Automatic monitoring and alarm method based on acoustic emission data classification

Abstract

The present invention proposes an automatic monitoring and alarm method based on acoustic emission data classification, including: obtaining the acoustic emission data of the monitored object within a sampling time period; determining the type of acoustic emission parameter data participating in the rating according to the industry to which the application scenario of the monitored object belongs; and dividing the intervals within the selected parameter value range according to national standards or industry standards for the type of selected acoustic emission parameter data, setting different actual intensity, statistical intensity, statistical activity level and comprehensive level criteria, and making corresponding alarm level responses according to the signals collected by the sensor in real time; finally, the alarm level of the monitored object is pushed through the communication network to realize unattended monitoring alarm, and the user determines the performance or status of the monitored object according to the alarm level. The statistical time and criteria provided in the present invention can be set to achieve standard evaluation and rating, reduce the technical requirements for users, and have great significance for the large-scale application of acoustic emission technology in industry.

Description

Technical Field

[0001] The invention relates to an acoustic emission data processing method, and in particular to an automatic monitoring and alarming method and system based on acoustic emission data classification. Background Art

[0002] Acoustic emission inspection / monitoring technology, with its characteristics of non-stop production, online, and overall defect evaluation, has incomparable advantages in industrial applications, especially in metal atmospheric pressure storage tank inspection and metal pressure vessel inspection. However, traditional acoustic emission inspection relies on professional acoustic emission technicians to deeply analyze acoustic emission data in order to evaluate and rate the inspection object. This method is not only time-consuming and labor-intensive, but also has the characteristics of many application industries, great differences, and non-universal technical standards, which makes acoustic emission inspection and monitoring technology only applicable in some research fields and difficult to popularize in industrial applications. Summary of the invention

[0003] In view of the deficiencies in the prior art, the present invention provides an automatic acoustic emission data classification algorithm and implementation. The automatic acoustic emission data classification algorithm obtains the intensity level, activity level, and comprehensive level of intensity and activity from the acoustic emission impact parameters, and automatically converts the complex acoustic emission data into easy-to-understand alarm levels. Through embedded chip hardware and software implementation, the alarm level results are output to the cloud, pushed to mobile phones, local clients and other platforms, and unattended monitoring and alarm are realized. The statistical time and criterion can be set to achieve standard evaluation and rating, reduce the technical requirements for users, and have great significance for the large-scale application of acoustic emission technology in industry.

[0004] In order to solve the above technical problems, the present invention proposes an automatic monitoring and alarm method based on acoustic emission data classification, comprising the following steps:

[0005] Step 1, obtaining the acoustic emission data of the monitored object: converting the electrical signal collected by the real-time online sensor into acoustic emission parameter data, the types of acoustic emission parameter data include: amplitude, ring count, duration, energy, rise count, rise time, root mean square RMS, average signal level ASL, centroid frequency, peak frequency and impact number;

[0006] Step 2, determine the alarm level of the monitored object: select the type of acoustic emission parameter data participating in the rating according to the application scenario of the monitored object; set different real-time intensity level criteria for the selected type of acoustic emission parameter data according to national standards or industry standards; according to the real-time intensity level criterion, determine all real-time intensity levels of the monitored object in the sampling time period based on the value of the acoustic emission parameter collected in real time during the sampling time period in units of sensor sampling step length; according to the levels of all real-time intensities in the above sampling time period, take the highest real-time intensity level as the statistical intensity level in the sampling time period; according to the number of different real-time intensity levels in the sampling time period, set the criterion of statistical activity levels of different intensities, and take the statistical activity level with the largest number of occurrences as the statistical activity level in the sampling time period; determine the comprehensive level according to the statistical intensity level and statistical activity level in the sampling time period; the alarm level includes one or more of the real-time intensity level, statistical intensity level, statistical activity level, and comprehensive level;

[0007] Step 3: Push the alarm level of the monitored object to one or more of the cloud server, mobile phone and local client terminal through the communication network to realize unattended monitoring alarm push. The user determines the performance or status of the monitored object according to the alarm level.

[0008] Furthermore, the automatic monitoring and alarm method of the present invention comprises:

[0009] The national standards or industry standards mentioned above include at least: "JT / T 1037-2022 Technical Specifications for Monitoring of Highway Bridge Structures"; "JB / T 10764-2007 Nondestructive Testing of Atmospheric Pressure Metal Storage Tank Bottom Plate Acoustic Emission Detection and Evaluation"; "NB / T47013-2015 Nondestructive Testing of Pressure Equipment" Part 9: Acoustic Emission Monitoring; "GBT34370.8-2020 Nondestructive Testing of Amusement Facilities" Part 8: Acoustic Emission Detection; "NB / T10859-2021 Technical Conditions for Online Monitoring System of Metal Structure Equipment Status of Hydropower Engineering".

