Fixed underground coal mine rock wall vibration monitoring and alarming device and method

Abstract

The invention provides a fixed underground coal mine rock wall vibration monitoring and alarming device and method, and belongs to the field of mine vibration monitoring and early warning. The problems that the mine environment is complex, most collected signals are noisy signals, the monitoring accuracy of a conventional linear fusion algorithm is poor, and full early warning cannot be performed under the condition that obvious crack propagation does not occur are solved. Based on a multi-sensor fusion technology of a linear frequency modulation continuous wave radar and an acoustic emission piezoelectric micro-vibration sensor, through a micro-vibration monitoring system and a ground receiving end, real-time monitoring is carried out on deformation displacement, micro-vibration frequency and energy and seismic source information of a rock mass in combination with multiple sensors; fusion processing is carried out on data of multiple sensors by means of a prediction filtering algorithm, interference of environmental factors on monitoring results is reduced, and efficient and comprehensive monitoring and early warning are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of mine vibration monitoring and early warning, and in particular to a fixed coal mine underground rock wall vibration monitoring and alarm device and method. Background Art

[0002] In the process of mining, rock wall instability is a common and dangerous phenomenon. Mining activities disturb the original geological balance underground, resulting in the redistribution of stress in the rock layer. This change may cause the rock wall to vibrate, cracks to expand, and even large-scale collapse, seriously threatening the personal safety of miners and causing property losses. Therefore, in mining activities, effective monitoring and early warning of rock wall instability are of great significance to ensure mine safety.

[0003] Among the monitoring methods of rock wall stability in mines, the main technical support includes stress, strain, displacement, microseismic and acoustic emission methods. Among them, traditional monitoring and early warning schemes such as stress, strain and displacement have serious deficiencies in dynamic characteristics, monitoring lags, and cannot track the entire evolution process of rock wall instability in real time and dynamically. Microseismic monitoring is suitable for the overall monitoring of the mining area, but the lower monitoring frequency domain makes it difficult to capture the high-frequency micro-vibration signals on the local working surface, which to a certain extent affects its effective early warning ability for local disasters. In comparison, acoustic emission monitoring has significant advantages in capturing micro-vibration signals and determining the source of earthquakes with its higher monitoring frequency domain. However, due to the complex environment under the mine, the complex rock wall vibration is equivalent to a mobile target whose motion state is difficult to estimate, and its motion model has uncertainty. The conventional linear fusion algorithm is not effective, resulting in the acoustic emission monitoring of the sound wave signal is mostly a noisy complex signal, and it may not provide sufficient early warning when there is no obvious crack expansion. Summary of the invention

[0004] The technical problems to be solved by the present invention are:

[0005] In order to solve the problem that the collected signals are mostly noisy signals due to the complex mine environment, the monitoring accuracy using conventional linear fusion algorithms is poor, and sufficient early warning cannot be provided when there is no obvious crack expansion.

[0006] The present invention adopts the following technical solutions to solve the above technical problems:

[0007] The present invention provides a fixed underground coal mine rock wall vibration monitoring and alarm device, comprising a fixed support seat, a microseismic monitoring system and a ground receiving end.

[0008] The fixed support seat includes a base and a connecting rod assembly, a triangular fixing anchor is provided below the base for fixing to the ground or a wall, a connecting rod assembly is provided above the base, the connecting rod assembly includes a vertical rod and a horizontal rod, the bottom end of the vertical rod is arranged at the center of the upper end of the base, the upper end of the vertical rod is connected to the horizontal rod through a rotating shaft, one end of the horizontal rod is connected to an anchor rod through a rotating shaft, the anchor rod is used to be inserted into the ground or a wall, the other end of the horizontal rod is connected to a pan-tilt stabilizer, and the pan-tilt stabilizer is connected to the shell of the radar in the microseismic monitoring system through a connecting arm;

[0009] The microseismic monitoring system includes a controller, a communication module, a signal processing module, a radar, a posture sensor and a microseismic sensor. The radar is provided with a posture sensor, which is connected to the controller, and the controller is connected to the gimbal stabilizer. When the posture sensor detects that the horizontal angle exceeds a threshold, the controller sends a signal to the controller, and the controller controls the gimbal stabilizer to drive the connecting arm to adjust the position of the radar to the initial position.

[0010] The communication module, signal processing module and power supply module are all arranged on the base. The communication module is used to transmit the vibration data collected by the radar and the microseismic sensor to the signal processing module. After the signal processing module completes the processing of the vibration data, the rock wall vibration information is packaged and transmitted to the ground receiving end to realize real-time communication with the ground receiving end; the signal processing module is used to denoise, smooth and normalize the cyclic ranging information obtained by radar monitoring and the internal vibration information of the rock wall measured by the microseismic sensor, and then use the Kalman filtering algorithm to fuse the data of multiple sensors, analyze the fused data to evaluate the vibration condition of the wall and provide early warning information.

[0011] Furthermore, the ground receiving end includes a display module and an alarm module. The display module is used to receive and display in real time the microseismic data processing results of the current time, and to display the alarm information provided by the alarm module; the alarm module is used to alarm when the current microseismic data exceeds the risk threshold or when the horizontal angle deviation is too large, and the alarm information is displayed in real time on the display module of the ground terminal.

[0012] Furthermore, it also includes a power module and a data storage module. The power module is used to supply power to the controller, communication module, signal processing module, radar, attitude sensor, microseismic sensor, display module and early warning module; the data storage module is used to upload the original data to the cloud for storage, and the storage content includes the angle data of the attitude sensor and the periodic vibration data of the monitored rock wall. The periodic vibration data of the monitored rock wall includes amplitude, frequency, phase and vibration waveform.

[0013] An alarm method of a fixed coal mine underground rock wall vibration monitoring and alarm device comprises the following steps:

[0014] S100, install radar and microseismic sensors to collect on-site monitoring data of the rock wall to be monitored;

[0015] S200, using the on-site monitoring data to analyze the response of special geological events, establish the threshold values ​​of microseismic displacement, microseismic frequency and microseismic energy of the rock wall 1;

[0016] S300, using finite element software to simulate rock mass overload, calculate the maximum allowable value of the measuring point data, and establish threshold 2;

[0017] S400, selecting the smaller value between threshold 1 and threshold 2 as the warning threshold of the measuring point;

[0018] S500, after wavelet preprocessing of the field monitoring data obtained in step S100, the optimal state estimation is obtained using the Kalman filter algorithm, and the abnormal changes of the microseismic activity of the rock wall are analyzed, and then the warning judgment logic is set in combination with the warning threshold set in step S400.

