Coal mine rock burst dynamic early warning method and device based on microseism

Abstract

The invention provides a coal mine rock burst dynamic early warning method and device based on microseism. According to the method, dynamic prediction and real-time early warning of the rock burst risk can be realized, the adaptability of the early warning model to the mining process and the geological condition change is improved, the prediction hysteresis is effectively reduced, and the early warning accuracy is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of dynamic pre-warning of rock burst in coal mines, and in particular to a dynamic pre-warning method and system of rock burst in coal mines based on microseism. BACKGROUND

[0002] As an important pillar of energy supply, the safety monitoring technology in the mining process of coal mines is particularly crucial. With the popularization of deep mining and high-intensity mining operations, the frequency and destructiveness of rock burst disasters are increasingly prominent, becoming a core problem restricting the safety production of coal mines. In related technologies, a surrounding rock stability monitoring system is constructed through the collaborative work of displacement monitoring, acoustic emission monitoring and microseismic monitoring. Specifically, the system covers the whole process from geological condition evaluation, sensor deployment, signal acquisition, feature extraction to risk pre-warning, including key links such as data preprocessing, multi-source information fusion, model construction and feedback optimization. Among them, the microseismic monitoring technology, which can detect the internal micro-fracture activity of rock mass in real time, becomes an important support means for rock burst pre-warning, and is widely used in coal mine safety prevention and control under complex geological conditions such as high stress and deep mining.

[0003] However, in the existing microseismic pre-warning method, the static "one side one index" setting method is directly used, and there is no dynamic adjustment according to the mining progress, geological condition changes and microseismic signal evolution trend, which may lead to pre-warning lag or misjudgment, or lack of adaptability between different working faces, thereby affecting the real-time and accuracy of the pre-warning system. Specifically, the traditional method has a low update frequency of the pre-warning index during mining, which is difficult to reflect the instantaneous changes of the stress field inside the rock mass and the burst characteristics of the microseismic events, especially in the aspects of multi-dimensional information fusion such as the location of the seismic source, energy release and frequency spectrum distribution. In addition, due to the fact that the microseismic signal is easily disturbed by noise and different sensor data have dimension differences, the existing technology also faces certain challenges in data standardization and feature extraction, which limits the generalization ability and prediction accuracy of the pre-warning model. SUMMARY

[0004] The present application aims to at least solve one of the technical problems in the related art to some extent.

[0005] To this end, a first object of the present application is to propose a dynamic pre-warning method of rock burst in coal mines based on microseism.

[0006] A second object of the present application is to propose a dynamic pre-warning device of rock burst in coal mines based on microseism.

[0007] To achieve the above object, the first aspect of the present application provides a coal mine rock burst dynamic early warning method based on microseism, comprising: S1, according to the geological conditions of the mine, the mining method and progress, and the occurrence of historical rock burst, dynamically arranging microseismic sensors and collecting microseismic signals in real time;

[0008] S2, the collected microseismic signals are processed and analyzed in real time, and multi-dimensional feature information related to rock burst incubation is extracted, and a comprehensive monitoring and early warning index with high sensitivity is selected;

[0009] S3, based on the reliability theory and standardization method, the early warning index is uniformly processed, the membership function of the early warning level is established combined with the Gaussian membership function, and different coefficients are given according to the early warning level to calculate the weight of each early warning index;

[0010] S4, a rock burst microseismic multi-dimensional information fusion monitoring and early warning model is constructed, the model is dynamically adjusted according to the weight, and the real-time prediction of the rock burst occurrence possibility and the near time is realized.

[0011] In an embodiment of the present application, the S1 comprises:

[0012] S11, according to the mining area, depth, surrounding rock type and occurrence of historical rock burst of the mine, the representative monitoring points and high-risk areas are selected to arrange the microseismic sensors, and a network system is formed to monitor and collect the microseismic signals;

[0013] S12, the microseismic sensors are installed at the preset positions in the mine roadway, and the system is debugged, and the microseismic sensors are calibrated and maintained regularly.

[0014] In an embodiment of the present application, the S2 comprises:

[0015] S21, the original microseismic signal is filtered by using the Gaussian filter algorithm, the interference signal is removed, and the data of different sensors is uniformly standardized;

[0016] S22, the occurrence frequency of microseismic events is statistically analyzed, irrelevant or irrelevant small amplitude vibration is removed, and frequent microseismic signals are often precursors of rock burst; the spatial position of the source is determined by three-dimensional positioning technology, and whether the source is close to the mining area or there is a stress concentration area is evaluated.

[0017] In an embodiment of the present application, the S3 comprises:

[0018] S31, according to the reliability theory, the early warning index is brought into the exponential distribution function, and the expression of each standard early warning index is obtained;

[0019] S32, the data of the early warning index is standardized by using a standardization method, a maximum-minimum standardization formula is used for a high-value abnormal index, and a reverse maximum-minimum standardization formula is used for a low-value abnormal index.

[0020] In one embodiment of the present application, further comprising:

[0021] S5, a qualitative evaluation of the early warning index is quantitatively converted based on a comprehensive evaluation using fuzzy mathematics theory, an index set and an impact danger evaluation vector are constructed, a fuzzy comprehensive evaluation vector is calculated, and an impact danger level is determined through the maximum membership principle.

[0022] To achieve the above purpose, the second aspect of the present application provides a coal mine rock burst dynamic early warning device based on microseism, comprising:

[0023] The sensor arrangement module is used for dynamically arranging microseismic sensors and collecting microseismic signals in real time according to the geological conditions of the mine, the mining method and progress, and the occurrence of historical rock bursts;

[0024] The signal processing and feature extraction module is used for real-time processing and analysis of the collected microseismic signals, extraction of multi-dimensional feature information related to rock burst incubation, and screening of high-sensitivity comprehensive monitoring and early warning indicators;

[0025] The weight calculation and membership modeling module is used for unified processing of the early warning indicators based on reliability theory and standardization method, establishment of the membership function of the early warning level combined with the Gaussian membership function, and calculation of the weight of each early warning indicator according to different coefficients of the early warning level;

[0026] The multi-dimensional information fusion early warning module is used for constructing a rock burst microseismic multi-dimensional information fusion monitoring and early warning model, dynamically adjusting the model according to the weight, and realizing real-time prediction of the possibility and near time of rock burst occurrence.

[0027] The method and device of the present application can realize dynamic real-time early warning of coal mine rock burst, effectively overcome the problems of poor adaptability and slow response of traditional static early warning indicators, and improve the early warning precision and system adaptive ability.

