Acoustic emission signal noise filtering and hazard comprehensive judgment method based on multi-system fusion
Through the multi-system fusion method, noise is identified and filtered down the coal mine, combined with gas and roof monitoring results, real-time monitoring and early warning of coal rock power disasters is achieved, and the acoustic emission signal extraction is solved in the existing technology under the influence of noise interference. Pure problems improve the accuracy and reliability of analysis results.
Patent Information
- Application Number
- CN202010437494.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2020-05-21
- Publication Date
- 2024-11-05
- Estimated Expiration
- 2040-05-21
AI Technical Summary
In the case of many noise types and complex environments, it is difficult to quickly identify and filter out random human-acting noise and most mechanical noise, resulting in impure extraction of acoustic emission signals, affecting the analysis results. Accuracy and reliability.
Using a multi-system fusion method, the acoustic emission signal is continuously monitored and preprocessed through the acoustic emission monitor, combined with the real-time communication of the personnel positioning system, production management system and safety monitoring system, the noise source is locked and noise identification is performed, and the acoustic emission signal is calculated. Eigenvector value, combined with gas and roof monitoring results, real-time monitoring and comprehensive early warning.
It realizes the rapid and accurate identification and filtering of noise signals, improves the purity of acoustic transmitted signals, ensures real-time monitoring and reliable early warning of dynamic disasters, and improves the accuracy of analysis results and the real-time data processing.
Smart Images

Figure CN111832604B_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of acoustic emission monitoring of coal-rock dynamic disasters, and relates to a noise filtering and hazard comprehensive judgment method of a coal mine acoustic emission monitoring system with multi-system fusion identification. Background Art
[0002] Coal rock dynamic disaster is a rapid instability phenomenon with dynamic effects and disaster consequences caused by coal rock mass in mines under the action of ground stress and gas. During the gestation, stimulation and development stages of the disaster evolution process, the cracks in the coal rock mass will release acoustic emission information. By continuously monitoring and analyzing the acoustic emission information released during the coal rock mass destruction process, dynamic disaster prediction of the working face can be achieved.
[0003] Although my country has conducted research in this area for many years, due to the complex underground environment of coal mines and the variety of noise types, the same type of noise has large differences in different environments, and the acoustic emission signal and the noise signal are often mixed together, which brings great interference to the extraction of acoustic emission signals. The existing isolation and noise filtering technology can well isolate low-frequency environmental noise, filter out electrical noise, and filter out some mechanical noise through the known time and frequency characteristics of the noise signal. However, for random human activity noise and most mechanical noise, it is difficult to quickly identify them, so the real-time filtering effect is not good, and it is difficult to obtain a pure acoustic emission signal, which affects the accuracy and reliability of the analysis results. Summary of the invention
[0004] In view of this, the purpose of the present invention is to provide an acoustic emission signal noise filtering and comprehensive hazard judgment method based on multi-system fusion, which can quickly filter out noise interference, accurately extract the characteristic vector value of the acoustic emission signal, and combine the gas and roof monitoring results to realize real-time monitoring and comprehensive early warning of coal and rock dynamic disasters.
[0005] In order to achieve the above object, the present invention provides the following technical solutions:
[0006] A method for filtering noise of acoustic emission signals and comprehensive hazard determination based on multi-system fusion specifically comprises the following steps:
[0007] S1: The underground acoustic emission monitor in coal mines continuously monitors the acoustic emission signals released during the destruction of coal and rock mass; the original acoustic emission full waveform signals triggered and collected are pre-processed and uploaded to the host computer software via Ethernet;
[0008] S2: The AE host computer software communicates with the personnel positioning system, production management system, and safety monitoring system in real time. The production management system provides the daily production and maintenance shift schedules of the mine. The personnel positioning system provides information on personnel activities within the AE monitoring range for each shift. The safety monitoring system provides equipment operation and start / stop status.