[0019] Furthermore, in step S500, it specifically includes:

[0020] S510, using an interactive multi-model based Kalman filtering algorithm to perform interactive multi-model Kalman filtering on the field monitoring data after wavelet preprocessing;

[0021] Assume the basic motion model and observation equation of the target to be:

[0022] x k+1 =A k+1 x k +B k+1 μ k+1 +w k+1 (1)

[0023] z k+1 =H k+1 x k+1 +v k+1 (2)

[0024] Among them, x k is the state vector of the system at time k, A k+1 is the state transfer matrix of the system at time k+1, μ k+1 is the state control vector at time k+1, B k+1 is the control variable matrix at time k+1; w k+1 is the noise of the control system at time k+1, which obeys Gaussian distribution: w k ~N(0,Q), Q is the covariance matrix of the system noise; z k+1 is the measurement vector at time k+1, H k+1v is the transformation matrix from the state vector to the measurement vector at time k+1; k+1 is the noise measured at time k+1, which obeys Gaussian distribution: w k ~N(0,R), R is the covariance matrix of the measurement noise;

[0025] The Kalman filter algorithm includes prediction and update, where:

[0026] Prediction process:

[0027]

[0028] Update process:

[0029]

[0030] in, and are the posterior state estimates at time k+1 and time k, respectively, which are the optimal estimates; represents the prior state estimate at time k, and predicts the state at time k+1 based on the optimal estimate at time k; P k+1 and P k are the posterior estimated covariances at time k+1 and time k respectively; is the prior estimated covariance at time k+1; K k+1 is the Kalman gain;

[0031] S520, using multiple Kalman filters for parallel processing, assuming that the vibration has n motion models, the transfer probability matrix is:

[0032]

[0033] Among them, p ij is the transition probability of the target from motion model i to motion model j;

[0034] S530, the interactive multi-model method includes input interaction, Kalman filtering, model probability update and output interaction, specifically including:

[0035] S531, interactive input, at time k-1, the optimal estimate of the state of model j is The estimated covariance matrix is set up For Model At time point k, the state value of the current time point is predicted to be a mixture of various models in a certain proportion. is the covariance matrix obtained from the mixture of each model;

[0036] set up and These are the initial conditions after the interaction of model j at time k:

[0037]

[0038] in, is the correlation coefficient from model i to j;

[0039]

[0040] in, is the correlation coefficient of model i;

[0041] S532, the Kalman filter of model j is used to filter and predict the target in multi-model interactive tracking. The input of model j at time k-1 is and The multi-model Kalman filtering process includes,

[0042] The first-order prediction equation of Kalman filter is:

[0043]

[0044] in, is the predicted value of the state at time k by model j based on the state at time k-1; Φ j is the j model state transfer matrix;

[0045] Prediction covariance matrix:

[0046]

[0047] in, is the state estimation covariance matrix at time k-1; Q j is the covariance matrix of the j model prediction process;

[0048] Kalman filter gain:

[0049]

[0050] Among them, H j is the j-model observation matrix, which maps the state space to the observation space; R j is the covariance matrix of the observation noise of model j;

[0051] The system state estimation and covariance estimation corresponding to the output of a single model filter at time k are:

[0052]

[0053] Among them, z k is the observed value at time k;

[0054] S533, model probability update, using maximum likelihood estimation to update the model, by calculating the similarity between the current model and the current target state to give the weight of the most suitable tracking model at present;

[0055] The most matching maximum likelihood function of model i at time k is:

[0056]

[0057] Where N is the dimension of the observation vector;

[0058]

[0059] The update probability of model j is:

[0060]

[0061] Where c is the normalization constant:

[0062]

[0063] S534, output interaction, the overall state estimation equation is:

[0064]

[0065] The total updated error covariance matrix is:

[0066]

[0067] Furthermore, in step S500, the wavelet transform preprocessing is a wavelet threshold denoising method, including:

[0068] Wavelet analysis, for the field monitoring data collected in step S100, wavelet basis functions and decomposition levels are selected to perform wavelet decomposition on the actual vibration signal to obtain wavelet coefficients of different scales, specifically including:

[0069] When the number of decomposition layers is N, the frequency band range is:

[0070] F S / 2 N+1 =f max / 2 N (twenty four)

[0071] Among them, F S is the sampling frequency, f max is the maximum signal frequency;

[0072] The wavelet decomposition is to decompose the original signal into signals of different frequency bands. The wavelet function formula is as follows, the frequency is located by parameter a, and the time is located by parameter b:

[0073]

[0074] in, is the wavelet function, which is the Symlets wavelet; x(τ) is the original signal;

[0075] The inverse wavelet transform is:

[0076]

[0077] in, C ψ is the constant of the wavelet function;

[0078] After wavelet transform, a threshold with adaptive adjustment performance is selected to process the noisy leakage acoustic emission signal. The threshold is:

[0079]

[0080] Wherein, σ is the standard deviation of noise; J is the number of decomposition layers; D is the signal length; ln represents the logarithmic function with e as the base;

[0081] The threshold function includes a soft threshold and a hard threshold:

[0082]

[0083] in, is the wavelet coefficient after threshold function processing, ω j,k is the original wavelet coefficient, and λ is the selected threshold.

[0084] Furthermore, in step S100, when using radar to collect on-site detection data, the mathematical model of the transmitted signal and the received signal in each frequency sweep cycle is:

[0085]

[0086] in,

[0087]

[0088] Where A(t) and B(t) are the amplitudes of the transmitted and received signals, respectively, f0 is the initial frequency of the transmitted signal, K is the modulation slope, is the initial phase, t is the time, a0 is the electromagnetic wave propagation speed, B is the signal bandwidth, T is the single frequency sweep cycle time, R0 is the initial distance between the radar (233) and the target, and x(t) is the vibration displacement of the target object;

[0089] The echo signal received by the radar is obtained through a low-noise amplifier to obtain a high-gain echo signal, and then mixed with the local oscillator signal and low-pass filtered to obtain the baseband signal:

[0090]

[0091] in,

[0092]

[0093] The baseband beat signal is simplified to:

[0094]

[0095] Where, C(t) is the amplitude of the baseband signal;

[0096] Assuming that the displacement x(t) of the object within a frequency sweep cycle is a constant, the baseband beat signal within the qth frequency sweep cycle is:

[0097]

[0098] in,

[0099]

[0100] Among them, C q is the amplitude of the baseband beat signal in the qth frequency sweep cycle, f c is the center carrier frequency, λ0 and λ c are the wavelengths corresponding to the initial and center carrier frequencies, respectively, and f B is the beat frequency, and x(qT) are the initial phase of the beat signal and the target vibration displacement time series in the qth frequency sweep cycle, respectively;

[0101] The variation of the vibration displacement time domain information of adjacent sweep frequency cycles is obtained as follows:

[0102]

[0103] Among them, Δx(t) is the change of vibration displacement, is the variation of the baseband signal phase information in adjacent frequency sweep cycles, θ is the angle between the target vibration direction and the radar beam line of sight;

[0104] The time domain information of target vibration displacement is inverted from the displacement variation.

