Underground rock wall fracture positioning system and method based on multiple distributed probes

Abstract

The invention discloses an underground rock wall fracture positioning system and method based on multiple distributed probes, and belongs to the technical field of underground exploration. The positioning system comprises an acquisition probe and a conveying mechanism, the acquisition probe is connected with an acquisition device capable of acquiring sound and images, and the acquisition probe is provided with a connection interface; the conveying mechanism comprises a conveying motor, the conveying motor is connected with a conveying shaft and a conveying wheel, the conveying wheel is connected with a propelling structure, the propelling structure is arranged in the axial direction of the conveying mechanism, the conveying mechanism is connected with a processing unit and a cable storage device, and the processing unit comprises a processing terminal and an acquisition circuit. The underground acoustic emission signal real-time monitoring system can be applied to coal mine safety detection, helps to detect underground parameters and observe the underground working environment, ensures that equipment is in a normal state or operates under the optimal condition, finally achieves high-precision, high-reliability and strong-anti-interference underground acoustic emission signal real-time monitoring, and provides key technical support for deep well exploration, disaster early warning and resource development.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of underground exploration, more particularly, to an underground rock wall fracture positioning system and method based on distributed multi-probe. BACKGROUND

[0002] The underground detection probe is an important support equipment for image monitoring in mine operation. Before formal operation, comprehensive and detailed monitoring of the bottom environment is an important prerequisite to ensure smooth operation and personnel safety. Process monitoring uses an underground detection probe to capture and analyze the geological structure, water level and potential safety hazards in real time, so as to scientifically assess whether the operation conditions are met.

[0003] The existing probe has high maintenance cost and signal interference in use. Most traditional underground probes lack automatic entry function and cannot observe the specific conditions of the underground probe in real time. The operation process is complex. For example, the patent with application number CN202321351360.7 discloses an underground exploration probe, which includes a detection probe, a counterweight pipe and a steel rope. The upper and lower ends of the counterweight pipe are closed, and the inside is filled with fillers. The processing circuit and the acquisition probe are packaged together, which has high maintenance cost. For example, the patent with application number CN202323094053.8 discloses a pulley conveying underground hole transient electromagnetic probe, which relies on the friction between the pulley and the hole wall foundation. The working environment is limited, the complexity is high, and the equipment maintenance cost is large. SUMMARY

[0004] The present application is aimed at the technical problems existing in the prior art, and provides an underground rock wall fracture positioning system and method based on distributed multi-probe.

[0005] To solve the above technical problems, the present application provides the following technical solutions: The present application first provides an underground rock wall fracture positioning system based on distributed multi-probe, which includes an acquisition probe and a conveying mechanism. The acquisition probe is connected with an acquisition device for sound and image acquisition. The acquisition probe is provided with a connection interface. The conveying mechanism includes a conveying motor, a conveying shaft and a conveying wheel connected to the conveying motor, a propulsion structure connected to the conveying wheel, the propulsion structure being arranged along the axial direction of the conveying mechanism, a processing unit and a cable storage device connected to the conveying mechanism, the processing unit including a processing terminal and an acquisition circuit, and the cable being connected to the connection interface through the cable storage device.

[0006] Further, the present application also provides an underground rock wall fracture positioning method based on distributed multi-probe, which utilizes the positioning system and includes the following steps: S100, drilling a collection hole corresponding to the collection probe in the plane to be detected, and placing the collection probe in the installation hole; S200, sending the collection probe to the collection position by the transmission mechanism, and extending the length of the collection probe by the circular tube and the connecting ring; S300, when the collection probe reaches the designated position, the video collection device starts to work and records the changes of the surrounding environment in the borehole in real time; at the same time, the three groups of two mutually perpendicular piezoelectric ceramics start to collect the acoustic emission signals, and transmit the collected signals to the collection circuit on the transmission mechanism by the cable, upload to the computer for data processing after being processed by the processing circuit; S400, eliminating the interference of mechanical vibration by Kalman filtering, and fusing the data collected by the multiple probes and locating the sound source by TDOA algorithm; S500, after the computer processes the data, the results of the processing and the screen collected by the video collection device are displayed on the host computer page, and the collected data is stored according to the date number; S600, after the collection probe completes the first collection, the collection circuit and the cable storage device are disassembled together with the collection probe, and the collection hole is filled.