[0009] S3: The acoustic emission host computer software triggers the reception of the original acoustic emission full waveform signal downhole, locks the possible noise within the monitoring range according to step S2, and performs noise identification;
[0010] S4: Locate the noise source according to the noise identification result, filter out the noise interference, and calculate the characteristic vector value of the acoustic emission signal.
[0011] S5: The characteristic vector value of the acoustic emission signal is combined with the gas and roof monitoring results shared by the fusion system to conduct real-time monitoring and comprehensive early warning of coal and rock dynamic disasters.
[0012] Furthermore, in step S1, the preprocessing includes: signal acquisition, resampling, A\D conversion, low-pass filtering, DC notch removal, limiting filtering, and waveform event extraction.
[0013] Furthermore, in step S2, during the production shift, the personnel type information of the personnel positioning system is integrated with the equipment operating status to lock in possible noises such as blasting, coal mining, tunneling, fan operation or drilling; during the maintenance shift, by obtaining personnel positioning information and equipment operating status, possible human activity noise and fan operation noise, etc.
[0014] Furthermore, in step S3, noise identification specifically includes:
[0015] 1) If there is no environmental noise, it is marked as no noise and marked as an acoustic emission event;
[0016] 2) If there is a single noise source, the time and frequency characteristics are compared with the noise samples in the noise library. If the similarity is ≥ 60%, it is marked as containing noise and the noise type is marked; if the similarity is < 60%, it is marked as no noise and an acoustic emission event;
[0017] 3) If there are multiple noise sources, the time and frequency characteristics are compared with the noise samples in the noise library according to the principle of giving priority to the noise signal strength. If the similarity is ≥ 60%, it is marked as containing noise and the noise type is marked; if the similarity is < 60%, it is marked as no noise and marked as an acoustic emission event.
[0018] Furthermore, in step S4, the characteristic vector values involved in the similarity comparison mainly include: rise time, duration, peak value, average value, energy, number of ringings, wave hit count and main frequency distribution.
[0019] Furthermore, according to the hazard judgment rules, the noise filtering model mainly filters the event energy value. For the acoustic emission events marked as containing noise, the predefined filtering rules when the noise is stored are used to choose to completely filter out or filter out the noise influence through the noise filtering algorithm; the filtering algorithm is: acoustic emission event energy = original acoustic emission event energy - impact factor × event duration × noise energy rate
[0020] Furthermore, the characteristic vector value of the acoustic emission event marked as noise-free can be directly used for subsequent analysis, or it can be determined by manual audio-visual combined with offline analysis function to determine whether it is a type of noise that has not been included. The steps of manual re-examination specifically include:
[0021] 1) Confirm that it is a pure acoustic emission signal and end the re-inspection.
[0022] 2) If it is a noise signal and the noise library has been included, save it as a noise sample to enrich the noise library;
[0023] 3) If it is a noise signal and the noise type is not stored in the database, the noise is identified by combining the associated system or time-frequency domain features, the characteristic vector value of the noise event is calculated, and a similarity judgment rule is established. A new noise type is created as a noise sample and stored in the noise signal database;
[0024] 4) If it is a noise signal but does not meet the filtering conditions, it is stored in the noise library as a random noise sample, and the manual re-inspection ends.
[0025] Furthermore, in step S5, the comprehensive determination method of power disasters specifically includes:
[0026] 1) Real-time alarm for single acoustic emission signal: When the energy rate of the acoustic emission event exceeds the set value, it will be a major event. At this time, an audible and visual alarm will be triggered to prompt whether to filter out the noise and the type of filtered noise;
[0027] 2) Multi-system comprehensive warning: A comprehensive alarm is carried out every shift, and a graded warning is given for dynamic disasters through the accumulated acoustic emission energy, number of rings, average confidence level, combined with the gas parameters of the safety monitoring system, the anchor stress value of the roof monitoring system and the separation monitoring value.
[0028] The on-duty energy rate exceeds the H limit, yellow warning, warning level III.
[0029] The on-duty energy rate exceeds the HH limit, orange warning, warning level II.