[0105] Further, in step S100, when installing the radar, according to the on-site working conditions, a radar beacon is installed in the key monitoring area, and periodic scanning is performed to obtain a set of rock wall vibration displacement data; when installing the microseismic sensor, it is set on the bottom plate of the track lane and the coal seam of the transport lane, and a drill hole is set every 1m-5m according to the layout requirements. The layout requirements are:

[0106] ① In the weak rock burst danger zone, the hole spacing is between 4m and 5m;

[0107] ② In the medium rock burst danger area, the hole spacing is between 2m and 4m;

[0108] ③ In the strong rock burst danger zone, the hole spacing is between 1m and 2m;

[0109] Using the three-point measurement method, when a vibration event occurs, each measuring point obtains the distance from the sensor measuring point to the station after receiving the P wave and S wave; for one measuring point, the source is located on an underground hemispherical surface with the measuring point as the center and the distance between the source and the measuring point as the radius; the intersection of the two hemispherical surfaces is the source trajectory line, and its projection on the ground is the epicenter trajectory line. Using the third measuring point, another epicenter trajectory line is obtained, and the intersection of the two is the epicenter position.

[0110] Furthermore, in step S200, the special geological events include ground collapse, ground fissures, earthquakes and rock collapse, and special geological events with data records are selected, and the P wave and S wave of the seismic wave signal are processed by seismic wave processing software to determine the location, magnitude, released energy and seismic moment information of the mine earthquake; wherein, the released energy is estimated by the duration of the seismic phase, and the arrival time and amplitude information of the seismic wave signal are used to obtain the velocity model and attenuation structure of the target area through the inversion method, and the propagation characteristics of the seismic wave in the crust and the deformation displacement caused are obtained, and then the threshold 1 is determined.

[0111] Furthermore, in step S300, the mine environment is modeled in combination with the underground geological engineering map data of the coal mine, and the overload method is used to simulate the ultimate bearing capacity of the rock wall. The overload condition is continuously tried using K times the pressure until the model calculation does not converge, and the deformation displacement data and stress measurement point data of the rock wall at this moment are retained as the parameter value of threshold 2.

[0112] Compared with the prior art, the present invention has the following beneficial effects:

[0113] The present invention discloses a fixed underground rock wall vibration monitoring and alarm device and method for a coal mine. The microseismic monitoring system includes a pan-tilt stabilizer, an attitude sensor, a link component, and a monitoring system module. A microseismic detection data acquisition system is formed by a linear frequency modulation continuous wave radar and an acoustic emission piezoelectric microvibration sensor. The two sensors are subjected to data fusion processing through a predictive filtering algorithm, thereby realizing comprehensive monitoring of the external deformation displacement and internal microvibration of the rock wall, so that the vibration state of the rock wall can be more accurately evaluated. This method not only effectively eliminates the problem of underground time series data being disturbed by noise, but also greatly improves the overall performance of rock wall stability monitoring, and makes up for the limitations of traditional monitoring methods in data coverage. In addition, a threshold value obtained by analyzing the response to a specific geological event based on field measured data is compared with a threshold value obtained by simulating rock mass overload using finite element software, so as to finely adjust the final early warning standard. This method of combining actual observation data with theoretical models significantly improves the accuracy of prediction of the probability of potential rock wall instability and geological disasters.

[0114] The rock wall vibration monitoring device and monitoring method based on millimeter wave radar and piezoelectric microseismic sensors proposed in the present invention solve the current problem of monitoring the stability of rock walls. Through the microseismic monitoring system and the ground receiving end, the deformation displacement, microseismic frequency, energy and earthquake source of the rock mass are monitored in real time in combination with multiple sensors. The problem that a single sensor is only sensitive to microseismic responses within a specific frequency range and that traditional monitoring schemes are easily affected by environmental factors such as temperature, humidity and electromagnetic interference is overcome. High-precision vibration parameters are obtained, and data from different sensors are fused and processed with the help of a predictive filtering algorithm, which reduces the interference of environmental factors on the monitoring results. The storage data adopts a double insurance strategy of cloud storage and local storage to achieve efficient and comprehensive monitoring and early warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0115] Figure 1 It is a structural schematic diagram of a fixed coal mine underground rock wall vibration monitoring and alarm device in an embodiment of the present invention;

[0116] Figure 2 This is a structural block diagram of a fixed underground coal mine rock wall vibration monitoring and alarm device in an embodiment of the present invention;

[0117] Figure 3 This is a flow chart of a fixed coal mine underground rock wall vibration monitoring and alarm method according to an embodiment of the present invention;

[0118] Figure 4 A vibration model diagram in an embodiment of the present invention;

[0119] Figure 5 is a flow chart of an interactive multiple model filtering algorithm according to an embodiment of the present invention;

[0120] Figure 6 Schematic diagram of determining the epicenter position using the three-point positioning method in an embodiment of the present invention.

[0121] Description of reference numerals:

[0122] 11. Base; 12. Connecting rod assembly; 13. Rotating shaft; 14. Triangular fixing anchor; 21. Gimbal stabilizer; 22. Connecting arm; 231. Communication module; 232. Signal processing module; 233. Radar; 234. Power module; 235. Attitude sensor; 236. Microseismic sensor; 31. Display module; 32. Alarm module. DETAILED DESCRIPTION

[0123] In the description of the present invention, it should be noted that the terminology nouns in the various embodiments, such as "up", "down", "front", "back", "left", "right", etc., which indicate directions, are only for simplifying the description of the positional relationship based on the drawings in the specification, and do not mean that the referred elements and devices must be operated according to the specific directions and defined operations and methods and structures in the specification. Such directional nouns do not constitute a limitation on the present invention.

[0124] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0125] Specific implementation plan 1: Combine Figure 1 and Figure 2 As shown, the present invention provides a fixed coal mine underground rock wall vibration monitoring and alarm device, including a fixed support base, a microseismic monitoring system and a ground receiving end.

[0126] The fixed support seat includes a base 11 and a connecting rod assembly 12. A triangular fixing anchor 14 is provided below the base 11 for fixing to the ground or a wall. A connecting rod assembly 12 is provided above the base 11. The connecting rod assembly 12 includes a vertical rod and a horizontal rod. The bottom end of the vertical rod is arranged at the center of the upper end of the base 11. The upper end of the vertical rod is connected to the horizontal rod through a rotating shaft 13, so as to realize that the horizontal rod can swing at the connection point as the axis. One end of the horizontal rod is connected to an anchor rod through the rotating shaft 13. The anchor rod is used to be inserted into the ground or a wall for further fixing. The other end of the horizontal rod is connected to a gimbal stabilizer 21. The gimbal stabilizer 21 is connected to the shell of the radar 233 in the microseismic monitoring system through a connecting arm 22.

