Rock mass instability stage early warning method, system, device and medium

Abstract

The application relates to a rock mass instability stage early warning method, system, device and medium. The method comprises the following steps: obtaining microseismic event monitoring data of a target rock mass region; calculating magnitude frequency distribution index sampling time series data, cumulative energy index sampling time series data and cumulative microseismic event number sampling time series data of the target rock mass region based on the microseismic event monitoring data of each monitoring time point; calculating magnitude frequency distribution index increment sampling time series data, cumulative energy index increment sampling time series data and cumulative microseismic event number increment sampling time series data according to the magnitude frequency distribution index sampling time series data, the cumulative energy index sampling time series data and the cumulative microseismic event number sampling time series data; and generating global instability stage classification information of the target rock mass region of each monitoring time point in combination with a magnitude frequency distribution index early warning threshold. The method can integrate multi-dimensional monitoring features and improve the accuracy of rock mass instability stage division.

Description

Technical Field

[0001] This application belongs to the field of rock mass early warning, and in particular relates to a method, system, equipment and medium for early warning of rock mass instability stage. Background Technology

[0002] Rock mass instability in mines primarily manifests as geological hazards such as rock bursts, surrounding rock collapses, and seismic stress. The frequency of these hazards increases with mining depth. These hazards typically cause fatalities and damage mining facilities and equipment. Given the severe consequences of rock mass instability, there is significant interest in developing efficient early warning models. These models are crucial tools for mine safety, providing timely warnings before hazards such as rock mass instability occur, thereby mitigating losses.

[0003] However, most existing early warning models rely on a single parameter, using only sudden changes or threshold exceedances as early warning indicators. Furthermore, the timeliness of these warnings is measured in days, making real-time alerts difficult to achieve. Because different monitoring systems employ varying principles, methods, and priorities, multi-source parameter early warning results may be inconsistent, leading to false alarms. Simultaneously, the inconsistent quality of early warning indicators creates difficulties for end-users and presents new challenges to the prevention and control of rock mass instability in underground mines. Current research has not yet fully integrated the advantages of various systems and parameters to develop a comprehensive early warning model. Summary of the Invention

[0004] Therefore, it is necessary to provide a method, system, equipment, and medium for predicting and realizing real-time dynamic early warning of rock mass instability stages, which can address the above-mentioned technical problems.

[0005] Firstly, this application provides a method for early warning of rock mass instability stages, including:

[0006] Acquire microseismic event monitoring data for the target rock mass area;

[0007] Based on the microseismic event monitoring data within the sampling time window of each monitoring time point, the sampling time series data of the magnitude frequency distribution index, the sampling time series data of the cumulative energy index, and the sampling time series data of the cumulative number of microseismic events in the target rock mass area are calculated.

[0008] Based on the sampling time series data of magnitude frequency distribution index, cumulative energy index, and cumulative microseismic event count, incremental sampling time series data of magnitude frequency distribution index, cumulative energy index, and cumulative microseismic event count are calculated respectively.

[0009] Based on the sampling time series data of magnitude frequency distribution index, cumulative energy index, cumulative microseismic event count, incremental sampling time series data of magnitude frequency distribution index, incremental sampling time series data of cumulative energy index, and incremental sampling time series data of cumulative microseismic event count, combined with the preset magnitude frequency distribution index early warning threshold, global instability stage classification information of the target rock mass area at each monitoring time series point is generated.

[0010] In one embodiment, the rock mass instability stage early warning method further includes:

[0011] The microseismic intensity index sampling time series data of each early warning grid in the target rock mass area is calculated based on the microseismic event monitoring data within the sampling time window of each monitoring time series point;

[0012] The incremental sampling time series data of the microseismic intensity index is calculated based on the sampling time series data of the microseismic intensity index.

[0013] Based on the global instability stage classification information, microseismic intensity index sampling time series data, and microseismic intensity index incremental sampling time series data, instability early warning information for each early warning grid at each monitoring time series point is generated.

[0014] In one embodiment, the expression for the microseismic intensity index sampling time series data is:

[0015]

[0016]

[0017]

[0018]

[0019]

[0020]

[0021]

[0022] In the formula, The timing sequence number is The grid number in the target rock mass region within the sampling time window of the monitoring time series point is The microseismic intensity index sampling time series data of the early warning grid, The timing sequence number is The sampling time window of the monitoring time series point The microseismic event belongs to the first The grid number of each sensor channel is The microseismic intensity index of the early warning grid, Record the microseismic intensity trigger threshold for microseismic events. The timing sequence number is The sampling time window of the monitoring time series point The first microseismic event Corrected Mercalli intensity coefficient for each sensor channel and The time sequence numbers are respectively The sampling time window of the monitoring time series point The first microseismic event The path coefficient and damping coefficient of each sensor channel, and The time sequence numbers are respectively The sampling time window of the monitoring time series point The micro-earthquake event is the same as the first The fitted source distance and actual straight-line distance for each sensor channel To fit the source depth, The timing sequence number is The sampling time window of the monitoring time series point The Richter magnitude of a micro-earthquake event The timing sequence number is The sampling time window of the monitoring time series point The microseismic event corresponds to the first... Seismic wave path attenuation correction term for each sensor channel, , , , , and For calibration coefficients, For the first The effective monitoring radius of each sensor channel This is the path coefficient attenuation scaling factor. is the damping coefficient and attenuation rate factor.