[0030] If the pressure of the roof monitoring anchor exceeds the limit, the separation monitoring value exceeds the limit, or the gas exceeds the limit continuously, the warning level will be upgraded by one level. The highest warning level is red, warning level I.
[0031] The beneficial effects of the present invention are:
[0032] (1) The method of the present invention uses multi-system fusion data to lock possible noise sources and locations, and combines noise samples in the noise library for similarity judgment, thereby achieving rapid and accurate identification of noise signals. According to the hazard judgment rules, the filtering of known noise signals is targeted, and the noise filtering model is simple and the amount of calculation is small, which improves the noise filtering effect while ensuring the real-time nature of data processing.
[0033] (2) During the application process, the method of the present invention obtains noise signals in actual application scenarios as noise library samples, thereby reducing the same type of noise, such as the differences between drilling rigs from different manufacturers in different application environments.
[0034] (3) It has real-time acoustic emission alarm, supplemented by environmental gas, anchor stress value, and delamination monitoring value to achieve disaster classification warning, and the warning result is more reliable.
[0035] (4) The classification and identification of noise signals in the method of the present invention also helps system maintenance personnel to remotely and real-time understand the operation status of personnel and equipment near the monitoring point.
[0036] Other advantages, objectives and features of the present invention will be described in the following description to some extent, and to some extent, will be obvious to those skilled in the art based on the following examination and study, or can be taught from the practice of the present invention. The objectives and other advantages of the present invention can be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS
[0037] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be described in detail below in conjunction with the accompanying drawings, wherein:
[0038] Figure 1 This is a schematic diagram of the structure of the acoustic emission monitoring host computer involved in the present invention;
[0039] Figure 2 The present invention relates to a multi-system fusion system data flow diagram;
[0040] Figure 3 This is the basic flow chart of noise identification and filtering of the present invention;
[0041] Figure 4 This is the basic flow chart of the manual re-inspection and storage of noise signals in the present invention. DETAILED DESCRIPTION
[0042] The following describes the embodiments of the present invention by specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments only illustrate the basic concept of the present invention in a schematic manner, and the following embodiments and features in the embodiments can be combined with each other without conflict.
[0043] See also Figure 1 to Figure 4 , a method for noise filtering and comprehensive hazard determination of acoustic emission signals based on multi-system fusion, specifically including:
[0044] Step 1: The underground acoustic emission monitor in coal mines continuously monitors the acoustic emission signals released during the destruction of coal and rock mass; the original acoustic emission full waveform signal collected by the trigger is pre-processed and uploaded to the host computer software via Ethernet. The pre-processing includes: signal acquisition, resampling, A\D conversion, low-pass filtering, DC notch removal, floating threshold noise reduction, and event extraction.
[0045] Step 2: The AE host computer software communicates in real time with the personnel positioning system, production management system, and safety monitoring system. The production management system provides the daily production and maintenance schedule of the mine, the personnel positioning system provides information on personnel activities within the AE monitoring range for each shift, and the safety monitoring system provides equipment operation and start-stop status. During the production shift, the comprehensive personnel positioning system's personnel type information, combined with the equipment operation status, locks in possible noises such as blasting, coal mining, tunneling, fan operation, or drilling; during the maintenance shift, by obtaining personnel positioning information and equipment operation status, locks in possible human activity noises and fan operation noises.
[0046] Step 3: Reference Figure 3 , the acoustic emission host computer software triggers the reception of the original acoustic emission event full waveform signal, locks the possible noise within the monitoring range according to step 2, and performs noise identification.
[0047] 1) If there is no environmental noise at this time, it is marked as no noise and marked as an acoustic emission event.
[0048] 2) If there is a single noise source at this time, the time and frequency characteristics are compared with the noise samples in the noise library. If the similarity is ≥ 60%, it is marked as containing noise and the noise type is marked; if the similarity is < 60%, it is marked as no noise and an acoustic emission event.