[0127] The microseismic monitoring system includes a controller, a communication module 231, a signal processing module 232, a radar 233, a power module 234, a posture sensor 235, a microseismic sensor 236 and a data storage module. The radar 233 is provided with a posture sensor 235, and the posture sensor 235 is connected to the controller through a cable. The controller is connected to the gimbal stabilizer 21, and is used to send a signal to the controller after the posture sensor 235 detects that the horizontal data deviates by more than 10°, and the controller controls the gimbal stabilizer 21 to drive the connecting arm 22 to adjust the position of the radar 233 to the initial position. The gimbal stabilizer is a DJI Osmo Mobile SE stabilizer.

[0128] The communication module 231, the signal processing module 232 and the power module 234 are all arranged on the base 11. The communication module 231 includes a 485 bus and a Lora submodule, and overcomes the failure problem of a single method through a communication method combining wired and wireless communication. The communication module 231 is used to transmit the vibration data collected by the radar 233 and the microseismic sensor 236 to the signal processing module 232. After the signal processing module 232 completes the processing of the vibration data, the rock wall vibration information is packaged and transmitted to the ground receiving end through the 485 bus and the Lora submodule, so as to realize real-time communication with the ground receiving end.

[0129] The signal processing module 232 is used to process the radar cyclic ranging information. Cyclic ranging means that the radar 233 works for 1 minute and rests for 30 seconds, and repeats this process. After the periodic data of the non-contact microseismic information obtained and the internal vibration information of the rock wall measured by the microseismic sensor 236 are de-noised, smoothed and normalized, the Kalman filter algorithm is used to fuse the multi-sensor data, and the fused data is analyzed to evaluate the vibration of the wall and provide necessary warning information.

[0130] The radar 233 is a linear frequency modulated continuous wave (LFMCW) radar, which has a center frequency of 81 GHz, a detection distance of 10 m, and a distance resolution of 3 mm. It provides basic data for determining the stability of the rock wall through distance perception and microseismic information extraction based on the LFMCW radar; the microseismic sensor 236 is an acoustic emission piezoelectric microseismic sensor, which has a measurement range of ±50 g, a sensitivity (25°C) of 100 mv / g (160 Hz), and an amplitude nonlinearity of ±1%. The sensor can capture tiny vibrations in seismic waves, including first arrival waves (P waves) and second arrival waves (S waves), thereby providing the necessary data basis for accurate vibration positioning;

[0131] The specific process of the periodic scanning includes: from the beginning of the current scan, the analysis of the deformation and microseismic data of the monitored rock wall and the traversal of the radar 233 image points, to the beginning of the next scan, this time interval is a cycle, and the time interval of each cycle is adjusted according to specific environmental factors. When it is detected that the stability data of the rock wall exceeds the set risk threshold and an alarm message is issued, the corresponding monitoring time interval should be shortened to carry out more intensive monitoring;

[0132] The data storage module is a 10T data mechanical hard disk, and uses a cloud storage service for backup, which is used to upload the original data to the cloud for storage, and is backed up and synchronized every week. The storage content is the angle data of the attitude sensor 235 and the periodic vibration data of the monitored rock wall, including amplitude, frequency, phase and vibration waveform;

[0133] The power module 234 uses a combination of an underground power line and a battery to supply power. When the external power line is powered off, it can immediately switch to the battery to ensure continuous operation of the system.

[0134] The ground receiving end includes a display module 31 and an alarm module 32. The display module is an LCD screen, which is used to receive and display in real time the microseismic data processing results of the current time, and display the alarm information provided by the alarm module 32; the alarm module 32 is used to alarm when the current microseismic data exceeds the risk threshold or when the horizontal angle deviation is too large. The alarm information will be displayed in real time on the display module 31 of the ground terminal through cables or wireless communication equipment.

[0135] Specific implementation plan 2: Combination Figures 1 to 6 As shown, the present invention provides a fixed coal mine underground rock wall vibration monitoring and alarm method, comprising the following steps:

[0136] S100, installing radar 233 and microseismic sensor 236, and collecting on-site data of the rock wall to be monitored, specifically including:

[0137] The installation of radar 233, according to the on-site working conditions, installs radar beacons in the key monitoring area. The radar beacon is an electronic device installed on a specific target, which can emit electromagnetic signals and work in conjunction with the radar system to achieve accurate detection and tracking of the target; and performs periodic scanning, because radar 233 reports all intensity thresholds in the irradiation area (radar 233 irradiates as a beam, the rock wall is a non-uniform flat surface, and several target points with close distance resolution can be obtained), a set of rock wall vibration displacement data can be obtained;

[0138] The microseismic sensor 236 is installed by placing a piezoelectric vibration sensor inside the rock mass to record the vibration signal of the rock mass within the scanning period of the radar 233, and calculate the number of periodic microseisms, referred to as the microseismic frequency, and the maximum energy of periodic microseismic events, referred to as the microseismic energy. The present invention adopts the classic three-point measurement method. When a vibration event occurs, after each measuring point receives the P wave and the S wave, since the speed difference between the P wave and the S wave is certain, the distance from the sensor measuring point to the earthquake source can be obtained according to the time difference between the P wave and the S wave by simple calculation. Combined with Figure 6 As shown in the figure, for one measuring point, the earthquake source is located on a hemispherical surface inside the rock mass with the measuring point as the center and the distance between the earthquake source and the measuring point as the radius; another hemispherical surface is obtained by using another measuring point, and the intersection of the two hemispherical surfaces is the earthquake source trajectory line, and its projection on the ground is the epicenter trajectory line. Then, another epicenter trajectory line is obtained by using a third measuring point, and the intersection of the two is the epicenter position;

[0139] When laying out the microseismic sensors 236, it is necessary to fully consider the geological conditions of the coal mine and set them on the bottom plate of the track lane and the coal seam of the transport lane; according to the layout requirements, a drill hole is set every 1-5m for installing the microseismic sensors 236. Based on the geological conditions and safety requirements of different rock burst danger zones, in order to ensure that the microseismic sensors 236 can effectively detect vibration activities, so as to provide timely warnings and take corresponding safety measures, the layout requirements of the sensors are as follows:

[0140] ① In the weak rock burst danger zone, the hole spacing should be between 4-5m;

[0141] ② In the medium rock burst danger zone, the hole spacing should be between 2-4m;