[0023] In one embodiment, based on the sampling time-series data of magnitude frequency distribution index, cumulative energy index, cumulative microseismic event count, incremental sampling time-series data of magnitude frequency distribution index, incremental sampling time-series data of cumulative energy index, and incremental sampling time-series data of cumulative microseismic event count, and combined with a preset magnitude frequency distribution index early warning threshold, global instability stage classification information of the target rock mass region at each monitoring time-series point is generated, including:

[0024] The time series data of the magnitude frequency distribution index warning indication tag is calculated based on the sampling time series data of the magnitude frequency distribution index and the warning threshold of the magnitude frequency distribution index.

[0025] The time series data of the magnitude frequency distribution index early warning indicator tag, the time series data of the cumulative energy index sampling, the time series data of the cumulative microseismic event sampling, the time series data of the incremental sampling of the magnitude frequency distribution index, the time series data of the incremental sampling of the cumulative energy index, and the time series data of the incremental sampling of the cumulative microseismic event are input into the machine learning model for the probability prediction sequence analysis of the instability stage to generate the probability distribution information of the instability stage prediction for each monitoring time series point.

[0026] Global instability stage classification information is generated based on the predicted probability distribution information of the instability stage.

[0027] In one embodiment, the rock mass instability stages include a development stage, a peak stage, and a quiescent stage. The predicted probability distribution information for the instability stages includes the predicted probability for the development stage, the predicted probability for the peak stage, and the predicted probability for the quiescent stage. The expression for the stage prediction probability distribution loss function of the machine learning model for instability stage probability prediction sequence analysis is as follows:

[0028]

[0029] In the formula, The loss function is the probability distribution for stage prediction. For training batch size, The time series length of a single training sample. For the first Category weights for rock mass instability stages For the first The time sequence number of the training batch is The monitoring time series point belongs to the first The true hallmark of the unstable stage of rock mass For the first The time sequence number of the training batch is The monitoring time series point belongs to the first Predicted probability of instability stage of rock mass.

[0030] In one embodiment, the expressions for the sampling time series data of the magnitude frequency distribution index and the timing data of the magnitude frequency distribution index early warning indication tag are as follows:

[0031]

[0032]

[0033] In the formula, The timing sequence number is The sampling time window of the monitoring time series points, The timing sequence number is Sampling time window of monitoring time series points The Richter magnitude is greater than or equal to The total number of microseismic events, The timing sequence number is Sampling time window of monitoring time series points Gutenberg-Richard intercept within the range, The time series data corresponding to the magnitude frequency distribution index sampling time series data is... Sampling time window of monitoring time series points The magnitude frequency distribution index, The magnitude is measured on the Richter scale. The earthquake frequency distribution index warning indicator tag time series data corresponds to the time series number. The earthquake magnitude frequency distribution index early warning indicator label at the monitoring time point, The magnitude frequency distribution index is the early warning threshold.

[0034] In one embodiment, the sampling time end point of the sampling time window for monitoring time points is the monitoring time point.

[0035] Secondly, this application also provides an early warning system for the rock mass instability stage, including:

[0036] The microseismic event monitoring module is used to acquire microseismic event monitoring data for the target rock mass area;

[0037] The index time series calculation module is used to calculate the magnitude frequency distribution index sampling time series data, cumulative energy index sampling time series data, and cumulative microseismic event number sampling time series data of the target rock mass area based on the microseismic event monitoring data within the sampling time window of each monitoring time series point.

[0038] The incremental time series calculation module is used to calculate the incremental sampling time series data of the magnitude frequency distribution index, the cumulative energy index, and the cumulative microseismic event count based on the sampling time series data of the magnitude frequency distribution index, the cumulative energy index, and the cumulative microseismic event count, respectively.

[0039] The instability stage classification module is used to generate global instability stage classification information for the target rock mass area at each monitoring time point based on the sampling time series data of magnitude frequency distribution index, cumulative energy index, cumulative microseismic event count, incremental sampling time series data of magnitude frequency distribution index, incremental sampling time series data of cumulative energy index, and incremental sampling time series data of cumulative microseismic event count, combined with the preset magnitude frequency distribution index early warning threshold.

[0040] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the method as described in any of the first aspects of this application.

[0041] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described in any of the first aspects of this application.

[0042] The aforementioned methods, systems, equipment, and media for early warning of rock mass instability stages can improve the real-time performance and accuracy of rock mass instability early warning through multi-parameter collaborative analysis and dynamic incremental modeling, and can also reduce the false judgment rate of the early warning model. By using time-driven dynamic identification of the entire process, and constructing time-series data and its incremental features within the sampling time window, it is possible to identify rock mass instability stages in stages and capture early signs of instability in a timely manner, thereby enhancing the accuracy and real-time performance of rock mass instability stage prediction. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 A flowchart illustrating a rock mass instability stage early warning method provided in one embodiment of this application. Figure 1 ;

[0045] Figure 2 A flowchart illustrating a rock mass instability stage early warning method provided in one embodiment of this application. Figure 2 ;

[0046] Figure 3 This is a schematic diagram of the structure of an early warning system for rock mass instability stage provided in one embodiment of this application. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0048] In one exemplary embodiment of this application, such as Figure 1As shown, a method for early warning of rock mass instability is provided. This embodiment illustrates the application of this method to an instability stage early warning terminal. It is understood that this method can also be applied to an instability stage early warning server, and further to an instability stage early warning system including both an instability stage early warning terminal and an instability stage early warning server, and is implemented through the interaction between the two. In this embodiment, the method includes the following steps:

[0049] Step S101: Obtain microseismic event monitoring data for the target rock mass area.