[0049] 3) If there are multiple noise sources at this time, compare the time and frequency characteristics with the noise samples in the noise library according to the principle of priority of noise signal strength. If the similarity is ≥ 60%, it is marked as containing noise and the noise type is marked; if the similarity is < 60%, it is marked as no noise and an acoustic emission event.
[0050] The eigenvector values involved in similarity comparison mainly include: rise time, duration, peak value, average value, energy, number of rings, wave hit count, and main frequency distribution.
[0051] According to theoretical and experimental research, the frequency and energy of events are the most important parameters reflecting the activity of acoustic emission. With the increase of stress, the number of acoustic emission events increases and becomes more and more intensive. Near the intensity value, the emission event rate and energy rate reach the maximum. Since the frequency and energy of events are the most important parameters reflecting the activity of acoustic emission, in order to improve the speed of noise filtering, this patent uses the hazard judgment rules to filter noise mainly for the event energy value. For acoustic emission events marked as containing noise, the predefined filtering rules when the noise is stored can be used to choose to completely filter out or filter out the influence through the following noise filtering algorithm: Acoustic emission event energy = original acoustic emission event energy - impact factor × event duration × noise energy rate
[0052] Step 4: Dynamic disaster determination method
[0053] 1) Real-time alarm for single acoustic emission signal: When the energy rate of the acoustic emission event exceeds the set value, it will be a major event. At this time, an audible and visual alarm will be triggered to prompt whether to filter out the noise and the type of filtered noise;
[0054] 2) Multi-system comprehensive warning: A comprehensive alarm is carried out every shift, and a graded warning is given to dynamic disasters through the accumulated acoustic emission energy and number of rings during the shift combined with the gas parameters of the safety monitoring system, the anchor stress value of the roof monitoring system and the separation monitoring value.
[0055] The on-duty energy rate exceeds the H limit, yellow warning, warning level III.
[0056] The on-duty energy rate exceeds the HH limit, orange warning, warning level II.
[0057] If the pressure of the roof monitoring anchor exceeds the limit, the separation monitoring value exceeds the limit, or the gas exceeds the limit continuously, the warning level will be upgraded by one level. The highest warning level is red, warning level I.
[0058] Noise storage:
[0059] Reference Figure 4 , the AE events marked as noise-free, their feature vector values can be used directly for subsequent analysis, or they can be judged by manual audio-visual combined with offline analysis functions to determine whether they are unrecorded noise types. The steps of manual re-inspection include:
[0060] 1) Confirm that it is a pure acoustic emission signal and end the re-inspection.
[0061] 2) If it is a noise signal and the noise library has been included, save it as a noise sample to enrich the noise library;
[0062] 3) If it is a noise signal and the noise type is not stored in the database, the noise is identified by combining the associated system or time-frequency domain features, the characteristic vector value of the noise event is calculated, and a similarity judgment rule is established. A new noise type is created as a noise sample and stored in the noise signal database;
[0063] 4) If it is a noise signal but does not meet the filtering conditions, it is stored in the noise library as a random noise sample, and the manual re-inspection ends.
[0064] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solution, which should be included in the scope of the claims of the present invention.