[0028] Additional aspects and advantages of the present application will be in part apparent and in part pointed out hereinafter. BRIEF DESCRIPTION OF DRAWINGS

[0029] The above and / or additional aspects and advantages of the present application will become apparent and be readily understood from the following description, taken in conjunction with the accompanying drawings, in which:

[0030] Figure 1A flowchart of a coal mine rock burst dynamic early warning method based on microseism provided by an embodiment of the present application is shown in FIG. 1.

[0031] Figure 2 A flowchart of another coal mine rock burst dynamic early warning method based on microseism provided by an embodiment of the present application is shown in FIG. 2.

[0032] Figure 3 A structure diagram of a coal mine rock burst dynamic early warning device based on microseism provided by an embodiment of the present application is shown in FIG. 3. DETAILED DESCRIPTION

[0033] It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0034] In order to enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should fall within the scope of protection of the present application.

[0035] A coal mine rock burst dynamic early warning method and device based on microseism according to an embodiment of the present application will be described below with reference to the accompanying drawings.

[0036] Embodiment 1

[0037] Figure 1 A flowchart of a coal mine rock burst dynamic early warning method based on microseism according to an embodiment of the present application is shown in FIG. 1, which includes: Figure 1

[0038] S1, according to the geological conditions of the mine, the mining method and progress, and the occurrence of historical rock burst, dynamically deploying microseismic sensors and collecting microseismic signals in real time.

[0039] Specifically, this step is a dynamic deployment of microseismic sensors based on multi-source geological information and real-time signal collection, which is the basic link of the entire rock burst dynamic early warning system. The technical implementation principle is based on the seismic wave emission characteristics of the rock mass micro-fracture process, combined with the mine geological structure, mining technology, working face advancing speed and spatial distribution characteristics of historical rock burst events, to realize the dynamic optimization configuration of the microseismic monitoring network.

[0040] ​In some implementations, the layout of the microseismic sensors needs to consider the mining area, mining depth, surrounding rock type (such as sandstone, shale, coal seam, etc.) and epicenter distribution of historical rock burst events of the mine. Preferably, the GIS (Geographic Information System) and geological modeling software (such as Surpac, MineSight) are used to analyze the three-dimensional geological model of the mine, and to identify potential stress concentration areas, fault zones, coal pillar edges and other high-risk areas. The layout density is usually 1 sensor per 50-100 meters of roadway, and the sensor spacing is not more than the error range of the seismic source positioning (generally ± 5 meters), to ensure the three-dimensional positioning accuracy.

[0041] The microseismic sensor usually uses a three-component accelerometer, with a sampling frequency of 1-10 kHz, a dynamic range of ± 2g, and a sensitivity of 100 mV / g to adapt to microseismic events of different intensities. The sensor is connected to the monitoring host through wired or wireless transmission (such as LoRa, ZigBee) to realize real-time data acquisition and transmission. The system supports remote control and parameter adjustment, and can dynamically migrate the sensor position according to the mining progress to ensure that the monitoring coverage always keeps pace with the working face advancement.

[0042] The technical value of this step is that through the dynamic layout strategy, the defects of the traditional static point layout that cannot adapt to the dynamic changes of mining are overcome, and the spatio-temporal resolution and representativeness of microseismic signal collection are improved, providing a high-quality and high-timeliness data basis for subsequent feature extraction, early warning index selection and risk assessment.

[0043] Further, S1 includes:

[0044] S11, according to the mining area, depth, surrounding rock type and occurrence of historical rock burst of the mine, selecting representative monitoring points and high-risk areas to layout the microseismic sensors to form a network system for monitoring and collecting the microseismic signals.

[0045] In some implementations, the layout of the microseismic sensors needs to consider the mining area, mining depth, surrounding rock type (such as sandstone, shale, coal seam, etc.) and spatial distribution and time characteristics of historical rock burst events. Through geological modeling and mining progress analysis, high-risk areas such as stress concentration areas, fault zones, coal pillar edges, goaf boundaries are identified, and representative monitoring points are laid out in these areas. The layout density of the monitoring points is usually adjusted according to the length of the roadway, the complexity of the geology and the frequency of rock burst history, and generally one sensor node is laid out every 50-100 meters in high-risk areas to form a three-dimensional spatial coverage microseismic monitoring network system.

[0046] Optionally, the microseismic sensor adopts a high-sensitivity three-component acceleration sensor, which has a frequency response range of 1-1000 Hz and a sampling rate of not less than 10 kHz, so as to effectively capture the high-frequency components of the microseismic signal. The sensor installation position should avoid strong noise sources (such as ventilation equipment, transportation systems, etc.), and be fixed in the stable rock mass of the roadway roof or side slope to reduce external interference.

[0047] Further, after the layout is completed, system debugging is required, including sensor signal synchronization calibration, data transmission channel testing, positioning system accuracy verification, etc. The regular calibration period is generally 15-30 days, to ensure the stability and data consistency of the long-term operation of the sensor.

[0048] This step plays a basic supporting role in the entire technical solution, providing high-quality and representative raw data for subsequent signal processing, feature extraction, and early warning model construction. By scientifically laying out the microseismic sensor, the identification ability of the rock burst precursor signal can be effectively improved, providing reliable data sources for dynamic early warning, thereby significantly improving the response speed and early warning accuracy of the mine safety monitoring system.

[0049] S12, installing the microseismic sensor at a predetermined position in the mine roadway and performing system debugging, regular calibration, and maintenance of the microseismic sensor.

[0050] Specifically, this step involves installing the microseismic sensor at a predetermined position in the mine roadway and performing system debugging, regular calibration, and maintenance, which is a key link for stable operation of the rock burst dynamic early warning system and quality assurance of data acquisition.

[0051] In some implementations, the layout of the microseismic sensor needs to be based on the geological structure, mining depth, mechanical properties of surrounding rock, and distribution characteristics of historical rock burst events of the mine, to select representative high-risk areas for deployment. The sensor is usually arranged in a distributed manner to form a three-dimensional monitoring network to ensure the spatial positioning accuracy of the microseismic event. During installation, the fixation method of the sensor should be considered, such as fixing it to the roof, side slope, or floor of the roadway using anchor bolts or special brackets, to reduce external interference and enhance signal capture capability. System debugging includes communication testing between the sensor and the data acquisition unit (DAQ), signal transmission stability verification, and data synchronization calibration, to ensure that all sensors work under a unified time reference with an error of within ±1 ms.