[0142] ③ In the high-impact rock pressure danger zone, the hole spacing should be between 1-2m;

[0143] In addition, the vertical depth of each borehole is at least 1m to ensure a more effective pressure relief effect and to effectively collect high-quality vibration data; when installing the microseismic sensor 236, it is necessary to ensure that the sensor can be stably placed at the bottom of the hole and ensure seamless fit; then prepare cement mortar and pour it into the hole to completely cover the microseismic sensor 236. After leaving it in the air for a period of time, wait for the cement mortar to gradually solidify, take out the installation tool, and then use cement to fill the hole to ensure the stability of the sensor;

[0144] S200, using the on-site monitoring data to analyze the response of special geological events, establish the threshold values ​​of microseismic displacement, microseismic frequency and microseismic energy of the rock wall 1;

[0145] The special geological events include ground collapse, ground fissures, earthquakes and rock collapse; Figure 4As shown, it is the acceptance of the vibration band under normal circumstances; in the present invention, seismic activities with detailed data records are selected, and the location, magnitude, released energy and seismic moment of the mine earthquake can be determined by processing and analyzing the P wave and S wave of the seismic wave signal, and the energy released by the entire event is estimated by the duration of the seismic phase; the propagation time of the P wave (seismic longitudinal wave) is marked by the SeismoSignal seismic wave processing software, and the signal spectrum is analyzed to summarize the occurrence frequency of microseismic events and the law of energy release; when seismic waves propagate in the crust, they will produce compression or shear stress on the surrounding rocks, thereby causing tiny deformation and displacement; using the arrival time and amplitude information of the seismic wave signal, the velocity model and attenuation structure of the target area can be obtained by the inversion method, and then the propagation characteristics of the seismic wave in the crust and the deformation and displacement caused by it can be obtained; the threshold 1 is determined according to the above data;

[0146] S300, using finite element software to simulate rock mass overload, calculate the maximum allowable value of the measuring point data, and establish threshold 2;

[0147] The finite element analysis software can be ABAQUS, which is combined with the underground geological engineering map data of the coal mine to carry out modeling, and the modeling is analyzed according to the actual materials; the overload method is used to simulate the ultimate load-bearing condition of the rock wall, and the overload condition is continuously tried using K times the pressure until the model calculation does not converge (that is, the rock wall is considered to be unstable), and the deformation displacement data and stress measurement point data of the rock wall at this moment are retained as the parameter value of the threshold 2;

[0148] S400, selecting a smaller value between threshold 1 and threshold 2 as the warning threshold of the measuring point, thereby establishing a final threshold of the measuring point;

[0149] S500, after wavelet preprocessing of the field monitoring data, the optimal state estimation is obtained by using the Kalman filter algorithm, and the abnormal changes of the microseismic activity of the rock wall are analyzed, and the early warning judgment logic is set in combination with the early warning threshold set in step S400, including:

[0150] S510, the Kalman filter algorithm is an interactive multi-model Kalman filter algorithm (IMM-KF) algorithm. The complex rock wall vibration is equivalent to a maneuvering target whose motion state is difficult to estimate. Its motion model has uncertainty, and the conventional linear fusion algorithm is not effective. Therefore, multiple motion models are required to match the various vibration modes of the rock wall, that is, interactive multi-model Kalman filtering is performed on the monitoring data after wavelet preprocessing;

[0151] Kalman filtering has great advantages in acquiring vibration data and estimating vibration signals. The basic motion model and observation equation of the target are:

[0152] x k+1 =Ak+1 x k +B k+1 μ k+1 +w k+1 (1)

[0153] z k+1 =H k+1 w k+1 +v k+1 (2)

[0154] Among them, x k is the state vector of the system at time k; A k+1 is the state transition matrix of the system at time k+1, which describes how the system state transfers from time k to time k+1; μ k+1 is the state control vector, B k+1 is the control variable matrix; w k+1 is the noise of the control system, which obeys Gaussian distribution: k ~N(0,Q), Q is the covariance matrix of the system noise; z k+1 is the measurement vector, H k+1 is the transformation matrix from state vector to measurement vector; v k+1 is the measured noise, which follows a Gaussian distribution: w k ~N(0,R), R is the covariance matrix of the measurement noise;

[0155] The steps of the Kalman filter algorithm are divided into two steps, namely prediction and update:

[0156] Prediction process:

[0157]

[0158] Update process:

[0159]

[0160]

[0161] in, and The posterior state estimates at time k+1 and time k are one of the filtering results, that is, the updated result, also called the optimal estimate (because it is impossible to know the true value, a hat is added to indicate the optimal estimate); represents the prior state estimate at time k, and predicts the state at time k+1 based on the optimal estimate at time k; P k+1 and P k The posterior estimated covariances at time k+1 and time k are respectively and The covariance of , indicating the uncertainty of the state), is one of the filtering results; K is the prior estimated covariance at time k+1 (the covariance of the crowd); k+1 is the Kalman gain;

[0162] S520, using multiple Kalman filters for parallel processing, each filter corresponds to a different state space model, different state space models describe different target motion modes, so each filter has different state estimation of the target; the method contains a model probability estimator, an input interactor, a state estimation fusion device and Kalman filters corresponding to multiple motion states, multiple motion models are used to match multiple motion modes of rock mass vibration, the transition probability between different models is calculated based on Markov chain, the input interactor calculates the interaction of each Kalman filter through the model probability and the transition probability between different models, assuming that there are n motion models for vibration, and the transition probability matrix is:

[0163]

[0164] Among them, p ij is the transition probability of the target from motion model i to motion model j;

[0165] S530, the steps of the IMM method are divided into four steps, namely input interaction, Kalman filtering, model probability update and output interaction.

[0166] S531, interactive input,

[0167] At time k-1, the optimal estimate of the state of model j is And the estimated covariance matrix is Assumptions For Model At time point k, it is used to predict the state value of the current time point from each model mixed in a certain proportion, as well as its inaccuracy, which also comes from the covariance matrix obtained by mixing each model

[0168] at the same time and is the initial condition after the interaction of model j at time k, calculated as follows:

[0169]

[0170] in, is the correlation coefficient from model i to j,

[0171]

[0172] in, is the correlation coefficient of model i;

[0173] S532, Kalman filter (model j),

[0174] Since the Kalman filter algorithm works well, it is usually used in multi-model interactive tracking to filter and predict the target. The input of the model j at time k-1 is and The tracking process is similar to the single model Kalman filter algorithm. The Kalman filter process of the IMM algorithm is as follows:

[0175] The first-order prediction equation of Kalman filter is:

[0176]

[0177] in, is the predicted value of the state at time k by model j based on the state at time k-1; Φ j is the j-model state transition matrix, which describes how the system transfers from the previous moment to the current moment;