[0050] Specifically, the instability stage early warning terminal can communicate with various sensor channels in the target rock mass area. The terminal acquires microseismic event monitoring data of the target rock mass area through these sensor channels. This microseismic event monitoring data can be used to characterize the fracturing activity and energy changes within the target rock mass area.

[0051] Optionally, microseismic event monitoring data may include, but is not limited to, microseismic event timestamps, microseismic event source spatial coordinates, microseismic event seismic moments, and microseismic event microseismic energy.

[0052] Furthermore, the instability stage early warning terminal can calculate the Richter magnitude of microseismic events based on the seismic moment of the microseismic events.

[0053] Step S102: Based on the microseismic event monitoring data within the sampling time window of each monitoring time point, calculate the sampling time series data of the magnitude frequency distribution index, the cumulative energy index, and the cumulative number of microseismic events in the target rock mass area.

[0054] Specifically, the instability stage early warning terminal can identify microseismic event monitoring data within a preset sampling time window for each monitoring time point based on the microseismic event timestamp. The instability stage early warning terminal can calculate the magnitude frequency distribution index sampling time series data, cumulative energy index sampling time series data, and cumulative microseismic event number sampling time series data of the target rock mass area at each monitoring time point based on the microseismic event monitoring data within the preset sampling time window for each monitoring time point.

[0055] Optionally, the instability stage early warning terminal can obtain the Richter magnitude of microseismic events based on microseismic event monitoring data. Based on the Richter magnitude of microseismic events and combined with the Gutenberg-Rickett law, the terminal can construct a logarithmic relationship model between the Richter magnitude of microseismic events and the number of microseismic events by statistically analyzing the total number of microseismic events with a Richter magnitude greater than or equal to a specific Richter magnitude within the sampling time window. This allows the terminal to calculate the magnitude frequency distribution index corresponding to each monitoring time point in the sampling time series data.

[0056] Optionally, the expression for the Gutenberg-Richard law can be...

[0057]

[0058] In the formula, The timing sequence number is The sampling time window of the monitoring time series points, The timing sequence number is Sampling time window of monitoring time series points The Richter magnitude is greater than or equal to The total number of microseismic events, The timing sequence number is Sampling time window of monitoring time series points Gutenberg-Richard intercept within the range, The time series data corresponding to the magnitude frequency distribution index sampling time series data is... Sampling time window of monitoring time series points The magnitude frequency distribution index, The magnitude is indicated on the Richter scale.

[0059] Optionally, the instability stage early warning terminal can sum the microseismic energy of microseismic events in the microseismic event monitoring data within a preset sampling time window for each monitoring time point to calculate the cumulative energy index corresponding to each monitoring time point in the sampling time series data. The cumulative energy index can characterize the total scale of energy released by the rock mass within the sampling time window of the monitoring time point. Its trend is highly correlated with the evolution of the rock mass instability stage. During the development and peak periods, the cumulative energy index mostly shows an upward trend, while during the static period, the cumulative energy index decreases significantly or remains stable at a low level.

[0060] Optionally, the instability stage early warning terminal can generate a cumulative number of microseismic events corresponding to each monitoring time point in the cumulative microseismic event sampling time series data, based on the total number of microseismic events corresponding to the microseismic event monitoring data within a preset sampling time window for each monitoring time point. The cumulative number of microseismic events can characterize the frequency of rock mass fracturing activity within the sampling time window, and the increase or decrease in the cumulative number of microseismic events can characterize the intensity of rock mass stress adjustment.

[0061] Step S103: Based on the sampling time series data of magnitude frequency distribution index, cumulative energy index, and cumulative microseismic event count, the incremental sampling time series data of magnitude frequency distribution index, cumulative energy index, and cumulative microseismic event count are calculated respectively.

[0062] Optionally, the instability stage early warning terminal can calculate the forward difference of the sampling time series data of magnitude frequency distribution index, cumulative energy index, and cumulative microseismic event count corresponding to each monitoring time series point, to obtain the increment of magnitude frequency distribution index, cumulative energy index, and cumulative microseismic event count for each monitoring time series point. Based on the increment of magnitude frequency distribution index, cumulative energy index, and cumulative microseismic event count for each monitoring time series point, it can construct the sampling time series data of magnitude frequency distribution index increment, cumulative energy index increment, and cumulative microseismic event count increment.

[0063] Step S104: Based on the sampling time series data of magnitude frequency distribution index, cumulative energy index, cumulative microseismic event count, incremental sampling time series data of magnitude frequency distribution index, incremental sampling time series data of cumulative energy index, and incremental sampling time series data of cumulative microseismic event count, and combined with the preset magnitude frequency distribution index early warning threshold, generate global instability stage classification information of the target rock mass area at each monitoring time series point.

[0064] Optionally, the rock mass instability stage may include the development stage, peak stage, and quiescent stage, and the global instability stage classification information may include development stage classification information, peak stage classification information, and quiescent stage classification information.