[0052] The sampling frequency of the sensor is generally set to 1000 Hz to 20000 Hz to capture the high-frequency components of microseismic events. The sensor sensitivity is usually 100 mV / g to 500 mV / g, and the dynamic range is between ±2g and ±10g. The calibration period is recommended once every 30 days, and the maintenance period is once every 90 days, including checking the sensor connection status, power supply stability, signal noise level (signal-to-noise ratio should be ≥15dB) and data transmission channel integrity. The calibration process requires the use of a standard vibration source (such as the vibration reference signal specified in ISO 10816-1) for frequency response testing to ensure that the linearity error of the sensor in the key frequency band (5Hz-500Hz) is less than 5%.

[0053] This step is suitable for deep coal mining environment, especially in the roadway with faults, fracture zones or high stress concentration areas. Through scientific layout and continuous maintenance, the system can realize real-time monitoring of microseismic events, provide high-quality data input for subsequent feature extraction and early warning model construction, and improve the accuracy and response efficiency of rock burst prediction.

[0054] This step effectively improves the spatial resolution and time synchronization accuracy of the microseismic monitoring system by optimizing the sensor layout strategy and strict system debugging and calibration process, providing a reliable data basis for the precursor identification of rock burst, and is the key prerequisite for realizing adaptive adjustment of dynamic early warning model and multi-dimensional information fusion.

[0055] S2, real-time processing and analysis of the collected microseismic signals, extracting multi-dimensional feature information related to rock burst incubation, and screening out high sensitivity comprehensive monitoring and early warning indicators.

[0056] Specifically, this step is suitable for real-time monitoring of key areas such as high stress areas, mining and excavation faces, and fault zones in coal mines. After collecting microseismic signals, the system processes them in edge computing nodes or ground servers, determines the spatial location of the seismic source using three-dimensional positioning technology (such as inversion algorithms based on the time difference between P-wave and S-wave arrival times), and performs spatial matching analysis with the mining boundary and stress field distribution in the GIS system to determine whether it is in a high-risk area.

[0057] In terms of technical effects, this step effectively improves the accuracy and real-time performance of rock burst early warning through multi-dimensional feature extraction and sensitivity screening. Compared with traditional static early warning indicators, this method can adjust the weight of feature parameters according to the mining dynamics, enhance the adaptability of the model to complex geological conditions, and realize early identification and dynamic early warning of rock burst precursors.

[0058] Further, S2 includes:

[0059] S21, using a Gaussian filtering algorithm to filter the original microseismic signal noise, remove interference signals, and unify the data from different sensors.

[0060] Specifically, this step involves using a Gaussian filtering algorithm to filter the original microseismic signal noise and unifying the data from different sensors, which is a key step in the data preprocessing of the dynamic early warning method. In some implementations, this step uses a Gaussian filter to smooth the collected microseismic signal to remove high-frequency noise and unrelated interference signals, thereby improving the accuracy of subsequent feature extraction and early warning analysis.

[0061] After noise filtering, the microseismic data collected by different sensors is further standardized. Due to differences in the range, sampling accuracy, and installation location of each sensor, the original data is not comparable.

[0062] This step is deployed in the coal mine underground microseismic monitoring system in practical applications and is suitable for multi-sensor collaborative monitoring scenarios in complex geological environments. Through this step, the signal-to-noise ratio of the microseismic signal can be effectively improved, the stability of feature extraction can be enhanced, and high-quality, unified format data input can be provided for the subsequent rock burst early warning model, thereby significantly improving the real-time performance and reliability of the early warning system.

[0063] S22, statistical analysis of the frequency of microseismic events, eliminating irrelevant or unrelated small vibrations, frequent microseismic signals are often precursors of rock burst

[0064] Specifically, this step is a statistical analysis and feature selection process based on microseismic signals, and its core is to identify microseismic precursor signals related to the rock burst incubation process through frequency statistics and signal selection. In some implementations, this step first preprocesses the collected microseismic signal, including Gaussian filtering denoising and Z-score standardization processing, to eliminate sensor noise and data bias between devices. Subsequently, the system statistically analyzes the frequency of microseismic events, usually using a time window sliding statistical method (such as a 10-minute, 1-hour, or 24-hour time window), calculates the number of microseismic events in a unit of time (i.e., microseismic activity S), and combines energy distribution (ΔF) and time information entropy Qt to evaluate the concentration and burstiness characteristics of microseismic activity.

[0065] In terms of parameter setting, the screening of microseismic events is usually based on a magnitude threshold, for example, signals with a magnitude less than 1.5 are considered as irrelevant or irrelevant small vibrations and are rejected. In addition, the system determines the spatial coordinates of the source by three-dimensional positioning technology (such as inversion algorithm based on the time difference between P-wave and S-wave arrival time), and combines GIS system and mining area boundary data for spatial overlap analysis to identify whether it is located in the stress concentration area or the mining influence area. In terms of technical effect, this step effectively improves the signal-to-noise ratio of microseismic data and the accuracy of feature extraction by rejecting invalid signals, providing a high-quality data basis for subsequent construction of multi-dimensional information fusion early warning model, thereby enhancing the real-time and reliability of rock burst early warning.

[0066] S3, based on reliability theory and standardization method, the early warning index is uniformly processed, the membership function of early warning level is established combined with Gaussian membership function, and different coefficients are given according to early warning level to calculate the weight of each early warning index.

[0067] Specifically, this step uniformly processes the early warning index based on reliability theory and standardization method, establishes the membership function of early warning level combined with Gaussian membership function, and finally calculates the weight of each early warning index by giving different coefficients to different early warning levels, which is a key link in constructing rock burst dynamic early warning model.

[0068] In terms of technical implementation, first, the reliability theory is used to model and analyze each microseismic early warning index. Specifically, each early warning index is regarded as a reliability variable, and its abnormal probability distribution is described by exponential distribution function, so as to obtain its failure distribution function in the statistical time window.

[0069] Second, the standardization method is used to process the dimension of each early warning index. For high-value abnormal index, the maximum and minimum standardization formula x min x max x min is adopted; for low-value abnormal index, the reverse standardization formula x max x max x min is adopted. Through this processing, different indexes are compared in a unified scale, which improves the interpretability and consistency of the model.

[0070] Further, Gaussian membership function is introduced to fuzzify the early warning level.

[0071] In application scenarios, this step is applicable to the dynamic early warning system of rock burst in coal mining process, especially in complex working conditions such as working face advancing and frequent fault activity. Through scientific index processing and weight distribution, the adaptability and prediction accuracy of the early warning model are improved.

[0072] The technical effect of this step is that by introducing reliability theory and fuzzy membership function, the quantitative processing and risk level division of multi-source microseismic data are realized, the scientificity and practicality of the early warning model are effectively improved by combining F-cose evaluation and weight calculation, and a solid foundation is provided for subsequent multi-dimensional information fusion and dynamic adjustment.