[0007] Preferably, in step S300, the acoustic emission signals collected by the collection probe are transmitted to the collection circuit by the cable, wherein the collection circuit filters the collected acoustic emission signals by a low-pass filter, and then transmits the effective signal band to the processing unit.

[0008] Preferably, the transfer function of the low-pass filter used in the collection circuit can be represented as: H s

[0009] wherein, ω c = 2 πf c is the cutoff angular frequency; f c is the cutoff frequency, which is set according to the frequency band of the acoustic emission signal; Q is the quality factor, which is usually taken as to achieve the maximum flat response; s is the complex frequency domain variable; The relationship between the output signal y ( t ) of the filter and the input signal x ( t ) can be described by a differential equation: .​​

[0010] Preferably, before processing the data at the processing terminal, the acquired signals are synchronized in time using the generalized mutual GCC correlation algorithm, and the established three-channel signal model is as follows:

[0011] Where τk i is the signal arriving at the th ... k The latency of each channel, For noise, The source signal after time delay. i Indicates the first i One probe, i =1,2,…9; Cross-power spectrum based on the data acquired in the first channel:

[0012] in, for Fourier transform, for The complex conjugate of the Fourier transform; Choose phase transform weighting to sharpen peaks:

[0013] Weighted cross-power spectrum:

[0014] For the k One and z Weighted cross-correlation function of each channel for:

[0015] in, The kernel function of the inverse Fourier transform; The time delay difference is estimated by pairwise joint estimation of the three-channel acquired signals:

[0016] Alignment channel signals:

[0017] Output the aligned three-channel signal: .

[0018] Preferably, step S400 includes: S410, after the processing terminal receives the acoustic emission signal, the signal is preprocessed, the collected signal is extracted by using the improved MRCC algorithm, the time domain or frequency domain characteristics of the effective acoustic emission event are extracted, and the real sound source signal and noise are distinguished by characteristics; S420, aligning the multi-probe data by using the DTW algorithm, extracting the MFCC value of each channel, and calculating the 13-dimensional MFCC feature of the main channel of each probe; S430, positioning by using the TDOA algorithm, establishing the TDOA equation, assuming that the sound source position H s =( a , b , e ),probe position H i =( a i , b i , e i ),taking the first probe as the reference, and the time delay equation is:

[0019] wherein, c is the sound velocity, is the signal arrival time difference between the probes i and j ; The least square solution is: ; The matrix form is:

[0020] wherein, ; The least square solution is: ; S440, feature fusion is performed on the processed data, weight distribution is performed according to the signal-to-noise ratio of each probe, abnormal values are removed, multi-probe data is linearly combined according to the weight, decision-level fusion is performed, the initial position of the acoustic emission signal is output, and finally Kalman filtering is used for secondary optimization to smooth the positioning result.

[0021] Preferably, in step S410, it includes: the improved MRCC algorithm is used for three-layer wavelet packet decomposition on each aligned channel to obtain subband signals: , l is the number of subbands; The subband with an energy proportion greater than 5% is saved:

[0022] in, Indicates the first i The first probe k Through the sub-band l Energy on the surface No. i The probe k The first channel l Time-domain signals in sub-bands; For different probes, such as probes and For the corresponding channels, calculate the subband cross-correlation:

[0023] Weighted fusion subband results:

[0024] in, This is the optimal time offset. These are the sub-band weighting coefficients. For quality evaluation functions, Optimization objectives It has reached its maximum value.

[0025] Preferably, step S420 includes: extracting MFCC for each channel signal:

[0026] in, For Mel band energy, , c n No. n Number of MFCCs ; Calculate each probe and MFCC distance matrix And seek the optimal solution:

[0027] in, for Similarity distance measure between two points: ; Correcting the final latency:

[0028] in, For the initial alignment delay, f s The sampling frequency.