[0065] Indicatively, during the development phase, the magnitude frequency distribution index shows a continuous fluctuating upward trend, reflecting a gradual increase in the proportion of small-magnitude microseismic events, indicating that internal cracks in the rock mass are in the gestation and propagation stage. The cumulative energy index shows a steady upward trend, indicating that the rock mass is continuously accumulating stress and energy without any obvious release. The cumulative number of microseismic events gradually increases, and the frequency of microseismic activity continues to rise, reflecting the intensified fracturing activity of the rock mass after disturbance.

[0066] As an illustration, during peak periods, the magnitude frequency distribution index may turn around after reaching a certain peak, and in some cases approach or touch the warning threshold of the magnitude frequency distribution index. The proportion of large-magnitude microseismic events has increased significantly, and rock mass fissures have entered a critical stage of penetration and propagation. The cumulative energy index is growing at an accelerated rate, and a rapid short-term increase is possible, indicating that rock mass energy accumulation has reached a high level and is nearing a release threshold. The cumulative number of microseismic events is showing explosive growth, microseismic activity is extremely frequent, and rock mass fracturing activity has entered an active state.

[0067] Indicatively, during the quiescent period, the magnitude frequency distribution index tends to stabilize, with a significant reduction in fluctuation amplitude, gradually returning to normal levels, and the proportion of microseismic events of varying magnitudes returns to equilibrium. The cumulative energy index stagnates or shows a slow decline, indicating that the energy accumulated in the rock mass has been gradually dissipated through instability events or slow release. The growth of the cumulative number of microseismic events slows down until it stabilizes, the frequency of microseismic activity decreases, and rock mass fracturing activity enters a period of cessation.

[0068] In the aforementioned rock mass instability stage early warning method, by calculating the sampling time series data of the magnitude frequency distribution index, cumulative energy index, and cumulative microseismic event count of the target rock mass region based on microseismic event monitoring data, key characteristic parameters related to rock mass instability can be extracted, and a continuous and complete time series data sequence can be constructed. This lays a solid data foundation for the accurate judgment of subsequent rock mass instability stages and improves the feasibility and accuracy of rock mass instability stage early warning. By calculating the increments of the magnitude frequency distribution index, cumulative energy index, and cumulative microseismic event count, the method can accurately capture the core parameters of each rock mass instability stage. The changes in core parameters over time reveal the intrinsic correlation between stress accumulation, energy accumulation, and microseismic activity frequency in rock masses, reflecting the gradual evolution of rock mass instability from accumulation to criticality and then to stability. This provides a dynamic and real-time basis for classifying instability stages. Through multi-dimensional time-series data fusion and early warning threshold linkage analysis, the rock mass state corresponding to different data characteristics can be accurately identified, improving the accuracy of rock mass instability stage classification. This enables precise identification of the rock mass state at each monitoring time point, which in turn can provide guidance for the formulation of targeted protection measures, reducing casualties and property losses caused by rock mass instability disasters.

[0069] In an optional embodiment of this application, please refer to Figure 2 Early warning methods for rock mass instability stages also include:

[0070] Step S207: Calculate the microseismic intensity index sampling time series data of each early warning grid in the target rock mass region based on the microseismic event monitoring data within the sampling time window of each monitoring time series point.

[0071] Step S208: Calculate the incremental sampling time series data of the microseismic intensity index based on the microseismic intensity index sampling time series data.

[0072] Optionally, the instability stage early warning terminal can calculate the forward difference of the microseismic intensity index sampling time series data corresponding to each monitoring time series point, and obtain the microseismic intensity index increment for each monitoring time series point. The instability stage early warning terminal can construct microseismic intensity index increment sampling time series data based on the microseismic intensity index increments for each monitoring time series point.

[0073] Step S209: Based on the global instability stage classification information, the microseismic intensity index sampling time series data, and the microseismic intensity index incremental sampling time series data, generate instability early warning information for each early warning grid at each monitoring time series point.

[0074] For example, the instability stage early warning terminal can obtain the microseismic intensity index risk Boolean label for each early warning grid at each monitoring time point based on the microseismic intensity index sampling time series data corresponding to each monitoring time point, combined with a preset microseismic intensity index risk threshold. The instability stage early warning terminal can also obtain the microseismic intensity index increment risk Boolean label for each early warning grid at each monitoring time point based on the microseismic intensity index increment sampling time series data corresponding to each monitoring time point, combined with a preset microseismic intensity index increment risk threshold. Finally, the instability stage early warning terminal can generate instability early warning information for each early warning grid at each monitoring time point based on global instability stage classification information, microseismic intensity index risk Boolean labels, and microseismic intensity index increment risk Boolean labels.

[0075] Optionally, the instability phase early warning terminal can set the Boolean label of the microseismic intensity index increment risk at each monitoring time point for each early warning grid where the microseismic intensity index increment is greater than the microseismic intensity index increment risk threshold.

[0076] Optionally, the instability warning information for each early warning grid at each monitoring time point may include red warning information, yellow warning information, and green warning information.