[0073] Further, S3 comprises:

[0074] S31, according to the reliability theory, the early warning index is brought into the exponential distribution function, and the expression of each early warning index standard is obtained.

[0075] Specifically, this step involves introducing the screened rock burst comprehensive monitoring and early warning index into the exponential distribution function based on the reliability theory to establish the standard expression of each early warning index. The core is to model the failure probability of microseismic events using exponential distribution, thereby quantifying the abnormal membership of each index in different time windows, and providing a mathematical basis for subsequent early warning level division and weight calculation.

[0076] In terms of technical implementation, according to the reliability theory, each early warning index is regarded as a random variable, and its failure time obeys exponential distribution. Exponential distribution has no memory, and is suitable for describing the burstness and unpredictability of microseismic events in the time domain.

[0077] In application scenarios, this step is applicable to the dynamic early warning system of rock burst in the process of coal mining, especially in key areas such as high stress concentration area, fault zone, and boundary of goaf. Through exponential distribution modeling, the abnormal evolution trend of microseismic activity can be effectively identified, and the real-time response capability of the early warning model can be improved.

[0078] The technical effect of this step is that the early warning index is mathematically modeled by the exponential distribution function, which can scientifically quantify the abnormal degree, provide a reliable basis for subsequent membership function construction and weight allocation, and thus enhance the adaptability and prediction accuracy of the early warning model.

[0079] S32, the early warning index is standardized by using a standardized method, and for high-value abnormal index, a Min-Max normalization formula is used, and for low-value abnormal index, a reverse Min-Max normalization formula is used.

[0080] Specifically, this step uses a standardized method to standardize the data of the early warning index, specifically including using a Min-Max normalization formula for high-value abnormal index and a reverse Min-Max normalization formula for low-value abnormal index. The technical implementation principle is based on data normalization, aiming to eliminate the dimensional differences between different indexes, and improve the fusion ability and recognition accuracy of the early warning model for multi-dimensional data.

[0081] In terms of parameter setting, the standardized window is usually set as a sliding time window (such as 1 hour, 24 hours or 72 hours) to adapt to the data changes in the dynamic mining environment. The real-time data stream processing mechanism needs to be combined in the standardization process to ensure the timeliness and consistency of data updating.

[0082] This step plays a key role in data preprocessing in the whole dynamic early warning system, providing a unified data basis for subsequent weight allocation based on Gaussian membership function and confusion matrix precision evaluation, thereby improving the robustness and generalization ability of the early warning model. Through standardization processing, the system can more accurately identify abnormal features in microseismic signals and realize real-time and dynamic assessment of rockburst risk.

[0083] S4, a rockburst microseismic multi-dimensional information fusion monitoring and early warning model is constructed, and the model is dynamically adjusted according to the weight, realizing real-time prediction of rockburst occurrence possibility and near time.

[0084] Specifically, in some implementations, this step is based on the fusion analysis of multi-source microseismic data, and a multi-dimensional information fusion technology is used to construct a dynamic early warning model. Specifically, the model integrates key feature parameters such as magnitude, energy, frequency, epicenter distribution, time information entropy, total fault area A(t), Z-map value, b value, etc. of microseismic events to form a multi-dimensional feature vector. By introducing fuzzy mathematical theory and reliability analysis method, the qualitative early warning index is converted into a quantifiable risk evaluation value, and combined with exponential distribution function and Gaussian membership function, a rockburst early warning model with adaptive ability is constructed.

[0085] Further, the model uses the MMDP (Maximum Membership Degree Principle) maximum membership degree principle and VFPR (Variable Fuzzy Pattern Recognition) variable fuzzy pattern recognition principle to weight and fuse multi-dimensional features, realizing real-time prediction of rockburst occurrence possibility. Among them, the weight of each early warning index is dynamically calculated through the confusion matrix precision evaluation method (Confusion Matrix) and F-cose performance index, ensuring that the model has good adaptability and prediction accuracy under different mining stages and geological conditions.

[0086] In terms of specific parameter setting, the weight coefficients of each early warning index in the model need to meet the normalization condition, i.e. the sum of all weights is 1. For example, if 8 early warning indexes are selected, the weight of each index ranges from 0 to 1, and. The model prediction result usually takes rockburst occurrence probability (P) and near time (T) as output, where the value range of P is 0-1, the unit of T is hour or day, and the evolution rate of microseismic activity is dynamically estimated.

[0087] The step is deployed in the monitoring system of the key area in the coal mine in practical application, online calculation and updating are carried out in combination with real-time microseismic data flow, dynamic and accurate rock burst early warning support is provided for mine safety. By adjusting the index weight, the model can adapt to the geological conditions and mining progress of different regions, and the response speed and prediction accuracy of the early warning system are significantly improved, so that the safety risk caused by rock burst is effectively reduced.

[0088] The microseismic-based dynamic early warning method for coal mine rock burst of the embodiment can realize dynamic and real-time early warning of the coal mine rock burst, improve the adaptability and accuracy of the early warning model, and effectively reduce the threat of rock burst to mine safety.

[0089] Further comprising:

[0090] S5, using comprehensive evaluation based on fuzzy mathematics theory to quantitatively transform the qualitative evaluation of the early warning index, constructing an index set and a rock burst danger evaluation vector, calculating a fuzzy comprehensive evaluation vector, and determining a rock burst danger level through the maximum membership degree principle.

[0091] In summary, the microseismic-based dynamic early warning method for preventing and treating rock burst in a coal mine provided by the present application comprises the following steps: comprehensively analyzing the geological conditions, mining methods and progress, and historical rock burst occurrence in a mine, arranging microseismic sensors in key areas in the mine, collecting microseismic signals in real time, and adjusting the positions of the microseismic sensors according to the dynamic changes of actual mining operations; real-time processing and analyzing the collected microseismic signals, extracting feature information related to rock burst incubation, screening out rock burst comprehensive monitoring and early warning indexes, and screening the sensitivity of the early warning indexes based on the key areas; uniformly processing the dimensions of each monitoring and early warning index according to the reliability theory and in combination with a standardization method, dividing the early warning levels of microseismic precursors based on the early warning indexes, and providing corresponding prevention and treatment measures; using a Gaussian function to establish the membership function related to the early warning levels, giving different coefficients in combination with the early warning levels through a confusion matrix precision evaluation method, and calculating the weights of each early warning index in the early warning model; constructing a rock burst microseismic multi-dimensional information fusion monitoring and early warning model, using the early warning model to assess the risk of rock burst, predicting the possibility and near time of rock burst occurrence, and realizing the evaluation of the early warning model for each key area by adjusting the index weight; when the precursors of rock burst occur, timely early warning is given to the mine managers, and emergency response measures are started, and the early warning model and the monitoring strategy are adjusted in a timely manner according to the actual situation and the feedback after the occurrence of rock burst.