[0029] Preferably, in step S440, it comprises: in the calculation formula of the probe signal-to-noise ratio weight is:

[0030] Wherein, The first i The total number of probes, N The total number of probes, N =9 ，p The adjustment index is usually, p = 1 or 2, and the formula takes 2; The RANSAC algorithm is used again to remove outliers of the data collected by the probe, 3-4 probes are randomly selected to calculate the position of the initial sound source, the residual of the TDOA of other probes and the initial solution is counted, the abnormal probe with residual greater than the threshold is removed, and iterative optimization is carried out until convergence.

[0031] Preferably, in step S500, three areas are divided on the host computer page, which are real-time display waveform, video monitoring window and positioning window respectively; when saving files, hierarchical directories are established according to dates; the format of acoustic emission signal saving is CSV format, containing time stamp, energy, frequency, positioning coordinates, video files are attached with metadata, recording relevant acoustic emission time ID and acquisition parameters, and daily data is automatically compressed and backed up.

[0032] Compared with the prior art, the present application has the following beneficial effects: The embodiment of the present application adopts the combination of the conveying mechanism and the acquisition probe to realize the automatic advancement of the acquisition probe, the acquisition device of the acquisition probe is used to collect the acoustic emission signal, the acquisition device can capture sound and image information in real time, and then the information is transmitted to the acquisition circuit and the processing terminal through the cable, the position, waveform, real-time video and other information analysis of the acoustic emission signal are realized through the processing terminal, and the effective acquisition and analysis of the acoustic emission signal in the well are realized.

[0033] Moreover, the positioning method of the present application uses multiple acquisition probes in combination to collect the acoustic emission signal in the well and perform positioning, improves the positioning accuracy, has strong anti-interference ability, simultaneously uses the alignment of the three-channel signals in a single probe before the data processing of multiple probes to improve the accuracy and noise resistance, and optimizes the maintenance performance, separates the processing circuit from the acquisition probe, facilitates the maintenance of the subsequent circuit and reduces the cost. BRIEF DESCRIPTION OF DRAWINGS

[0034] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiments or prior art description will be briefly introduced. Obviously, the drawings in the following description only some embodiments of the present application, and for those skilled in the art, other drawings can be obtained without creative labor.

[0035] Figure 1 The structure schematic diagram of the positioning system in the embodiment of the present application; Figure 2 The structure schematic diagram of the transmission mechanism of the positioning system in the embodiment of the present application; Figure 3 The distribution schematic diagram of the acquisition probe of the positioning system in the embodiment of the present application Figure 4 The working flow schematic diagram of processing the acquired signal in the embodiment of the present application; Figure 5 The working flow schematic diagram of positioning the probe by using TDOA in the embodiment of the present application.

[0036] Explanation of the symbols in the drawings: 1, acquisition probe; 2, transmission mechanism; 3, connection interface; 4, transmission motor; 5, transmission shaft; 6, transmission wheel; 7, processing unit; 8, cable storage device; 9, processing terminal; 10, cable; 11, sound emission acquisition device; 12, image acquisition device; 13, support structure; 14, motor mounting box; 15, propelling circular tube; 16, circular tube groove; 17, circular tube connection port. DETAILED DESCRIPTION

[0037] In order to make the technical problems, technical solutions and beneficial effects of the present application more clearly, the following will further describe a kind of underground rock wall fracture positioning system and method based on distributed multi-probe provided by the present application in combination with drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application, and not used to limit the present application.

[0038] Embodiment 1 Please refer to Figure 1 The embodiment of the present application provides a kind of underground rock wall fracture positioning system based on distributed multi-probe, including acquisition probe 1 and transmission mechanism 2, acquisition probe 1 is connected with the acquisition device that can carry out sound and image acquisition, acquisition probe 1 is provided with connection interface 3; The conveying mechanism 1 comprises a conveying motor 4, a conveying shaft 5 and a conveying wheel 6 connected to the conveying motor 4, a pushing structure connected to the conveying wheel 6 and arranged along the axial direction of the conveying mechanism 2, a processing unit 7 and a cable storage device 8 connected to the conveying mechanism 2, the processing unit 7 comprising a processing terminal 9 and an acquisition circuit, and a cable 10 connected to the connecting interface 3 through the cable storage device 8.