[0077] Furthermore, the instability stage early warning terminal can set the instability early warning information at each monitoring time point of each early warning grid where the global instability stage classification information is the peak period classification information and either the microseismic intensity index risk Boolean label or the microseismic intensity index incremental risk Boolean label is 1, as a red early warning. The instability stage early warning terminal can set the instability early warning information at each monitoring time point of each early warning grid where the global instability stage classification information is the peak period classification information and both the microseismic intensity index risk Boolean label and the microseismic intensity index incremental risk Boolean label are 1, as a yellow early warning. The instability stage early warning terminal can set the instability early warning information at each monitoring time point of each early warning grid where the global instability stage classification information is the development period classification information and both the microseismic intensity index risk Boolean label and the microseismic intensity index incremental risk Boolean label are 1, as a yellow early warning. The instability stage early warning terminal can set the instability early warning information at each monitoring time point of each early warning grid where the global instability stage classification information is the development period classification information and not both the microseismic intensity index risk Boolean label and the microseismic intensity index incremental risk Boolean label are 1, as a green early warning. The instability phase early warning terminal can set the instability early warning information of each early warning grid at each monitoring time point to green early warning information, where the global instability phase classification information is the quiescent period classification information.

[0078] Indicatively, a green alert indicates that the target rock mass area is in a safe and controllable state, allowing for normal mining operations, while continuously monitoring changes in indicators. A yellow alert indicates that the target rock mass area is in a critical risk state, requiring the investigation of potential risk points and the implementation of measures such as support and reinforcement to eliminate hidden dangers. A red alert indicates that the target rock mass area is at risk of instability, requiring the cessation of operations, evacuation to a safe area, implementation of risk control measures, and close monitoring of dynamic changes in indicators.

[0079] In an optional embodiment of this application, the expression for the microseismic intensity index sampling time series data can be:

[0080]

[0081]

[0082]

[0083]

[0084]

[0085]

[0086]

[0087] In the formula, The timing sequence number is The grid number in the target rock mass region within the sampling time window of the monitoring time series point is The microseismic intensity index sampling time series data of the early warning grid, The timing sequence number is The sampling time window of the monitoring time series point The microseismic event belongs to the first The grid number of each sensor channel is The microseismic intensity index of the early warning grid, Record the microseismic intensity trigger threshold for microseismic events. The timing sequence number is The sampling time window of the monitoring time series point The first microseismic event Corrected Mercalli intensity coefficient for each sensor channel and The time sequence numbers are respectively The sampling time window of the monitoring time series point The first microseismic event The path coefficient and damping coefficient of each sensor channel, and The time sequence numbers are respectively The sampling time window of the monitoring time series point The micro-earthquake event is the same as the first The fitted source distance and actual straight-line distance for each sensor channel To fit the source depth, The timing sequence number is The sampling time window of the monitoring time series point The Richter magnitude of a micro-earthquake event The timing sequence number is The sampling time window of the monitoring time series point The microseismic event corresponds to the first... Seismic wave path attenuation correction term for each sensor channel, , , , , and For calibration coefficients, For the first The effective monitoring radius of each sensor channel This is the path coefficient attenuation scaling factor. is the damping coefficient and attenuation rate factor.

[0088] For example, the effective monitoring radius of a sensor channel It can be 50 to 100 meters. This value is merely an example; it represents the effective monitoring radius of the sensor channel. The value can be any value that satisfies the effective monitoring of the sensor channel, and no restrictions are imposed here.

[0089] For example, in deeply buried hard rock mines, the path loss during microwave propagation mainly comes from scattering through the tiny pores of the rock medium, rather than strong absorption by soft rock. Experimental measurements show that when microwaves propagate 100 meters, the intensity attenuation due to path loss is approximately 10% to 15%. At this point, the path coefficient attenuation scaling factor... It can take the value 100, using Characterizing the cumulative effect of path loss (e.g.) =100 meters, =2; =50 meters, =1.5).

[0090] For example, in deeply buried hard rock mines, the energy attenuation of microwaves (mainly shear waves) follows an "exponential law," and the attenuation rate is slower than that of soft rock / shallow rock masses. Field measurements in an exemplary mining area revealed that the damping coefficient attenuation rate factor... The value can be 10.

[0091] For example, field measurements in a specific mining area revealed that the calibration coefficient could take the following values: .

[0092] Optionally, microseismic event recording microseismic intensity trigger threshold. The value can be 0.2.

[0093] Indicative Can represent index variables Starting from 1, gradually increasing to ,in, The timing sequence number is The total number of microseismic events within the sampling time window of the monitoring time series points.

[0094] Indicative Can represent index variables Starting from 1, gradually increasing to ,in, This represents the total number of sensor channels.

[0095] In an optional embodiment of this application, please refer to Figure 1 and Figure 2Step S104: Based on the sampling time-series data of magnitude frequency distribution index, cumulative energy index, cumulative microseismic event count, incremental sampling time-series data of magnitude frequency distribution index, incremental sampling time-series data of cumulative energy index, and incremental sampling time-series data of cumulative microseismic event count, and combined with the preset magnitude frequency distribution index early warning threshold, global instability stage classification information of the target rock mass area at each monitoring time-series point is generated, which may include:

[0096] Step S204: Calculate the magnitude frequency distribution index warning indication tag time series data based on the sampling time series data of the magnitude frequency distribution index and the magnitude frequency distribution index warning threshold.

[0097] Step S205: Input the time series data of the magnitude frequency distribution index early warning indicator tag, the time series data of the cumulative energy index sampling, the time series data of the cumulative microseismic event sampling, the time series data of the magnitude frequency distribution index incremental sampling, the time series data of the cumulative energy index incremental sampling, and the time series data of the cumulative microseismic event incremental sampling into the machine learning model for instability stage probability prediction sequence analysis to generate instability stage prediction probability distribution information for each monitoring time series point.