[0178] Prediction covariance matrix:

[0179]

[0180] in, is the state estimation covariance matrix at time k-1, reflecting the uncertainty of the state estimation at time k-1; Q j is the covariance matrix of the j model prediction process;

[0181] Kalman filter gain:

[0182]

[0183] Among them, H j is the j-model observation matrix, which maps the state space to the observation space; R j is the covariance matrix of the observation noise of model j, which ignores the uncertainty in the observation data;

[0184] The system state estimation and covariance estimation corresponding to the output of a single model filter at time k are:

[0185]

[0186] Among them, z k is the observed value at time k;

[0187] S533, model probability update,

[0188] The maximum likelihood estimation is used to update the model, and the weight of the most suitable tracking model is given by calculating the similarity between the current model and the current target state;

[0189] The most matching maximum likelihood function of model i at time k is:

[0190]

[0191] Where N is the dimension of the observation vector;

[0192]

[0193]

[0194] The update probability of model j is:

[0195]

[0196] Where c is the normalization constant:

[0197]

[0198] S534, output interaction,

[0199] The overall state estimation equation is:

[0200]

[0201] The total updated error covariance matrix is:

[0202]

[0203] The other combinations and connection relationships of this embodiment are the same as those of the first embodiment.

[0204] Specific implementation scheme three: Different from specific implementation scheme two, the wavelet transform preprocessing is a wavelet threshold denoising method, which is used to optimize the monitoring data so as to provide higher quality data input for subsequent Kalman filtering and improve filtering performance. The steps of the wavelet threshold denoising method are:

[0205] Wavelet analysis: receiving the original monitoring data, selecting the appropriate wavelet basis function and decomposition layer number to perform wavelet decomposition on the actual vibration signal, and obtaining wavelet coefficients of different scales;

[0206] The larger the number of decomposition layers, the more obvious the difference between signal and noise, which is more conducive to signal-noise decomposition, but at the same time the reconstructed signal will be more distorted, affecting the denoising effect; the frequency band range of wavelet decomposition is related to the sampling frequency. When the number of decomposition layers is M, the frequency band range is:

[0207] F S / 2 M+1 =f max / 2 M (twenty four)

[0208] Among them, F S is the sampling frequency, f max is the maximum signal frequency;

[0209] The wavelet decomposition is a process of wavelet change, that is, decomposing the original signal into signals of different frequency bands, which can realize the analysis of signals of different frequencies at different times; the wavelet function formula is as follows, the frequency is located by parameter a, and the time is located by parameter b:

[0210]

[0211] in, is a wavelet function, in which the wavelet function is a Symlets wavelet; x(τ) is the original signal;

[0212] The corresponding inverse wavelet transform is:

[0213]

[0214] in, C ψ is the constant of the wavelet function;

[0215] Threshold processing: After decomposition, the appropriate threshold is selected and the threshold function is used to quantize the wavelet coefficients of each layer. The coefficients greater than the threshold are retained and those less than the threshold are discarded, so as to suppress interference noise.

[0216] The threshold processing, after wavelet transform, the wavelet coefficient of the useful signal in the wavelet domain is greater than the wavelet coefficient of the noise signal. The noise can be set to 0 by selecting the threshold. However, if the threshold is too high, part of the useful signal will be deleted. If the threshold is too low, the noise processing is not comprehensive. Therefore, a threshold with adaptive adjustment performance should be selected to process the noisy leakage acoustic emission signal. The threshold selected in the present invention is:

[0217]

[0218] Wherein, σ is the standard deviation of noise; J is the number of decomposition layers; D is the signal length; ln represents the logarithmic function with e as the base;

[0219] Commonly used threshold functions are divided into soft threshold and hard threshold. Soft threshold denoising is to subtract the threshold from the wavelet coefficients whose absolute values ​​are greater than the "threshold" threshold, and set the wavelet coefficients whose absolute values ​​are less than the "threshold" threshold to 0, as shown in formula (28); hard threshold denoising is to retain the wavelet coefficients whose absolute values ​​are greater than the "threshold" threshold, and set the wavelet coefficients whose absolute values ​​are less than the "threshold" threshold to 0, as shown in formula (29):

[0220]

[0221] in, is the wavelet coefficient after threshold function processing, ω j,k is the original wavelet coefficient, and λ is the selected threshold.

[0222] Specific implementation plan 4: Different from the specific implementation plan 2, the core of the vibration monitoring technology based on LFMCW radar is to utilize the nonlinear modulation of electromagnetic wave signals caused by vibration, analyze the interference phase in different frequency sweep cycles, and reversely infer the vibration signal based on it to extract the time domain information of vibration displacement;

[0223] The present invention uses a sawtooth wave frequency modulated LFMCW radar for non-contact vibration sensing, which is simplified by using complex form. The mathematical model of the transmitted and received signals in each frequency sweep cycle can be defined as:

[0224]

[0225] in,

[0226]

[0227] Where A(t) and B(t) are the amplitudes of the transmitted and received signals, respectively, f0 is the initial frequency of the transmitted signal, K is the modulation slope, is the initial phase, t is the time, a0 is the electromagnetic wave propagation speed, B is the signal bandwidth, T is the single frequency sweep cycle time, R0 is the initial distance between the radar 233 and the target, x(t) is the vibration displacement of the target object, and j represents the sign of the imaginary part;

[0228] Millimeter-wave radar signals will attenuate and lose during the transmission process. The echo signal received by the receiving antenna is obtained through a low-noise amplifier to obtain a high-gain echo signal, and then mixed with the local oscillator signal and low-pass filtered to obtain the baseband signal:

[0229]

[0230] in, YesS R The conjugate number of (t);

[0231]

[0232] The square term can be ignored, and the baseband beat signal can be simplified as:

[0233]

[0234] Where C(t) is the amplitude of the baseband signal;

[0235] Since the frequency sweep cycle is often short, the change in vibration displacement within a single frequency sweep cycle T is very small. Therefore, it can be assumed that the displacement x(t) of the object within a frequency sweep cycle is a constant. Then, within the qth frequency sweep cycle, the baseband beat signal can be expressed as:

[0236]

[0237] in,

[0238]

[0239] Among them, C q is the amplitude of the baseband beat signal in the qth frequency sweep cycle, f c is the center carrier frequency, λ0 and λ c are the wavelengths corresponding to the initial and center carrier frequencies, respectively, and f B is the beat frequency, and x(qT) are the initial phase of the beat signal and the target vibration displacement time series in the qth frequency sweep cycle, respectively;

[0240] Through the above signal model representation and formula derivation, it can be seen that the interference phase information of the baseband signal is linearly related to the vibration displacement time domain information of the target object; therefore, by tracking the interference phase evolution of the baseband signal, the change in the vibration displacement time domain information of adjacent sweep cycles can be obtained as follows:

[0241]

[0242] Among them, Δx(t) is the change of vibration displacement, is the variation of the baseband signal phase information in adjacent sweep cycles, θ is the angle between the target vibration direction and the radar 233 beam line of sight, and then the target vibration displacement time domain information is inverted from the displacement variation.