[0039] The embodiment of the present application combines the conveying mechanism 2 with the acquisition probe 1 to realize automatic pushing of the acquisition probe 1, acquires the acoustic emission signal through the acquisition device of the acquisition probe 1, can capture sound and image information in real time, and then transmits the information to the acquisition circuit and the processing terminal 9 through the cable 10, realizes information analysis of the position, waveform and real-time video of the acoustic emission signal through the processing terminal 9, and realizes effective acquisition and analysis of the acoustic emission signal in the well.

[0040] Specifically, as shown in the figure, Figure 1 the acquisition probe 1 adopts a piezoelectric sensing probe, the acquisition device is arranged at the front end of the acquisition probe 1, the acquisition device comprises an acoustic emission acquisition device 11 and an image acquisition device 12, three groups of piezoelectric ceramics perpendicular to each other are arranged in the acoustic emission acquisition device 11, and the image acquisition device 12 is embedded in the edge of the acoustic emission acquisition device 11 and arranged in the same plane.

[0041] The inner end of the acquisition probe 1 is provided with a circumferential array arranged support structure 13, the support structure 13 is a support wheel, the support structure 13 plays a supporting role, when the acquisition probe 1 contacts with the acquisition pipeline, the support wheel rolls closely to the measured surface, can support the acquisition probe 1, and reduces the abrasion of the whole probe by the rock wall in the well.

[0042] The connecting interface 3 is arranged at the inner end of the acquisition probe 1, the cable storage device 8 is provided with an automatic cable feeding structure and a cable storage structure, the cable 10 is connected to the acquisition probe 1 through the connecting interface 3, and the other end of the cable 10 is connected to the automatic cable feeding structure on the conveying mechanism 2.

[0043] In the embodiment, as shown in the figure, Figure 1 , Figure 2 the conveying motor 4 is arranged in a motor mounting box 14, the end of the conveying motor 4 is connected to the conveying wheel 6 through the conveying shaft 5, the pushing structure of the conveying mechanism 2 is a pushing circular pipe 15, the pushing circular pipe 15 is arranged along the axial direction of the conveying mechanism 2, the pushing circular pipe 15 is arranged in a circular pipe groove 16, and the pushing circular pipe 15 is axially pushed in the circular pipe groove 16 under the cooperation of the conveying motor 4 and the conveying shaft 5 and the conveying wheel 6.

[0044] Further, the two ends of the advancing pipe 15 are provided with threaded connection structures, facilitating disassembly and installation, and can be assembled in multiple sections as needed to extend the detection length of the collection probe 1. The transmission mechanism 2 comprises a connecting ring, which is connected to the advancing pipe 15. The advancing pipe 15 and the connecting ring can extend the collection length of the collection probe.

[0045] Further, the inner end of the collection probe 1 is provided with a pipe connection port 17, and the collection probe 1 is connected to the end of the advancing pipe 15 of the transmission mechanism 2 through the pipe connection port 17.

[0046] In this embodiment, the collection circuit and the cable storage device 8 are connected to the transmission mechanism in a detachable manner, and the processing terminal 9 is a computer processing unit, which is connected to the collection circuit.

[0047] Further, as shown in Figure 3 , the probe distribution is shown in the figure. In this embodiment, nine collection probes 1 are arranged.