[0098] Optionally, the machine learning model for probability prediction sequence analysis of the instability phase can be constructed based on one or more of the following models: gated recurrent unit model, long short-term memory network model, and temporal convolutional network model.

[0099] Step S206: Generate global instability stage classification information based on the predicted probability distribution information of the instability stage.

[0100] In an optional embodiment of this application, the rock mass instability stage may include a development stage, a peak stage, and a quiescent stage. The instability stage prediction probability distribution information may include the development stage prediction probability, the peak stage prediction probability, and the quiescent stage prediction probability. The expression for the stage prediction probability distribution loss function of the instability stage probability prediction sequence analysis machine learning model can be:

[0101]

[0102] In the formula, The loss function is the probability distribution for stage prediction. For training batch size, The time series length of a single training sample. For the first Category weights for rock mass instability stages For the first The time sequence number of the training batch is The monitoring time series point belongs to the first The true hallmark of the unstable stage of rock mass For the first The time sequence number of the training batch is The monitoring time series point belongs to the first Predicted probability of instability stage of rock mass.

[0103] Optional, It can predict probabilities during the development phase. It can predict the probability of peak periods. It can be used to predict probabilities during the quiescent period.

[0104] Optional, This can be used to assign category weights during the development phase. This can be used to assign category weights during peak periods. This can be the class weight during the quiescent period.

[0105] In an optional embodiment of this application, the expressions for the sampling time series data of the magnitude frequency distribution index and the timing data of the warning indication tag of the magnitude frequency distribution index can be:

[0106]

[0107]

[0108] In the formula, The timing sequence number is The sampling time window of the monitoring time series points, The timing sequence number is Sampling time window of monitoring time series points The Richter magnitude is greater than or equal to The total number of microseismic events, The timing sequence number is Sampling time window of monitoring time series points Gutenberg-Richard intercept within the range, The time series data corresponding to the magnitude frequency distribution index sampling time series data is... Sampling time window of monitoring time series points The magnitude frequency distribution index, The magnitude is measured on the Richter scale. The earthquake frequency distribution index warning indicator tag time series data corresponds to the time series number. The earthquake magnitude frequency distribution index early warning indicator label at the monitoring time point, The magnitude frequency distribution index is the early warning threshold.

[0109] Optional, magnitude frequency distribution index warning threshold The value can be 0.457.

[0110] Furthermore, the instability stage early warning terminal can calculate the magnitude frequency distribution index based on the maximum likelihood estimation method or the least squares method.

[0111] In an optional embodiment of this application, the sampling time end point of the sampling time window for monitoring time points can be the monitoring time point itself.

[0112] For example, the time interval between adjacent monitoring time points may be, but is not limited to, 1 hour, and the time window length of the sampling time window may be, but is not limited to, 6 hours, 8 hours, 12 hours or 24 hours.

[0113] In one example embodiment of this application, such as Figure 2 As shown, a method for early warning of rock mass instability is provided, including:

[0114] Step S201: Obtain microseismic event monitoring data for the target rock mass area.

[0115] Step S202: Based on the microseismic event monitoring data within the sampling time window of each monitoring time point, calculate the sampling time series data of the magnitude frequency distribution index, the cumulative energy index, and the cumulative number of microseismic events in the target rock mass area.

[0116] Step S203: Based on the sampling time series data of magnitude frequency distribution index, cumulative energy index, and cumulative microseismic event count, the incremental sampling time series data of magnitude frequency distribution index, cumulative energy index, and cumulative microseismic event count are calculated respectively.

[0117] Step S204: Calculate the magnitude frequency distribution index warning indication tag time series data based on the sampling time series data of the magnitude frequency distribution index and the magnitude frequency distribution index warning threshold.

[0118] Step S205: Input the time series data of the magnitude frequency distribution index early warning indicator tag, the time series data of the cumulative energy index sampling, the time series data of the cumulative microseismic event sampling, the time series data of the magnitude frequency distribution index incremental sampling, the time series data of the cumulative energy index incremental sampling, and the time series data of the cumulative microseismic event incremental sampling into the machine learning model for instability stage probability prediction sequence analysis to generate instability stage prediction probability distribution information for each monitoring time series point.

[0119] Step S206: Generate global instability stage classification information based on the predicted probability distribution information of the instability stage.

[0120] Step S207: Calculate the microseismic intensity index sampling time series data of each early warning grid in the target rock mass region based on the microseismic event monitoring data within the sampling time window of each monitoring time series point.

[0121] Step S208: Calculate the incremental sampling time series data of the microseismic intensity index based on the microseismic intensity index sampling time series data.

[0122] Step S209: Based on the global instability stage classification information, the microseismic intensity index sampling time series data, and the microseismic intensity index incremental sampling time series data, generate instability early warning information for each early warning grid at each monitoring time series point.

[0123] In the aforementioned early warning method for rock mass instability stages, by constructing a multi-dimensional time-series data system and its incremental characteristics, including magnitude frequency distribution index, cumulative energy index, and cumulative number of microseismic events, and combining it with a machine learning model for probability prediction sequence analysis of instability stages, the accuracy and real-time performance of rock mass instability stage classification can be improved. By calculating the microseismic intensity index and its increment for each early warning grid, refined dynamic identification of rock mass instability risk can be achieved, thereby reducing the misjudgment rate of the early warning model, improving the spatial resolution of the early warning model, and ensuring dynamic and scientific early warning of rock mass instability stages in underground mines.