[0092] Preferably, the representative monitoring points and high-risk areas are selected according to the mining area, depth, surrounding rock type and occurrence of historical rock burst of the mine, and the microseismic sensors are arranged to form a network system for monitoring and collecting the microseismic signals; the microseismic sensors are installed at the preset positions in the mine roadway and are subjected to system debugging, regular calibration and maintenance.

[0093] Preferably, the original microseismic signals are subjected to noise filtering using a Gaussian filtering algorithm to remove interference signals, and the data of different sensors are subjected to unified standardization processing; the occurrence frequency of microseismic events is statistically analyzed, and irrelevant or irrelevant small vibrations are removed, and frequent microseismic signals are often precursors of rock burst; the spatial position of the seismic source is determined through three-dimensional positioning technology to evaluate whether the seismic source is close to the mining area or there is a stress concentration area; the epicenter distribution of the microseismic events is analyzed to speculate whether there is a phenomenon of frequent occurrence of strong microseismic in a specific area.

[0094] The magnitude and energy of the microseismic event are estimated according to the vibration waveform to analyze whether the threshold value that may trigger rock burst is reached; the magnitude of each microseismic event is statistically analyzed to determine the frequency distribution of the microseismic signals by analyzing the spectral characteristics; cross analysis is performed using seismological knowledge, the selected seismological indicators are applied to the early warning indicators, and the characteristic information related to rock burst preparation is extracted based on the above analysis.

[0095] Preferably, the change degree of the early warning indicators is compared and analyzed by analyzing the mining process of the key area and the evolution trend of the early warning indicators; the early warning indicators with more sensitive low-value anomalies or more sensitive high-value anomalies are screened out to determine the universal indicators of rock burst in the key area.

[0096] Preferably, the early warning indicators are brought into an exponential distribution function according to the reliability theory to obtain the expression of the standard of each early warning indicator; the data of the early warning indicators are standardized by using a standardization method; and the indicator membership functions of different early warning levels are obtained by referring to a Gaussian membership function.

[0097] Preferably, the performance indicators based on the confusion matrix are used to evaluate each early warning indicator by using F-cose; different coefficients are given to different early warning levels to calculate the F-cose values and weights of each early warning indicator.

[0098] Preferably, the microseismic data of historical rock burst events is used to establish a relationship between magnitude and rock burst intensity; the qualitative evaluation of the early warning indicators is quantitatively converted based on fuzzy mathematics theory; the MMDP principle and VFPR principle are used to obtain a rock burst microseismic multi-dimensional information fusion monitoring and early warning model; and the risk of different prediction results is evaluated to assess the influence range and severity.

[0099] Preferably, the weight of each early warning indicator in the early warning model is recalculated according to the general indicators of rock burst in the key area determined above; the early warning model is reconstructed using the weight of each early warning indicator calculated above; and the newly constructed early warning model is applied to the evaluation process of the corresponding key area.

[0100] Preferably, when the early warning signal of rock burst is detected, the early warning information is automatically sent to the mine management personnel and the on-site operators through the mine communication system; the corresponding emergency response measures are started according to the early warning level; the microseismic data changes are continuously monitored, and the response strategy is adjusted in a timely manner; and an operable emergency plan is provided to help the management personnel quickly judge and take protective measures.

[0101] Preferably, the accuracy, timeliness and emergency response effect of each early warning event are evaluated to analyze the success and shortcomings of the early warning; the early warning effect and the actual situation of the mine are used as feedback information to optimize the early warning model, the risk evaluation process and the data analysis method; and the microseismic sensor layout, the data processing algorithm and the model parameters, and the emergency response process are adjusted according to the actual monitoring results and the emergency response effect.

[0102] Embodiment 2

[0103] As shown in Figure 2 Another microseismic-based dynamic early warning method for preventing and treating rock burst in coal mines according to an embodiment of the present application includes:

[0104] S101, based on the geological conditions of the mine, the mining method and progress, and the occurrence of historical rock burst, microseismic sensors are arranged in key areas in the mine to collect microseismic signals in real time, and the positions of the microseismic sensors are adjusted according to the dynamic changes of the actual mining operation;

[0105] S102, the microseismic signals collected above are processed and analyzed in real time to extract feature information related to the incubation of rock burst, screen out rock burst comprehensive monitoring and early warning indicators, and screen the sensitivity of the early warning indicators based on the key areas.

[0106] S103, according to the reliability theory and combined with the standardization method, the dimensions of each monitoring and early warning index are uniformly processed, the early warning levels of microseismic precursors are determined based on the early warning indexes, and the corresponding prevention measures are provided;

[0107] S104, a membership function related to the above early warning level is established using a Gaussian function, and the weights of each early warning index in the early warning model are calculated by combining the early warning levels with different coefficients through the confusion matrix precision evaluation method;

[0108] S105, a microseismic multi-dimensional information fusion monitoring and early warning model of rock burst is constructed, the risk of rock burst is evaluated using the early warning model, the possibility and near time of rock burst occurrence are predicted, and the evaluation of each key area by the early warning model is realized through fine-tuning of the index weight;

[0109] S106, when the precursor of rock burst occurs, timely warning is given to the mine management personnel, and emergency response measures are started, and the early warning model and monitoring strategy are adjusted in time according to the actual situation and feedback after the occurrence of rock burst.

[0110] Specifically, according to the mining area, depth, surrounding rock type and historical rock burst occurrence of the mine, representative monitoring points and high-risk areas are selected to lay microseismic sensors, forming a network system for monitoring and collecting microseismic signals; the microseismic sensors are installed at the preset positions in the mine roadway and are debugged, calibrated and maintained regularly.

[0111] The representative monitoring points refer to the monitoring positions that can effectively reflect the influence of geological conditions, surrounding rock, mining and other conditions in the mine on the safety and stability of the mine, and have certain representativeness and universality. The selection of these monitoring points is usually based on factors such as the mining area, depth, surrounding rock type, historical rock burst occurrence of the mine, etc. The purpose is to ensure that the monitoring work can cover potential high-risk areas and provide key data support to ensure the safety production of the mine.