[0243] Although the present invention is disclosed as above, the protection scope of the present invention is not limited thereto. Those skilled in the art may make various changes and modifications without departing from the spirit and scope of the present invention, and these changes and modifications will fall within the protection scope of the present invention.

Claims

1. A fixed coal mine underground rock wall vibration monitoring and alarm device, characterized in that: It includes a fixed support base, a microseismic monitoring system and a ground receiving terminal. The fixed support seat comprises a base (11) and a connecting rod assembly (12); a triangular fixing anchor (14) is provided below the base (11) for fixing to the ground or a wall; a connecting rod assembly (12) is provided above the base (11); the connecting rod assembly (12) comprises a vertical rod and a horizontal rod; the bottom end of the vertical rod is arranged at the center of the upper end of the base (11); the upper end of the vertical rod is connected to the horizontal rod via a rotating shaft (13); one end of the horizontal rod is connected to an anchor rod via the rotating shaft (13); the anchor rod is used to be inserted into the ground or a wall; the other end of the horizontal rod is connected to a pan-tilt stabilizer (21); the pan-tilt stabilizer (21) is connected to a shell of a radar (233) in a microseismic monitoring system via a connecting arm (22); The microseismic monitoring system comprises a controller, a communication module (231), a signal processing module (232), a radar (233), a posture sensor (235) and a microseismic sensor (236); the radar (233) is provided with a posture sensor (235); the posture sensor (235) is connected to the controller; the controller is connected to a gimbal stabilizer (21); and after the posture sensor (235) detects that a horizontal angle exceeds a threshold, the controller sends a signal to the controller, and the controller controls the gimbal stabilizer (21) to drive the connecting arm (22) to adjust the position of the radar (233) to an initial position. The communication module (231), the signal processing module (232) and the power supply module (234) are all arranged on the base (11); the communication module (231) is used to transmit the vibration data collected by the radar (233) and the microseismic sensor (236) to the signal processing module (232); after the signal processing module (232) completes the processing of the vibration data, the rock wall vibration information is packaged and transmitted to the ground receiving end, so as to realize real-time communication with the ground receiving end; the signal processing module (232) is used to perform denoising, smoothing and normalization processing on the cyclic ranging information monitored by the radar (233) and the internal vibration information of the rock wall measured by the microseismic sensor (236), and then use the Kalman filtering algorithm to perform data fusion on multiple sensors, analyze the fused data, so as to evaluate the vibration condition of the wall and provide early warning information.

2. A fixed coal mine underground rock wall vibration monitoring and alarm device according to claim 1, characterized in that: The ground receiving end comprises a display module (31) and an alarm module (32); the display module (31) is used to receive and display in real time the microseismic data processing result at the current time, and to display the alarm information provided by the alarm module (32); the alarm module (32) is used to alarm when the current microseismic data exceeds the risk threshold or when the horizontal angle deviation is too large, and the alarm information is displayed in real time on the display module (31) of the ground terminal.

3. A fixed coal mine underground rock wall vibration monitoring and alarm device according to claim 2, characterized in that: It also includes a power module (234) and a data storage module. The power module (234) is used to supply power to the controller, the communication module (231), the signal processing module (232), the radar (233), the attitude sensor (235), the microseismic sensor (236), the display module (31) and the early warning module. The data storage module is used to upload the original data to the cloud for storage. The storage content includes the angle data of the attitude sensor (235) and the periodic vibration data of the monitored rock wall. The periodic vibration data of the monitored rock wall includes the amplitude, frequency, phase and vibration waveform.

4. An alarm method according to any one of claims 1 to 3, characterized in that: The following steps are involved: S100, installing a radar (233) and a microseismic sensor (236) to collect on-site monitoring data of the rock wall to be monitored; S200, using the on-site monitoring data to analyze the response of special geological events, establish the threshold values ​​of microseismic displacement, microseismic frequency and microseismic energy of the rock wall 1; S300, using finite element software to simulate rock mass overload, calculate the maximum allowable value of the measuring point data, and establish threshold 2; S400, selecting the smaller value between threshold 1 and threshold 2 as the warning threshold of the measuring point; S500, after wavelet preprocessing of the field monitoring data obtained in step S100, the optimal state estimation is obtained using the Kalman filter algorithm, and the abnormal changes of the microseismic activity of the rock wall are analyzed, and then the warning judgment logic is set in combination with the warning threshold set in step S400.

5. A fixed coal mine underground rock wall vibration monitoring and alarm method according to claim 4, characterized in that: In step S500, specifically including: S510, using an interactive multi-model based Kalman filtering algorithm to perform interactive multi-model Kalman filtering on the field monitoring data after wavelet preprocessing; Assume the basic motion model and observation equation of the target to be: x k+1 =A k+1 x k +B k+1 μ k+1 +w k+1 (1) z k+1 =H k+1 x k+1 +v k+1 (2) Among them, x k is the state vector of the system at time k, A k+1 is the state transfer matrix of the system at time k+1, μ k+1 is the state control vector at time k+1, B k+1 is the control variable matrix at time k+1; w k+1 is the noise of the control system at time k+1, which obeys Gaussian distribution: w k ~N(0,Q), Q is the covariance matrix of the system noise; z k+1 is the measurement vector at time k+1, H k+1 v is the transformation matrix from the state vector to the measurement vector at time k+1; k+1 is the noise measured at time k+1, which obeys Gaussian distribution: w k ~N(0,R), R is the covariance matrix of the measurement noise; The Kalman filter algorithm includes prediction and update, where: Prediction process: Update process: in, and are the posterior state estimates at time k+1 and time k, respectively, which are the optimal estimates; represents the prior state estimate at time k, and predicts the state at time k+1 based on the optimal estimate at time k; P k+1 and P k are the posterior estimated covariances at time k+1 and time k respectively; is the prior estimated covariance at time k+1; K k+1 is the Kalman gain; S520, using multiple Kalman filters for parallel processing, assuming that the vibration has n motion models, the transfer probability matrix is: Among them, p ij is the transition probability of the target from motion model i to motion model j; S530, the interactive multi-model method includes input interaction, Kalman filtering, model probability update and output interaction, specifically including: S531, interactive input, at time k-1, the optimal estimate of the state of model j is The estimated covariance matrix is set up For Model At time point k, the state value of the current time point is predicted to be a mixture of various models in a certain proportion. is the covariance matrix obtained from the mixture of each model; set up and These are the initial conditions after the interaction of model j at time k: in, is the correlation coefficient from model i to j; in, is the correlation coefficient of model i; S532, the Kalman filter of model j is used to filter and predict the target in multi-model interactive tracking. The input of model j at time k-1 is respectively and The multi-model Kalman filtering process includes, The first-order prediction equation of Kalman filter is: in, is the predicted value of the state at time k by model j based on the state at time k-1; Φ j is the j model state transfer matrix; Forecast covariance matrix: in, is the state estimation covariance matrix at time k-1; Q j is the covariance matrix of the j model during prediction; Kalman filter gain: Among them, H j is the j-model observation matrix, which maps the state space to the observation space; R j is the covariance matrix of the observation noise of model j; The system state estimation and covariance estimation corresponding to the output of a single model filter at time k are: Among them, z k is the observed value at time k; S533, model probability update, using maximum likelihood estimation to update the model, by calculating the similarity between the current model and the current target state to give the weight of the most suitable tracking model at present; The most matching maximum likelihood function of model i at time k is: Where N is the dimension of the observation vector; The update probability of model j is: Where c is the normalization constant: S534, output interaction, the overall state estimation equation is: The total updated error covariance matrix is:

6. A fixed coal mine underground rock wall vibration monitoring and alarm method according to claim 5, characterized in that: In step S500, the wavelet transform preprocessing is a wavelet threshold denoising method, including: Wavelet analysis, for the field monitoring data collected in step S100, wavelet basis functions and decomposition levels are selected to perform wavelet decomposition on the actual vibration signal to obtain wavelet coefficients of different scales, specifically including: When the number of decomposition layers is N, the frequency band range is: F S / 2 N+1 =f max / 2 N (24) Among them, F S is the sampling frequency, f max is the maximum signal frequency; The wavelet decomposition is to decompose the original signal into signals of different frequency bands. The wavelet function formula is as follows, the frequency is located by parameter a, and the time is located by parameter b: in, is the wavelet function, which is the Symlets wavelet; x(τ) is the original signal; The inverse wavelet transform is: in, C ψ is the constant of the wavelet function; After wavelet transform, a threshold with adaptive adjustment performance is selected to process the noisy leakage acoustic emission signal. The threshold is: Wherein, σ is the standard deviation of noise; J is the number of decomposition layers; D is the signal length; ln represents the logarithmic function with e as the base; The threshold function includes a soft threshold and a hard threshold: in, is the wavelet coefficient after threshold function processing, ω j,k is the original wavelet coefficient, and λ is the selected threshold.

7. A fixed coal mine underground rock wall vibration monitoring and alarm method according to claim 6, characterized in that: In step S100, when using the radar (233) to collect on-site detection data, the mathematical model of the transmitted signal and the received signal in each frequency sweep cycle is: in, Where A(t) and B(t) are the amplitudes of the transmitted and received signals, respectively, f0 is the initial frequency of the transmitted signal, K is the modulation slope, is the initial phase, t is the time, a0 is the electromagnetic wave propagation speed, B is the signal bandwidth, T is the single frequency sweep cycle time, R0 is the initial distance between the radar (233) and the target, and x(t) is the vibration displacement of the target object; The echo signal received by the radar (233) is passed through a low noise amplifier to obtain a high-gain echo signal, and then mixed with the local oscillator signal and low-pass filtered to obtain a baseband signal: in, The baseband beat signal is simplified to: Where, C(t) is the amplitude of the baseband signal; Assuming that the displacement x(t) of the object within a frequency sweep cycle is a constant, the baseband beat signal within the qth frequency sweep cycle is: in, Among them, C q is the amplitude of the baseband beat signal in the qth frequency sweep cycle, f c is the center carrier frequency, λ0 and λ c are the wavelengths corresponding to the initial and center carrier frequencies, respectively, and f B is the beat frequency, and x(qT) are the initial phase of the beat signal and the target vibration displacement time series in the qth frequency sweep cycle, respectively; The variation of the vibration displacement time domain information of adjacent sweep frequency cycles is obtained as follows: Among them, Δx(t) is the change of vibration displacement, is the variation of the baseband signal phase information in adjacent frequency sweep cycles, θ is the angle between the target vibration direction and the radar (233) beam line of sight; The time domain information of target vibration displacement is inverted from the displacement variation.

8. A fixed coal mine underground rock wall vibration monitoring and alarm method according to claim 7, characterized in that: In step S100, when installing the radar (233), according to the on-site working conditions, a radar beacon is installed in the key monitoring area, and periodic scanning is performed to obtain a set of rock wall vibration displacement data; when installing the microseismic sensor (236), it is set on the bottom plate of the track lane and the coal seam of the transport lane, and a drill hole is set every 1m-5m according to the layout requirements. The layout requirements are: ① In the weak rock burst danger zone, the hole spacing is between 4m and 5m; ② In the medium rock burst danger area, the hole spacing is between 2m and 4m; ③ In the strong rock burst danger zone, the hole spacing is between 1m and 2m; Using the three-point measurement method, when a vibration event occurs, each measuring point obtains the distance from the sensor measuring point to the station after receiving the P wave and S wave; for one measuring point, the source is located on an underground hemispherical surface with the measuring point as the center and the distance between the source and the measuring point as the radius; the intersection of the two hemispherical surfaces is the source trajectory line, and its projection on the ground is the epicenter trajectory line. Using the third measuring point, another epicenter trajectory line is obtained, and the intersection of the two is the epicenter position.

9. A fixed coal mine underground rock wall vibration monitoring and alarm method according to claim 8, characterized in that: In step S200, the special geological events include ground collapse, ground fissures, earthquakes and rock collapse. Special geological events with data records are selected, and the P waves and S waves of the seismic wave signals are processed by seismic wave processing software to determine the location, magnitude, released energy and seismic moment information of the mine earthquake; wherein, the released energy is estimated by the duration of the seismic phase, and the arrival time and amplitude information of the seismic wave signal are used to obtain the velocity model and attenuation structure of the target area through the inversion method, and the propagation characteristics of the seismic wave in the crust and the deformation displacement caused are obtained, and then the threshold 1 is determined.

10. A fixed coal mine underground rock wall vibration monitoring and alarm method according to claim 9, characterized in that: In step S300, the mine environment is modeled in combination with the underground geological engineering map data of the coal mine, and the overload method is used to simulate the ultimate bearing capacity of the rock wall. The overload condition is continuously tried using K times the pressure until the model calculation does not converge. The deformation displacement data and stress measurement point data of the rock wall at this moment are retained as the parameter value of threshold 2.

Images