[0124] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0125] Based on the same inventive concept, this application also provides a rock mass instability stage early warning system for implementing the aforementioned rock mass instability stage early warning method. The solution provided by this system is similar to the solution described in the above method; therefore, the specific limitations in one or more rock mass instability stage early warning system embodiments provided below can be found in the limitations of the rock mass instability stage early warning method described above, and will not be repeated here.

[0126] In one exemplary embodiment, such as Figure 3 As shown, a rock mass instability stage early warning system 300 is provided, including:

[0127] The microseismic event monitoring module 301 can be used to acquire microseismic event monitoring data of the target rock mass area.

[0128] The index time series calculation module 302 can be used to calculate the magnitude frequency distribution index sampling time series data, cumulative energy index sampling time series data, and cumulative microseismic event number sampling time series data of the target rock mass area based on the microseismic event monitoring data within the sampling time window of each monitoring time series point.

[0129] The incremental time series calculation module 303 can be used to calculate the incremental sampling time series data of the magnitude frequency distribution index, the incremental sampling time series data of the cumulative energy index, and the incremental sampling time series data of the cumulative microseismic events based on the sampling time series data of the magnitude frequency distribution index, the sampling time series data of the cumulative energy index, and the sampling time series data of the cumulative microseismic events.

[0130] The instability stage classification module 304 can be used to generate global instability stage classification information for target rock mass areas at each monitoring time point based on the sampling time series data of magnitude frequency distribution index, cumulative energy index, cumulative microseismic event count, incremental sampling time series data of magnitude frequency distribution index, incremental sampling time series data of cumulative energy index, and incremental sampling time series data of cumulative microseismic event count, combined with a preset magnitude frequency distribution index early warning threshold.

[0131] In an optional embodiment of this application, the rock mass instability stage early warning system 300 can also be used for:

[0132] Based on the microseismic event monitoring data within the sampling time window of each monitoring time point, the sampling time series data of the microseismic intensity index of each early warning grid in the target rock mass region are calculated.

[0133] The incremental sampling time series data of the microseismic intensity index is calculated based on the sampling time series data of the microseismic intensity index.

[0134] Based on the global instability stage classification information, microseismic intensity index sampling time series data, and microseismic intensity index incremental sampling time series data, instability early warning information for each early warning grid at each monitoring time series point is generated.

[0135] In an optional embodiment of this application, the instability phase classification module 304 may also be used for:

[0136] The magnitude frequency distribution index early warning indication tag time series data is calculated based on the sampling time series data of the magnitude frequency distribution index and the magnitude frequency distribution index early warning threshold.

[0137] The time-series data of the magnitude frequency distribution index early warning indicator tag, the cumulative energy index sampling time-series data, the cumulative microseismic event sampling time-series data, the magnitude frequency distribution index incremental sampling time-series data, the cumulative energy index incremental sampling time-series data, and the cumulative microseismic event incremental sampling time-series data are input into the machine learning model for instability stage probability prediction sequence analysis to generate instability stage prediction probability distribution information for each monitoring time-series point.

[0138] Global instability stage classification information is generated based on the predicted probability distribution information of the instability stage.

[0139] In one embodiment, a computer device is provided, including a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement the steps of the rock mass instability stage early warning method as described above.

[0140] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the above method embodiments.

[0141] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The components described as separate parts may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this disclosure according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0142] The above-described embodiments are merely illustrative of several implementation methods of the embodiments of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of the patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the embodiments of this application, and these modifications and improvements all fall within the protection scope of the embodiments of this application.

Claims

1. A method for early warning of rock mass instability stages, characterized in that, The method includes: Acquire microseismic event monitoring data for the target rock mass area; Based on the microseismic event monitoring data within the sampling time window of each monitoring time point, the sampling time series data of the magnitude frequency distribution index, the sampling time series data of the cumulative energy index, and the sampling time series data of the cumulative number of microseismic events in the target rock mass area are calculated. Based on the sampling time series data of the magnitude frequency distribution index, the sampling time series data of the cumulative energy index, and the sampling time series data of the cumulative number of microseismic events, incremental sampling time series data of the magnitude frequency distribution index, incremental sampling time series data of the cumulative energy index, and incremental sampling time series data of the cumulative number of microseismic events are calculated respectively. Based on the sampling time series data of the magnitude frequency distribution index, the sampling time series data of the cumulative energy index, the sampling time series data of the cumulative number of microseismic events, the incremental sampling time series data of the magnitude frequency distribution index, the incremental sampling time series data of the cumulative energy index, and the incremental sampling time series data of the cumulative number of microseismic events, combined with the preset magnitude frequency distribution index early warning threshold, global instability stage classification information of the target rock mass region at each of the monitoring time series points is generated; Based on the microseismic event monitoring data within the sampling time window of each of the monitoring time points, calculate the microseismic intensity index sampling time series data of each early warning grid in the target rock mass region; The incremental sampling time series data of the microseismic intensity index is calculated based on the microseismic intensity index sampling time series data. Based on the global instability stage classification information, the microseismic intensity index sampling time series data, and the microseismic intensity index incremental sampling time series data, instability early warning information for each of the early warning grids at each of the monitoring time series points is generated; The expression for the sampling time series data of the microseismic intensity index is as follows: ； ； ； ； ； ； ； In the formula, The timing sequence number is The grid number in the target rock mass region within the sampling time window of the monitoring time series point is The microseismic intensity index sampling time series data of the aforementioned early warning grid. The timing sequence number is The sampling time window of the monitoring time series point The microseismic event belongs to the first The grid number of each sensor channel is The microseismic intensity index of the aforementioned early warning grid, Record the microseismic intensity trigger threshold for microseismic events. The timing sequence number is The sampling time window of the monitoring time series point The first micro-earthquake event Corrected Mercalli intensity coefficient for each sensor channel and The time sequence numbers are respectively The sampling time window of the monitoring time series point The first micro-earthquake event The path coefficient and damping coefficient of each sensor channel, and The time sequence numbers are respectively The sampling time window of the monitoring time series point The micro-earthquake event is the same as the first The fitted source distance and actual straight-line distance for each sensor channel To fit the source depth, The timing sequence number is The sampling time window of the monitoring time series point The Richter magnitude of a micro-earthquake event The timing sequence number is The sampling time window of the monitoring time series point The microseismic event corresponds to the first... Seismic wave path attenuation correction term for each sensor channel, , , , , and For calibration coefficients, For the first The effective monitoring radius of each sensor channel This is the path coefficient attenuation scaling factor. is the damping coefficient and attenuation rate factor.