Claims
1. A method for noise filtering and comprehensive risk assessment of acoustic emission signals based on multi-system fusion, characterized in that: The method specifically comprises the following steps: S1: Use the underground acoustic emission monitor to continuously monitor the acoustic emission signals released during the destruction of coal and rock mass. The original acoustic emission full waveform signal collected by the trigger is uploaded to the host computer software through Ethernet after preprocessing; the preprocessing includes: signal acquisition, resampling, A\D conversion, low-pass filtering, DC notch removal, limiting filtering, and waveform event extraction; S2: The AE host computer software communicates with the personnel positioning system, production management system, and safety monitoring system in real time. The production management system provides the daily production and maintenance shift schedules of the mine. The personnel positioning system provides information on personnel activities within the AE monitoring range for each shift. The safety monitoring system provides equipment operation and start / stop status. S3: The acoustic emission host computer software triggers the reception of the original acoustic emission full waveform signal downhole, locks the possible noise within the monitoring range according to step S2, and performs noise identification; Noise identification includes: 1) If there is no environmental noise, it is marked as no noise and marked as an acoustic emission event; 2) If there is a single noise source, the time and frequency characteristics are compared with the noise samples in the noise library. If the similarity is ≥ 60%, it is marked as containing noise and the noise type is marked; if the similarity is < 60%, it is marked as no noise and an acoustic emission event; 3) If there are multiple noise sources, compare the time and frequency characteristics with the noise samples in the noise library according to the principle of noise signal strength priority. If the similarity is ≥ 60%, it is marked as containing noise and the noise type is marked; if the similarity is < 60%, it is marked as no noise and an acoustic emission event; S4: Locate the noise source according to the noise identification result, filter out the noise interference, and calculate the eigenvector value of the acoustic emission signal; according to the danger judgment rule, the noise filtering model filters the event energy value, and for the acoustic emission events marked as containing noise, choose to completely filter them out or filter out the influence through the noise filtering algorithm according to the predefined filtering rules when the noise is stored; the noise filtering algorithm is: acoustic emission event energy = original acoustic emission event energy - impact factor × event duration × noise energy rate; The eigenvector value of the acoustic emission event marked as noise-free is directly used for subsequent analysis, or it can be judged whether it is a type of noise not included in the record through manual audio-visual combined with offline analysis function; the specific steps of manual re-inspection include: 1) Confirm that it is a pure acoustic emission signal and end the retest; 2) If it is a noise signal and the noise library has been included, save it as a noise sample to enrich the noise library; 3) If it is a noise signal and the noise type is not stored in the database, the noise is identified by combining the associated system or time-frequency domain features, the characteristic vector value of the noise signal is calculated and a similarity judgment rule is established. A new noise type is created as a noise sample and stored in the noise signal database; 4) If it is a noise signal but does not meet the filtering conditions, it will be stored in the noise library as a random noise sample, and the manual re-inspection ends; S5: The characteristic vector value of the acoustic emission signal is combined with the gas and roof monitoring results shared by the fusion system to conduct real-time monitoring and comprehensive early warning of coal and rock dynamic disasters; The comprehensive determination method of dynamic disaster risk specifically includes: 1) Real-time alarm for a single acoustic emission signal: When the energy rate of an acoustic emission event exceeds the set value, it will be considered a major event. At this time, an audible and visual alarm will be triggered to indicate whether to filter out the noise and the type of noise to be filtered out. 2) Multi-system integrated alarm: A comprehensive early warning is carried out every shift. The accumulated acoustic emission energy and number of rings during the shift are combined with the gas parameters of the safety monitoring system, the anchor stress value of the roof monitoring system and the separation monitoring value to carry out graded early warning of coal and rock dynamic disasters.
2. The method for acoustic emission signal noise filtering and comprehensive risk determination based on multi-system fusion according to claim 1 is characterized in that: In step S2, during the production shift, the personnel type information of the personnel positioning system is integrated with the equipment operating status to lock in the possible noises from blasting, coal mining, tunneling, fan operation or drilling; during the maintenance shift, the possible noises from human activities and fan operation are locked in by obtaining the personnel positioning information and the equipment operating status.
3. The method for acoustic emission signal noise filtering and comprehensive risk determination based on multi-system fusion according to claim 1 is characterized in that: In step S4, the characteristic vector values involved in the similarity comparison include: rise time, duration, peak value, average value, energy, number of rings, wave hit count and main frequency distribution.
Citation Information
Patent Citations
Method and device for identifying environmental noise sources
CN104952449A
Noise source identification method and system
CN105489215A
Method for building city noise map
CN105513489A
Strain type rockburst early warning method based on acoustic signal spectral analysis
CN105676267A
Work environment improvement system
JP2017146517A