[0048] Embodiment 2 As shown in Figure 4 , Figure 5 , the present application provides a downhole rock wall fracture positioning method based on distributed multi-probe. The method is performed by using a positioning system and comprises the following steps: S100, using a drill bit to drill a collection hole corresponding to a collection probe in a plane to be detected, and placing the collection probe in the installation hole; S200, using a transmission mechanism to send the collection probe to a collection position, and using a pipe and a connecting ring to extend the length of the collection probe; S300, when the collection probe reaches the specified position, a video collection device starts to work to record the changes in the surrounding environment in the borehole in real time; at the same time, three groups of two mutually perpendicular piezoelectric ceramics start to collect acoustic emission signals, and the collected signals are transmitted to the collection circuit on the transmission mechanism through a cable, and then uploaded to a computer for data processing after being processed by a processing circuit; S400, using Kalman filtering to eliminate the interference of mechanical vibration, and fusing the data collected by multiple probes, and using a TDOA algorithm to locate the sound source; S500, after the computer finishes processing the data, the results of the processing and the screen collected by the video collection device are displayed on the host computer page, and the collected data is stored according to the date number; S600, after the collection probe completes the first collection, the detachable collection circuit and the cable storage device are detached from the transmission mechanism, so that the collection circuit and the cable storage device are left together with the collection probe, facilitating subsequent continuous collection and monitoring, and the collection hole is filled to avoid external interference.

[0049] The positioning method of this invention uses multiple sets of acquisition probes in combination to collect acoustic emission signals from the well and perform positioning, which improves positioning accuracy and has strong anti-interference ability. At the same time, the method of aligning the three-channel signals in a single probe before processing the data from multiple probes improves accuracy and noise resistance. Moreover, it optimizes maintenance performance by separating the processing circuit from the acquisition probe, which facilitates the maintenance of subsequent circuits and reduces costs.

[0050] Specifically, in step S300, the acoustic emission signal acquired by the acquisition probe is transmitted to the acquisition circuit via a cable. The acquisition circuit uses a low-pass filter to filter the acquired acoustic emission signal, which can eliminate high-frequency noise interference, retain the effective signal frequency band, and then transmit the effective signal frequency band to the computer processing unit.

[0051] Furthermore, the transfer function of the low-pass filter used in the acquisition circuit. H ( s This can be represented as:

[0052] in, ω c = 2 πf c The cutoff angular frequency; f c The cutoff frequency is set according to the acoustic emission signal frequency band; Q The quality factor is typically set to a value of [value missing]. To achieve the maximum flat response; s For complex frequency domain variables; The time-domain response is obtained through the inverse Laplace transform, and the filter's output signal is... y ( t ) and input signal x ( t The relationship between the two sides can be described by differential equations: .

[0053] Furthermore, before the processing terminal processes the data, the acquired signals are synchronized in time using the generalized mutual GCC correlation algorithm, and the established three-channel signal model is as follows:

[0054] Where τk i is the signal arriving at the th ... k The latency of each channel, For noise, The source signal after time delay. i Indicates the first i One probe, i =1,2,…9; Cross power spectrum based on the first channel acquisition:

[0055] wherein, is the Fourier transform of is the complex conjugate of the Fourier transform of Selecting phase transform weighting to sharpen the peak:

[0056] Weighted cross power spectrum:

[0057] For the first k and z channel weighted cross-correlation function is:

[0058] wherein, is the kernel function of the inverse Fourier transform; Estimating the time delay difference by pairwise joint estimation of three-channel acquisition signals:

[0059] Aligning channel signals:

[0060] Outputting the aligned three-channel signals: .

[0061] In this embodiment, step S400 comprises: S410, after the computer processing terminal receives the acoustic emission signal, the signal is preprocessed first, and then the improved MRCC algorithm is used to extract the features of the collected signal, extract the time domain or frequency domain features of the effective acoustic emission event, distinguish the real sound source signal from the noise through the features, and avoid false triggering of the positioning algorithm; S420, in order to solve the time delay fluctuation caused by the non-stationary signal, the DTW algorithm is used to align the multi-probe data, extract the MFCC value of each channel, and calculate the 13-dimensional MFCC feature of the main channel (channel 1) of each probe; S430, using TDOA algorithm positioning, establishing TDOA equation, assuming that the sound source position H s =( a , b , e ),probe position H i =(a i , b i , e i ), with the first probe as the reference, the time delay equation is:

[0062] wherein, c is the sound speed, is the signal arrival time difference between the probes i and j ; The least square solution is:

[0063] In matrix form:

[0064] wherein, ; The least square solution is: ; S440, the feature fusion is performed on the data after processing, the weight distribution is performed according to the signal-to-noise ratio (SNR) of each probe, the abnormal value is eliminated, the multi-probe data is linearly combined according to the weight, the initial position of the acoustic emission signal is output in the decision stage, and finally the Kalman filter is used for secondary optimization to smooth the positioning result.