[0112] The original microseismic signal is filtered using a Gaussian filter algorithm to remove interference signals, and the data of different sensors are uniformly standardized;

[0113] The Gaussian filter is realized by constructing a weight matrix (usually a convolution kernel). Each element of the convolution kernel corresponds to the weight value of the Gaussian function in space. The form of the Gaussian function is:

[0114]

[0115] Where σ is the standard deviation, which determines the width of the filter. The larger the standard deviation, the stronger the smoothing effect of the filter, and the more obvious the denoising effect.

[0116] The constructed Gaussian filter (convolution kernel) is convolved with the original microseismic signal. This step can effectively filter out high-frequency noise components in the signal, leaving only the main trend or low-frequency part of the signal.

[0117] Standardizing the data collected by different sensors is to eliminate the range differences of different devices, so that the data can be compared and analyzed under the same standard. Z-score standardization (standard deviation standardization): the signal of each sensor is normalized by mean and standard deviation, so that the mean of the data is 0 and the standard deviation is 1.

[0118]

[0119] Where X is the original signal data, μ is the mean of the signal, and σ is the standard deviation of the signal.

[0120] Statistical analysis of the frequency of microseismic events, elimination of irrelevant or irrelevant small amplitude vibration; through three-dimensional positioning technology to determine the spatial position of the source, to evaluate whether the source is close to the mining area or there is a stress concentration area;

[0121] Using three-dimensional positioning technology, the spatial position of the source is determined according to the arrival time of the seismic wave (P wave, S wave, etc.), and the spatial coordinates (longitude, latitude, depth) of the source are calculated to obtain the position data of the source. In complex geological conditions, the inversion algorithm is used to further optimize the source position, considering the influence of complex geological structure underground.

[0122] Using the results of seismic positioning and the spatial coordinates of the mining area, the spatial overlap analysis of the source and the mining area is carried out. The boundary and ore body distribution information of the mining area is usually provided by the geological exploration data or remote sensing data of the mining area, which can be compared with the source position and mining area data through GIS (Geographic Information System) technology, and the calculation of whether the source is located in the mining area or nearby. If the source is located in the mining area, further analysis of the relationship between the source and the mining depth, mining method (such as blasting, drilling, etc.).

[0123] Through the simulation of underground stress field (such as using numerical simulation tools such as finite element method, discrete element method, etc.), combined with the stress change in the mining process, the possible stress concentration area is identified. Stress concentration areas often appear in mining areas, fault intersections, deep ore veins, etc. These places are high-risk areas for earthquakes or microseisms. Based on the stress change, the stress test, microseismic activity distribution, etc. Data are used to evaluate the stress state of the underground rock mass and identify possible high stress areas or stress concentration areas.

[0124] According to the spatial position of the seismic source, the relationship with the mining area or stress concentration area, the impact of the seismic source on the safety of the mine or engineering is evaluated. The evaluation content may include the ground vibration intensity that the seismic source may cause, the influence on the mine structure, the risk of equipment damage, the safety of mine operations, etc.

[0125] The epicenter distribution of microseismic events is analyzed to speculate whether there is a phenomenon of frequent occurrence of strong microseisms in a specific area.

[0126] The specific area refers to the area where the phenomenon of frequent occurrence of strong microseisms occurs, and the position of strong microseisms can be determined by the following standards:

[0127] 1. Microseisms with a magnitude of 3.0 or above;

[0128] 2. Epicenter dense area and concentrated distribution of epicenter depth;

[0129] 3. Relationship with mining activities and geological structures (such as fault zones, fracture zones, etc.);

[0130] 4. Area where strong microseisms occur within a certain time period (such as large-scale blasting, mining, etc. in a certain working face of the mine).

[0131] The magnitude and energy of microseismic events are estimated according to the vibration waveform, and it is analyzed whether the threshold value that may trigger rock burst is reached. The magnitude of each microseismic event is statistically analyzed, the frequency distribution of the microseismic signal is determined by analyzing the spectral characteristics, and cross-analysis is performed using seismological knowledge. The selected seismological indicators are applied to the early warning indicators, and the characteristic information related to the incubation of rock burst is extracted based on the above analysis content.

[0132] The threshold value needs to be set according to the current mining area, depth, surrounding rock type, etc. of the mine, and the specific situation of the current mining operation. The specific range is determined according to the actual situation. The selected seismological indicators include: b value, total fault area A(t), lack of earthquake, Z-map value; The final extracted characteristic information includes: b value, A(b) value, total fault area A(t), algorithm complexity AC, lack of earthquake, Z-map value, microseismic activity degree S, microseismic activity scale ΔF, time information entropy Q t .

[0133] By analyzing the mining process in the key area and the evolution trend of the early warning indicators, the change degree of the early warning indicators is compared and analyzed; the early warning indicators with more sensitive low value anomalies or more sensitive high value anomalies are selected, and the general indicators of rock burst in the key area are determined.

[0134] The change degree will vary according to the mining position, depth, surrounding rock type, etc. of the key area, and the different performances of the current mining operation. The specific range needs to be determined according to the actual situation.

[0135] According to the reliability theory, the early warning index is brought into the exponential distribution function, and the expression of each early warning index standard is obtained.

[0136]

[0137] In the formula, F(t) is the failure distribution function, λ ij (t) represents the abnormal membership of each index in the statistical time window t, and the value range is 0-1, λ ij represents the jth value of the ith early warning index after standardization.

[0138] The data of the early warning index is standardized by using the standardization method;

[0139] For high-value abnormal index, the standardization formula is as follows:

[0140] λ ij (t) = (Q ij -Q imin ) / (Q imax -Q imin )

[0141] For low-value abnormal index, the standardization formula is as follows:

[0142] λ ij (t) = (Q imin -Q ij ) / (Q imax -Q imin )

[0143] In the formula, Q ij represents the jth value of the ith early warning index, Q imin represents the minimum value of the i early warning indexes, and Q imax represents the maximum value of the ith early warning index.

[0144] The index membership function of different early warning levels is obtained by referring to the Gaussian membership function.

[0145] The relationship between the membership function G(W ij ) and the early warning index function W ij is analyzed by referring to the Gaussian membership function, and the value range is 0-1, then the index membership function of each early warning level is as follows:

[0146] No impact risk level:

[0147]

[0148] Weak impact risk level:

[0149]

[0150] Medium impact danger level:

[0151]

[0152] Strong impact danger level:

[0153]

[0154] Based on the performance index of confusion matrix, each early warning index is evaluated by F-cose; different coefficients are given to different early warning levels, and the F-cose value and weight of each early warning index are calculated.