[0010] In step 2, the criteria for setting different real-time intensity levels for the type of acoustic emission parameter data selected according to national standards or industry standards include: determining the numerical range for each selected acoustic emission parameter data, and arranging them from small to large and dividing them into N numerical intervals, N≤4, and the N numerical intervals are successively recorded as interval one, interval two, interval three and interval four, and the interval one, interval two, interval three and interval four correspond to the first level real-time intensity, interval two corresponds to the second level real-time intensity, interval three corresponds to the third level real-time intensity, and interval four corresponds to the fourth level real-time intensity. Any one or several intervals of the N numerical intervals can be selected as the real-time intensity level intervals for participating in the rating.

[0011] In step 2, according to the number of different real-time intensity levels in the sampling time period, the criteria for different statistical activity levels are set, including: the number of each level of real-time intensity level in the sampling time period is arranged from small to large and divided into M numerical intervals, M≤5, interval one corresponds to the first level statistical activity, interval two corresponds to the second level statistical activity, interval three corresponds to the third level statistical activity, interval four corresponds to the fourth level statistical activity, and interval five corresponds to the fifth level statistical activity. Any one or more intervals of the M numerical intervals can be selected as the statistical activity level intervals for participation in rating.

[0012] In step 2, N=4, M=5, the criterion for determining the comprehensive level is determined based on the statistical intensity level and the statistical activity level within the sampling period, including:

[0013] The statistical level of activities with the first-level statistical intensity is the first level, or the statistical level of activities with the second-level statistical intensity is the first level, and the comprehensive level is the first level;

[0014] If the statistical level of activities with the first-level statistical intensity is the second level, or the statistical level of activities with the second-level statistical intensity is the second level, or the statistical activity level of the third-level statistical intensity is the first level, then the comprehensive level is the second level;

[0015] If the statistical activity level of the first-level statistical intensity is level three, or the statistical activity level of the first-level statistical intensity is level four, or the statistical activity level of the second-level statistical intensity is level three, or the statistical activity level of the third-level statistical intensity is level two, or the statistical activity level of the fourth-level statistical intensity is level one, then the comprehensive level is level three;

[0016] If the statistical level of activities with statistical intensity of level 2 is level 4, or the statistical level of activities with statistical intensity of level 3 is level 3, or the statistical level of activities with statistical intensity of level 3 is level 4, or the statistical level of activities with statistical intensity of level 1 is level 5, or the statistical level of activities with statistical intensity of level 4 is level 2, or the statistical level of activities with statistical intensity of level 4 is level 3, then the comprehensive level is level 4;

[0017] If the statistical activity level of the second-level statistical intensity is level five, or the statistical activity level of the third-level statistical intensity is level five, or the statistical activity level of the fourth-level statistical intensity is level four, or the statistical activity level of the fourth-level statistical intensity is level five, then the comprehensive level is level five.

[0018] Compared with the prior art, the present invention has the following beneficial effects:

[0019] In the past, acoustic emission testing relied on professional acoustic emission technicians to analyze the data and evaluate and grade the defects. In addition, the technology is difficult to popularize on a large scale due to its many application industries, large differences, and non-universal technical standards. The automatic monitoring and alarm method based on acoustic emission data classification of the present invention has the characteristics of non-stop production, online, and overall defect evaluation, and has incomparable advantages in industrial applications, especially in metal atmospheric pressure storage tank detection, metal pressure vessel detection, etc.

[0020] The method of the present invention can automatically calculate and grade massive acoustic emission data. The algorithm is implemented through embedded hardware and software, so that the acoustic emission equipment, especially the monitoring equipment, can continuously output the grade, so as to quickly judge the state of the monitored / detected object, eliminating the trouble of traditional acoustic emission relying on professionals to analyze acoustic emission related data, so that users who do not understand acoustic emission can also quickly get started. In addition, through this automatic grading algorithm, users can set the criteria that meet the standards in different industries according to different industry standards, automatically evaluate and grade according to the standards, and output the results. It is of great significance for industrial applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a data flow chart of the automatic monitoring and alarm method of the present invention. DETAILED DESCRIPTION

[0022] The design concept of an automatic monitoring and alarm method based on acoustic emission data classification proposed by the present invention is: first, the acoustic emission data of the monitored object is obtained within a certain sampling time period; then, the type of acoustic emission parameter data participating in the rating is determined according to the industry to which the application scenario of the monitored object belongs; and after dividing the interval within the selected parameter value range according to the type of the selected acoustic emission parameter data according to national standards or industry standards, different real-time intensity levels, different statistical intensity levels, different statistical activity levels and comprehensive level criteria are set, and corresponding alarm level responses are made according to the signals collected by the sensor in real time; finally, the alarm level of the monitored object is pushed to one or more of the cloud server, mobile phone and local client terminal through the communication network, so as to realize unattended monitoring alarm push, and the user determines the performance or status of the monitored object according to the alarm level.