[0065] Further, in step S410, it includes: the improved MRCC algorithm is used to perform three-layer wavelet packet decomposition on each aligned channel to obtain a subband signal: , is the number of subbands; The subband with an energy proportion greater than 5% is saved:

[0066] wherein, represents the energy of the i-th probe on the j-th channel in the k-th subband, is the time domain signal of the i-th probe on the j-th channel in the k-th subband; For different probes, such as the corresponding channels of probes and , the subband cross-correlation is calculated:

[0067] ​​​​​​Weighted fusion sub-band results:

[0068] wherein, is the optimal time offset, is the sub-band weight coefficient, is the quality evaluation function, the optimization objective makes reach the maximum value.

[0069] Further, in step S420, it includes: extracting MFCC for each channel signal:

[0070] wherein, is the Mel band energy, , c n The first n MFCC number ; Calculate the MFCC distance matrix of each probe and , and seek the optimal solution:

[0071] wherein, is the similarity distance measure value between two points: ; Correct the final time delay:

[0072] wherein, is the initial alignment time delay, f s is the sampling frequency.

[0073] Further, in step S440, it includes: in the calculation of the probe signal-to-noise ratio weight, the calculation formula is:

[0074] wherein, is the signal-to-noise ratio of the i probe, N is the total number of probes, N =9 ，p is the adjustment index, usually, p = 1 or 2, and the formula takes 2; ​The RANSAC algorithm is used for removing outliers from the data collected by the probe, 3-4 probes are randomly selected to calculate the position of the initial sound source, the residual of the TDOA of other probes and the initial solution is counted, the abnormal probe with a residual greater than a threshold is removed, and iterative optimization is performed until convergence.

[0075] In this embodiment, in step S500, three areas are divided on the host computer page, which are real-time waveform display, video monitoring window and positioning window respectively; when saving a file, a hierarchical directory is established according to date, facilitating subsequent retrieval.

[0076] Furthermore, the acoustic emission signal is saved in CSV format, containing key parameters such as timestamp, energy, frequency and positioning coordinates, and the video file is attached with metadata (JSON) recording the relevant acoustic emission time ID and collection parameters, and the daily data is automatically compressed and backed up.

[0077] The present application provides a high-precision underground rock wall fracture positioning system and method based on distributed multi-probe, which solves the core problems of low signal-to-noise ratio, insufficient dynamic range and poor environmental adaptability of existing underground acoustic emission monitoring technology. The positioning system is composed of three groups of piezoelectric ceramic sensors, collection devices, support wheels, transmission mechanisms, processing circuits and processing terminals, etc. The three groups of piezoelectric ceramic sensors are responsible for collecting the acoustic emission signals in the well, the video collection device can capture video information in real time, the transmission mechanism is combined with the collection probe to realize the automatic advancement of the collection probe, the acoustic emission signals are collected by the collection device of the collection probe, the collection device can capture sound and image information in real time, and then the information is transmitted to the collection circuit and the processing terminal through the cable. The processing unit first aligns the three-channel collection signals with GCC, and then processes the aligned signals. The processing terminal realizes the analysis of the position, waveform, real-time video and other information of the acoustic emission signals. The host computer page displays the position, waveform, real-time video and other information of the acoustic emission signals. Each unit cooperates to realize effective collection and analysis of the acoustic emission signals in the well.