[0155] Based on the above early warning level risk, the weak, medium and strong impact dangers need to be considered at the same time, and the weak, medium and strong F-cose values corresponding to the weak, medium and strong impact dangers are represented by f r , f z and f q respectively. Different coefficients are given to the weak, medium and strong levels, and the early warning index F-cose value and weight are calculated using the following formula:

[0156]

[0157] In the formula, f r , f z and f q represent the weak, medium and strong level scores respectively, so the weight of the ith early warning index can be represented by the following formula:

[0158]

[0159] In the formula, F i represents the F-cose value of the ith early warning index, and ΣF i represents the sum of the F-cose values of each early warning index.

[0160] As can be seen from the above, the present application can real-time collect and analyze microseismic signals by arranging microseismic sensors in key areas and combining the geological conditions of the mine, the mining method and the historical rock burst data, and can identify the characteristic information related to rock burst in time; with the help of reliability theory and standardization method, the monitoring data are processed scientifically, the reasonable early warning level is demarcated, and the weight is allocated to the early warning model according to different monitoring indexes, so as to improve the accuracy and response speed of early warning.

[0161] Example 3

[0162] As shown in Figure 2 , the present embodiment is basically the same as the above-mentioned embodiments, and the difference lies in that the relationship between magnitude and rock burst intensity is established through the microseismic data of historical rock burst events; the qualitative evaluation of early warning indexes is quantitatively converted by using comprehensive evaluation based on fuzzy mathematics theory;

[0163] Firstly, the index set is constructed: W = {W1, W2, …, i, …, W m}, where m represents the number of early warning indexes, the impact risk evaluation vector V = {v1, v2, v3, v4} is constructed, corresponding to four risk levels of no, weak, medium and strong, and then the early warning evaluation vector matrix is constructed:

[0164]

[0165] In the formula, r i1 , r i2 , r i3 and r i4 represent the membership degree of the ith early warning index and the jth level in the judgment vector V, then the early warning index weight vector A = {a1, a2, …, a i , …, a m} is determined, the fuzzy comprehensive evaluation vector B = A·R = {b1, b2, b3, b4} is calculated, and finally the maximum membership degree principle evaluation method is used:

[0166]

[0167] In the formula, n V = 4, B max = max{b i}, B sceond = max j∈i {b j}.

[0168] The MMDP principle and the VFPR principle are used to obtain the microseismic multi-dimensional information fusion monitoring and early warning model of rock burst; the risk of different prediction results is evaluated, and the influence range and severity are evaluated.

[0169] The MMDP principle and the VFPR principle are combined together, the constructed early warning model is optimized and evaluated in multiple dimensions, the microseismic multi-dimensional information fusion monitoring and early warning model of rock burst is obtained, the early warning model is used for risk assessment of rock burst, and the possibility and near time of rock burst are predicted.

[0170] According to the general indexes of the key area rock burst determined above, the weights of each early warning index in the early warning model are recalculated; the early warning model is reconstructed by using the weights of each early warning index calculated above; the newly constructed early warning model is applied to the evaluation process of the corresponding key area.

[0171] When the early warning signal of rock burst is detected, the early warning information is automatically issued to the mine managers and the on-site operators through the mine communication system; the corresponding emergency response measures are started according to the early warning level; the microseismic data changes are continuously monitored, and the response strategy is timely adjusted.

[0172] The accuracy, timeliness and emergency response effect of each early warning event are evaluated, the success and shortcomings of the early warning are analyzed, the early warning effect and the actual situation of the mine are taken as feedback information, the early warning model, the risk assessment process and the data analysis method are optimized, and the microseismic sensor layout, the data processing algorithm and the model parameters, the emergency response process are adjusted according to the actual monitoring results and the emergency response effect.

[0173] As can be seen from the above, the present application can issue an alarm in the precursor stage of rock burst, provide sufficient early warning time for the mine managers, timely take emergency measures, and effectively reduce the threat of rock burst to the safety of the mine; the method has strong adaptability and flexibility, can be adjusted according to the actual situation of different working faces, and improves the accuracy and reliability of the early warning system.

[0174] Example 4

[0175] In order to realize the above-mentioned embodiments, as Figure 3 shown, the present embodiment also provides a coal mine rock burst dynamic early warning device 10 based on microseism, comprising:

[0176] A sensor layout module 100 is used for dynamically laying microseismic sensors and collecting microseismic signals in real time according to the geological conditions of the mine, the mining method and progress, and the occurrence of historical rock burst;

[0177] A signal processing and feature extraction module 200 is used for real-time processing and analysis of the collected microseismic signals, extraction of multi-dimensional feature information related to rock burst incubation, and screening of high-sensitivity comprehensive monitoring and early warning indicators;

[0178] A weight calculation and membership modeling module 300 is used for unified processing of the early warning indicators based on reliability theory and standardization method, establishment of membership function of early warning level combined with Gaussian membership function, and calculation of weight of each early warning indicator according to different coefficients of early warning level;

[0179] A multi-dimensional information fusion early warning module 400 is used for constructing a rock burst microseismic multi-dimensional information fusion monitoring and early warning model, dynamically adjusting the model according to the weight, and realizing real-time prediction of the possibility and approaching time of rock burst occurrence.

[0180] Further, the sensor layout module is also used for:

[0181] According to the mining area, depth, surrounding rock type and historical rock burst occurrence of the mine, representative monitoring points and high-risk areas are selected to arrange the microseismic sensor to form a network system for monitoring and collecting the microseismic signal.

[0182] The microseismic sensor is installed at a preset position in a mine roadway, and the system is debugged, calibrated and maintained regularly.

[0183] Further, the signal processing and feature extraction module is further used for:

[0184] The original microseismic signal is filtered by using a Gaussian filtering algorithm to remove interference signals, and the data of different sensors are uniformly standardized.

[0185] The occurrence frequency of the microseismic event is statistically analyzed, irrelevant or irrelevant small vibrations are removed, and frequent microseismic signals are often precursors of rock burst; the spatial position of the seismic source is determined by three-dimensional positioning technology, and whether the seismic source is close to the mining area or there is a stress concentration area is evaluated.

[0186] The weight calculation and membership modeling module is further used for:

[0187] According to the reliability theory, the early warning index is brought into the exponential distribution function to obtain the expression of each early warning index standard;

[0188] The early warning index is standardized by using a standardized method, and the maximum and minimum standardization formula is used for high-value abnormal index, and the reverse maximum and minimum standardization formula is used for low-value abnormal index.

[0189] Further, it further comprises:

[0190] The fuzzy comprehensive evaluation module is used for quantitative conversion of qualitative evaluation of the early warning index based on fuzzy mathematics theory, constructing an index set and an impact danger evaluation vector, calculating a fuzzy comprehensive evaluation vector, and determining an impact danger grade through a maximum membership principle.

