiliary information into a pre-trained target prediction model, and the target prediction model predicts the water inflow according to the multi-dimensional monitoring parameters and the corresponding auxiliary information to obtain a first predicted water inflow, and determines a target water hazard level of the monitoring area based on the first predicted water inflow, and issues an alarm message in response to the target water hazard level indicating a water inflow event in the monitoring area.

Images