[0078] The present application improves the precision and enhances the omnidirectional detection capability by arranging three groups of mutually perpendicular acoustic emission collection sensors in the collection probe, and the image collection device is fixed at the front end of the collection probe, which can not only realize data acquisition but also observe the internal situation of the mine to realize real-time recording. Moreover, the positioning method of the present application uses multiple collection probes jointly to collect the acoustic emission signals in the well and position them, which improves the positioning accuracy and has strong anti-interference ability. At the same time, the precision and anti-noise ability are improved by aligning the three-channel signals in a single probe first and then processing the data of multiple probes. Moreover, the maintenance performance is optimized by separating the processing circuit from the collection probe, which facilitates the maintenance of the subsequent circuit and reduces the cost.

[0079] The application can be applied to coal mine safety detection, helps to detect parameters and observe working environment of underground, ensures that equipment is in normal state or optimal condition operation, finally reaches high precision, high reliability, strong anti-interference real-time monitoring of acoustic emission signal of underground, provides key technical support for deep well exploration, disaster warning and resource development.

[0080] In the description of the present application, it should be understood that the orientation or positional relationship indicated by terms such as "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and are not intended to indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the present application.

[0081] In addition, the terms "first", "second" are only for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined with "first", "second" can explicitly or implicitly include one or more of the features. In the description of the present application, the meaning of "multiple" is two or more, unless otherwise specifically limited.

[0082] The above only describes the preferred embodiments of the present application and cannot be used to limit the present application, and any modifications, equivalent replacements and improvements made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A distributed multi-probe based borehole wall fracture location system, comprising: The collecting probe is connected with a collecting device capable of collecting sound and image, and a connecting interface is arranged on the collecting probe. The conveying mechanism comprises a conveying motor, a conveying shaft and a conveying wheel connected to the conveying motor, and a propelling structure connected to the conveying wheel and arranged along the axial direction of the conveying mechanism.

2. A method for locating a fracture in a borehole wall based on distributed multi-probe, characterized in that, The method comprises the following steps: S100, drilling a collecting hole corresponding to the collecting probe in a plane to be detected, and placing the collecting probe in the collecting hole; S200, conveying the collecting probe to a collecting position by using the conveying mechanism, and extending the length of the collecting probe by using a circular tube and a connecting ring; S300, when the collecting probe reaches the designated position, the video collecting device starts to work and records the changes of the surrounding environment in the borehole in real time; meanwhile, three groups of two mutually perpendicular piezoelectric ceramics start to collect acoustic emission signals, and the collected signals are transmitted to the collecting circuit on the conveying mechanism by using a cable, uploaded to a computer for data processing after being processed by a processing circuit; S400, eliminating the interference of mechanical vibration by using Kalman filtering, fusing the data collected by multiple probes, and locating the sound source by using a TDOA algorithm; S500, after the computer processes the data, displaying the processed results and the screen collected by the video collecting device on the host computer page, and storing the collected data according to the date number; S600, after the collecting probe completes the first collection, disassembling the collecting circuit and the cable storage device, leaving them together with the collecting probe, and filling the collecting hole.

3. The distributed multi-probe based borehole wall fracture location method of claim 2, wherein, In step S300, the acoustic emission signals collected by the collecting probe are transmitted to the collecting circuit by using a cable, wherein the collecting circuit filters the collected acoustic emission signals by using a low-pass filter, and then transmits the effective signal frequency band to the processing unit.