[0191] The coal mine rock burst dynamic early warning device based on microseismic of the embodiment can realize dynamic and real-time early warning of the coal mine rock burst, improve the adaptability and accuracy of the early warning model, and effectively reduce the threat of rock burst to mine safety.

[0192] In the description of the application, the description of the terms "one embodiment", "some embodiments", "an example", "a specific example", or "some examples" means that the particular feature, structure, material, or characteristic being described is included in at least one embodiment or example of the application. The illustrative examples of the above-mentioned terms do not necessarily refer to the same embodiment or example, although they can. Furthermore, the particular features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples. Moreover, the use of the terms "first", "second" or the like does not imply either a chronological or an ordinal nature, but is merely used to distinguish one embodiment or example from another. Furthermore, the terms "comprises", "comprising", "includes", "including" and the like used herein are specifically intended to be construed as "open-ended" terms, i.e., the terms do not exclude additional, unrecited elements or method steps. It is also noted that the terms "comprises", "comprising", "includes", "including" and the like can be used interchangeably with "consisting of" or "consisting essentially of".

[0193] Furthermore, the terms "first", "second", or the like do not denote any quantity or order, but are used to distinguish different features. Thus, these terms are used interchangeably with "one", "another", or "one or other". Furthermore, the meaning of "a", "an", and "the" includes plural references unless the context clearly dictates otherwise.

Claims

1. A microseismic-based dynamic pre-warning method for rock burst in a coal mine, characterized in that, The method comprises the following steps: S1, according to the geological conditions of the mine, the mining method and progress, and the occurrence of historical rock burst, dynamically arranging microseismic sensors and collecting microseismic signals in real time; S2, real-time processing and analysis of the collected microseismic signals, extracting multi-dimensional feature information related to rock burst incubation, and screening out high-sensitivity comprehensive monitoring and early warning indicators; S3, based on reliability theory and standardization method, the early warning indicators are uniformly processed, the membership function of the early warning level is established combined with Gaussian membership function, and different coefficients are given according to the early warning level to calculate the weight of each early warning indicator; S4, constructing a rock burst microseismic multi-dimensional information fusion monitoring and early warning model, dynamically adjusting the model according to the weight, and realizing real-time prediction of the possibility and near time of rock burst occurrence.

2. The method of claim 1, wherein, The S1 comprises: S11, according to the mining area, depth, surrounding rock type and occurrence of historical rock burst of the mine, selecting representative monitoring points and high-risk areas to arrange the microseismic sensors, forming a network system for monitoring and collecting the microseismic signals; S12, installing the microseismic sensors at the preset positions in the mine roadway and performing system debugging, periodic calibration and maintenance of the microseismic sensors.

3. The method of claim 1, wherein, The S2 comprises: S21, using Gaussian filtering algorithm to filter noise of original microseismic signals, removing interference signals, and uniformly standardizing data of different sensors; S22, statistical analysis of the occurrence frequency of microseismic events, elimination of irrelevant or irrelevant small vibrations, and frequent microseismic signals are often precursors of rock burst; the spatial position of the source is determined by three-dimensional positioning technology, and whether the source is close to the mining area or there is a stress concentration area is evaluated.

4. The method of claim 1, wherein, The S3 comprises: S31, according to the reliability theory, the early warning indicators are brought into the exponential distribution function to obtain the expression of each early warning indicator standard; S32, using standardization method to process data standardization of the early warning indicators, using maximum and minimum standardization formula for high-value abnormal indicators, and using reverse maximum and minimum standardization formula for low-value abnormal indicators.

5. The method of claim 1, wherein, Further comprising: S5, based on fuzzy mathematics theory, using comprehensive evaluation to quantitatively transform the qualitative evaluation of the early warning indicators, constructing index set and rock burst risk judgment vector, calculating fuzzy comprehensive evaluation vector, and determining rock burst risk level through the maximum membership principle.

6. A microseismic-based dynamic pre-warning device for rock burst in coal mines, characterized in that, The method comprises the following steps: A sensor arrangement module is arranged for dynamically arranging microseismic sensors and collecting microseismic signals in real time according to the geological conditions of the mine, the mining method and progress, and the occurrence of historical rock burst; A signal processing and feature extraction module is arranged for real-time processing and analysis of the collected microseismic signals, extracting multi-dimensional feature information related to rock burst incubation, and screening out high-sensitivity comprehensive monitoring and early warning indicators; A weight calculation and membership modeling module is arranged for uniformly processing the early warning indicators based on reliability theory and standardization method, establishing the membership function of the early warning level combined with Gaussian membership function, and calculating the weight of each early warning indicator according to different coefficients given according to the early warning level; The multi-dimensional information fusion early warning module is used for constructing a microseismic multi-dimensional information fusion monitoring and early warning model of rock burst, dynamically adjusting the model according to the weights, and realizing real-time prediction of the rock burst occurrence possibility and the near time.

7. The apparatus of claim 6, wherein, The sensor arrangement module is further used for: According to the mining area, depth, surrounding rock type and historical rock burst occurrence of the mine, the representative monitoring points and high-risk areas are selected to arrange the microseismic sensors, a network system is formed, and the microseismic signals are monitored and collected; The microseismic sensors are installed at the preset positions in the mine roadway, and the system is debugged, regularly calibrated and maintained.

8. The apparatus of claim 6, wherein, The signal processing and feature extraction module is further used for: The original microseismic signal is filtered by using a Gaussian filtering algorithm to remove interference signals, and the data of different sensors are uniformly standardized; The occurrence frequency of the microseismic event is statistically analyzed, irrelevant or irrelevant small vibrations are removed, and frequent microseismic signals are often precursors of rock burst; the spatial position of the seismic source is determined by using a three-dimensional positioning technology, and whether the seismic source is close to the mining area or there is a stress concentration area is evaluated.

9. The apparatus of claim 6, wherein, The weight calculation and membership modeling module is further used for: According to the reliability theory, the early warning indexes are brought into an exponential distribution function to obtain an expression of each early warning index standard; The data of the early warning indexes are standardized by using a standardization method, a maximum and minimum standardization formula is used for high-value abnormal indexes, and an inverse maximum and minimum standardization formula is used for low-value abnormal indexes.

10. The apparatus of claim 6, wherein, Further comprising: The fuzzy comprehensive evaluation module is used for quantitatively converting the qualitative evaluation of the early warning indexes based on the fuzzy mathematics theory, constructing an index set and a rock burst risk judgment vector, calculating a fuzzy comprehensive evaluation vector, and determining a rock burst risk grade through a maximum membership principle.

Images