[0023] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments, but the following embodiments are by no means intended to limit the present invention in any way.

[0024] Example 1

[0025] The automatic monitoring and alarm method based on acoustic emission data classification for monitoring storage tanks includes the following steps:

[0026] Step 1. Obtain the acoustic emission data of the monitored object: Install the acoustic emission sensor on the outside of the tank to collect the corrosion signal of the tank. The corrosion signal is collected by the acoustic emission collector and converted into a digital signal by AD, and the acoustic emission parameter data is extracted, including: amplitude (dB), ring count (times), duration (us), energy (kPj), rise count (times), rise time (us), root mean square RMS (mV), average signal level ASL (dB), centroid frequency (kHz), peak frequency (kHz) and number of impacts (times).

[0027] The on-site noise signal is XdB, and the threshold is set to (X+6)dB.

[0028] Step 2: Select the type of acoustic emission parameters to be rated, set the value range, the corresponding intensity level range, and perform real-time intensity rating for each set of real-time acoustic emission parameters; after determining the rating time length, determine the alarm level according to the parameter value of the acoustic emission data. The process is as follows:

[0029] 2-1) Select the amplitude (dB) among the acoustic emission parameters as the acoustic emission parameter involved in the rating.

[0030] 2-2) Select two intervals as the real-time intensity level intervals for participation in rating:

[0031] The amplitude value is set to be less than (X+6)dB as the first-level real-time intensity, and the amplitude value is greater than or equal to (X+6)dB as the second-level real-time intensity.

[0032] 2-3) Set to count all data within one hour once every hour according to the number of times corresponding to the two real-time intensity levels mentioned in 2-2). For example, the results within one hour are: Level 1 real-time intensity: 0 times; Level 2 real-time intensity: 8250 times; Level 3 real-time intensity: 0 times; the maximum real-time intensity level within one hour is Level 2, so the statistical intensity level within one hour is Level 2.

[0033] 2-4) In this embodiment 1, the five activity intervals set within the range of 0 to +∞ are:

[0034] When the activity is less than or equal to 500, it is the first-level statistical activity;

[0035] When the activity is greater than 500 and less than or equal to 5000, it is the secondary statistical activity;

[0036] When the activity is greater than 5000 and less than or equal to 50000, it is the third-level statistical activity;

[0037] When the activity is greater than 50,000 and less than or equal to 500,000, it is the fourth level statistical activity;

[0038] When the activity is greater than 500,000, it is level 5 statistical activity;

[0039] The second-level real-time intensity level appeared 8250 times within the one-hour period. It can be concluded that the statistical activity level within the one-hour period was the third-level statistical activity level.

[0040] 2-5) Determine the comprehensive level based on the statistical intensity level and statistical activity level;

[0041] The method for determining the comprehensive level according to the statistical intensity level and the statistical activity level in the present invention is shown in Table 1.

[0042] Table 1

[0043]

[0044] In this embodiment 1, within the one hour, the statistical intensity is level two, and the statistical activity is level three. According to the comprehensive level determination method set in the present invention, it is concluded that the corresponding rating alarm level within the one hour is a comprehensive level three.

[0045] In this embodiment 1, the alarm level within the one hour includes the second-level statistical intensity, the third-level statistical activity and the third-level comprehensive level; the output can be selected as needed, and this embodiment 1 only selects the comprehensive level output.

[0046] When the 1-hour rating period ends, the next 1-hour rating period will begin immediately, and the cycle will repeat to achieve 24-hour online monitoring.

[0047] Step 3: During the online monitoring process, the alarm level of the monitored object obtained at the end of each rating time period is pushed to one or more of the cloud server, mobile phone and local client terminal through the communication network at any time to realize unattended monitoring alarm push. Finally, the user determines the performance or status of the monitored object through the alarm level.