4. The distributed multi-probe based borehole wall fracture location method of claim 3, wherein, The transfer function of a low-pass filter used in the acquisition circuit H ( s ) can be represented as: wherein, ω c = 2 πf c is the cut-off angular frequency; f c is the cut-off frequency, set according to the frequency band of the acoustic emission signal; Q is the quality factor, typically taken as to achieve a maximum flat response; s is the complex frequency domain variable; The time-domain response is obtained through the inverse Laplace transform, and the filter's output signal is... y ( t ) and input signal x ( t The relationship between ) can be described by differential equations: 。 5. The distributed multi-probe based borehole wall fracture location method of claim 3, wherein, Before the data is processed by the processing terminal, the collected signals are time-synchronized by using a generalized mutual GCC correlation algorithm, and a three-channel signal model is established as follows: Where τk i is the signal arriving at the th ... k The latency of each channel, For noise, The source signal after time delay. i Indicates the first i One probe, i =1,2,…9; The mutual power spectrum based on the first channel collection as a reference: wherein is the Fourier transform of is the complex conjugate of the Fourier transform of The phase transformation weighting is selected to sharpen the peak value: ； The weighted mutual power spectrum: ； For the first k and z channel weighted cross-correlation function is: wherein kernel function of the inverse Fourier transform; The delay difference is estimated by two-by-two mutual estimation of the three-channel collected signals: ； The aligned channel signals: ； The three-channel signals after alignment are output: 。 6. The distributed multi-probe based borehole wall fracture location method of claim 2, wherein, Step S400 comprises: S410, after the processing terminal receives the acoustic emission signals, the signals are preprocessed, the collected signals are feature-extracted by using an improved MRCC algorithm, the time-domain or frequency-domain features of effective acoustic emission events are extracted, and the real sound source signals are distinguished from noises by using the features; S420, the data of multiple probes are aligned by using a DTW algorithm, the MFCC values of each channel are extracted, and 13-dimensional MFCC features of the main channel of each probe are calculated; S430, positioning by TDOA algorithm, establishing TDOA equation, setting sound source position H s =( a , b , e ), probe position H i =( a i , b i , e i ), taking the first probe as the reference, and the time delay equation is: wherein, c is the speed of sound, is the probe i the signal arrival time difference between j ​ Least square solution: ； Matrix form: wherein ; Least square solution: ； S440, the data after processing is fused, the weight is allocated according to the signal-to-noise ratio of each probe, and the abnormal value is removed, then the multi-probe data is linearly combined according to the weight, the initial position of the acoustic emission signal is output, finally the positioning result is smoothed by Kalman filtering secondary optimization.

7. The distributed multi-probe based borehole wall fracture location method of claim 6, wherein, In step S410, it includes: the improved MRCC algorithm is to each alignment channel Three-layer wavelet packet decomposition is carried out to obtain subband signals: , l is the number of subbands; Save the sub-band with energy ratio greater than 5%: wherein, represents the i th subband on the k th channel of the l th probe, the i th subband on the k th channel of the l th probe; For different probes, such as probes and corresponding channels, compute sub-band cross-correlations: ； The optimal time offset of the weighted fusion sub-band result is: wherein, is the optimal time offset, is the subband weight coefficient, is the quality evaluation function, the optimization objective is to maximize the maximum value.

8. The distributed multi-probe based borehole wall fracture location method of claim 6, wherein, In step S420, it includes: extracting MFCC for each channel signal: wherein, is the Mel band energy, , c n The number of MFCCs n ;​ Compute MFCC distance matrix for each probe and of the probes and seek the optimal solution: wherein is similarity distance measure value between two points: ； Correct the final time delay: wherein, is the initial alignment delay, f s is the sampling frequency.

9. The distributed multi-probe based borehole wall fracture location method of claim 6, wherein, In step S440, it includes: the formula for calculating the signal-to-noise ratio weight of the probe is: wherein, is the signal-to-noise ratio of the first i probe, N is the total number of probes, N = 9 ，p is the adjustment index, typically, p = 1 or 2, the formula takes 2. Then, the RANSAC algorithm is used to remove the abnormal value of the data collected by the probe, 3-4 probes are randomly selected to calculate the initial sound source position, the residual of the TDOA of other probes and the initial solution is counted, the abnormal probe with residual greater than the threshold is removed, and iterative optimization is performed until convergence.

10. The distributed multi-probe based borehole wall fracture location method of claim 2, wherein, In step S500, three areas are divided on the host computer page, which are real-time display waveform, video monitoring window and positioning window respectively; when saving files, hierarchical directories are established according to dates; the format of acoustic emission signal saved is CSV format, containing timestamp, energy, frequency, positioning coordinates, video file is accompanied by metadata, recording relevant acoustic emission time ID and acquisition parameters, daily data is automatically compressed and backed up.

Images