2. The method according to claim 1, characterized in that, The step of generating instability early warning information for each of the early warning grids at each of the monitoring time points based on the global instability stage classification information, the microseismic intensity index sampling time series data, and the microseismic intensity index incremental sampling time series data includes: Based on the microseismic intensity index corresponding to each monitoring time point in the sampling time series data of the microseismic intensity index of each early warning grid, and combined with the preset microseismic intensity index risk threshold, the microseismic intensity index risk Boolean label of each early warning grid at each monitoring time point is obtained. Based on the microseismic intensity index increment sampling time series data corresponding to each monitoring time series point in each of the early warning grids, and combined with the preset microseismic intensity index increment risk threshold, the microseismic intensity index increment risk Boolean label of each of the early warning grids at each of the monitoring time series points is obtained; Based on the global instability stage classification information, the microseismic intensity index risk Boolean label, and the microseismic intensity index incremental risk Boolean label, instability early warning information for each of the early warning grids at each of the monitoring time series points is generated.

3. The method according to claim 1, characterized in that, The method, based on the sampling time-series data of the magnitude frequency distribution index, the cumulative energy index, the cumulative microseismic event count, the incremental sampling time-series data of the magnitude frequency distribution index, the incremental sampling time-series data of the cumulative energy index, and the incremental sampling time-series data of the cumulative microseismic event count, and combined with a preset magnitude frequency distribution index early warning threshold, generates global instability stage classification information for the target rock mass region at each of the monitoring time-series points, including: The magnitude frequency distribution index early warning indication tag time series data is calculated based on the sampling time series data of the magnitude frequency distribution index and the early warning threshold of the magnitude frequency distribution index. The time-series data of the magnitude frequency distribution index early warning indicator tag, the time-series data of the cumulative energy index sampling, the time-series data of the cumulative microseismic event count sampling, the time-series data of the magnitude frequency distribution index incremental sampling, the time-series data of the cumulative energy index incremental sampling, and the time-series data of the cumulative microseismic event count incremental sampling are input into the machine learning model for instability stage probability prediction sequence analysis to generate instability stage prediction probability distribution information for each of the monitoring time-series points; The global instability stage classification information is generated based on the predicted probability distribution information of the instability stage.

4. The method according to claim 3, characterized in that, The rock mass instability stages include a development stage, a peak stage, and a quiescent stage. The predicted probability distribution information for each instability stage includes the predicted probability for the development stage, the peak stage, and the quiescent stage. The expression for the stage prediction probability distribution loss function of the machine learning model for instability stage probability prediction sequence analysis is as follows: ； In the formula, The loss function is the probability distribution prediction function for the stage. For training batch size, The time series length of a single training sample. For the first The category weights of the rock mass instability stages described above. For the first The time sequence number of the training batch is The monitoring time point mentioned above belongs to the first The true label of the rock mass instability stage described above For the first The time sequence number of the training batch is The monitoring time point mentioned above belongs to the first The predicted probability of the rock mass instability stage.

5. The method according to claim 3, characterized in that, The expressions for the sampling time series data of the magnitude frequency distribution index and the time series data of the early warning indication tag of the magnitude frequency distribution index are as follows: ； ； In the formula, The timing sequence number is The sampling time window of the monitoring time series points, The timing sequence number is The sampling time window of the monitoring time series points The Richter magnitude is greater than or equal to The total number of microseismic events, The timing sequence number is The sampling time window of the monitoring time series points Gutenberg-Richard intercept within the range, The time series data corresponding to the time series number of the magnitude frequency distribution index sampling data is... The sampling time window of the monitoring time series points The magnitude frequency distribution index, The magnitude is measured on the Richter scale. The earthquake frequency distribution index early warning indicator tag time series data corresponding to the time series number is The magnitude frequency distribution index early warning indicator tag for the monitoring time series points, The magnitude frequency distribution index is the early warning threshold.

6. The method according to any one of claims 1 to 5, characterized in that: The sampling time window of the monitoring time point is the sampling time end point.

7. A rock mass instability stage early warning system, characterized in that, The system includes various functional modules required to implement the method of any one of claims 1 to 5.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the method of any one of claims 1 to 5.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 5.

Images