[0048] According to the provisions of Table 3 in Clause 10 Test Results and Evaluation of Standard JB / T 10764-2007 Nondestructive Testing, Acoustic Emission Testing and Evaluation Methods for Atmospheric Pressure Metal Storage Tanks, the classification of the acoustic emission source of the tank bottom plate based on regional positioning analysis shows that the overall level of the tank is Level 3, with obvious signs of local corrosion and a moderate corrosion condition, suggesting that users need to consider maintenance.

[0049] Example 2

[0050] The automatic monitoring and alarm method based on acoustic emission data classification monitors whether the bridge wire rope is broken, including:

[0051] Step 1, obtain the acoustic emission data of the monitored object: install the acoustic emission sensor on the wire rope and collect the wire break signal. Once the wire rope breaks, the wire break signal is transmitted to the sensor through the suspension rod, collected by the acoustic emission collector and converted into a digital signal by AD, and the acoustic emission parameter data is extracted: amplitude (dB), ring count (times), duration (us), energy (kPj), rise count (times), rise time (us), root mean square RMS (mV), average signal level ASL (dB), center of mass frequency (kHz), peak frequency (kHz) and number of impacts (times).

[0052] Step 2: Select the type of acoustic emission parameters to be rated, set the value range, the corresponding intensity level range, and perform real-time intensity rating for each set of real-time acoustic emission parameters; after determining the rating time length, determine the alarm level based on the parameter value of the acoustic emission data. Including:

[0053] 2-1) The amplitude (dB) and energy (kPj) of the acoustic emission parameters are selected as the acoustic emission parameters participating in the rating. A set of acoustic emission parameters that simultaneously meet the requirements of amplitude and energy are used as effective alarm data to participate in the rating alarm.

[0054] Set the intensity alarm to:

[0055] 2-2) Select two intervals as the statistical strength level intervals for participation rating:

[0056] When the amplitude value is less than 90dB and the energy value is less than 10000kpj, the real-time intensity level is level one;

[0057] When the amplitude value is greater than or equal to 90dB and the energy value is greater than or equal to 10000kpj, the real-time intensity level is level 2;

[0058] 2-3) Select the external output alarm parameter as the real-time intensity level.

[0059] At a certain moment, when a set of parameters received is: amplitude = 95dB, energy = 18650kPj, according to the above rating criteria "when the amplitude value is greater than 90dB and the energy value is greater than 10000kpj, the real-time intensity level is level 2", the level 2 real-time intensity is output.

[0060] In this embodiment 2, according to the requirements of the structural change wire breakage in Table 9 of Section 11.3.3 of the Technical Specifications for Highway Bridge Structure Monitoring of JB / T 1037-2022, the monitoring data over-limit threshold setting table, when a wire break occurs, the over-limit alarm is level 2.

[0061] Step 3: During the online monitoring process, the alarm level of the monitored object obtained at the end of each rating time period is pushed to one or more of the cloud server, mobile phone and local client terminal through the communication network at any time to realize unattended monitoring alarm push. Finally, the user determines the performance or status of the monitored object through the alarm level.

[0062] In accordance with the requirements of structural change wire breakage in Table 9 of Monitoring Data Exceeding Limit Threshold Setting Table in Section 11.3.3 of JB / T 1037-2022 Technical Specifications for Highway Bridge Structure Monitoring, the alarm result is pushed to one or more of the cloud server, mobile phone and local client terminal through the communication network to realize unattended monitoring alarm push. Finally, the user knows through the alarm level that the current status of the wire rope is the second acoustic emission alarm and there is a wire breakage phenomenon.

[0063] Although the present invention is described above in conjunction with the accompanying drawings, the present invention is not limited to the above-mentioned specific embodiments, which are merely illustrative and not restrictive. Under the inspiration of the present invention, ordinary technicians in this field can also perform online real-time monitoring of monitored objects in different application scenarios without departing from the purpose of the present invention. The national standards or industry standards listed therein are only illustrative. Under the inspiration of the design concept of the monitoring method of the present invention, those skilled in the art can make specific alarm level (including real-time intensity level, statistical intensity level, statistical activity level and comprehensive level) judgment criteria according to different industries, thereby realizing automatic monitoring and alarm based on acoustic emission data classification in different industries, which are all within the protection of the present invention.

Claims

1. An automatic monitoring and alarm method based on acoustic emission data classification, characterized in that: The following steps are involved: Step 1: Obtain the acoustic emission data of the monitored object: Convert the electrical signal collected by the real-time online sensor into acoustic emission parameter data, wherein the types of acoustic emission parameter data include: amplitude, ring count, duration, energy, rise count, rise time, root mean square RMS, average signal level ASL, centroid frequency, peak frequency and number of impacts; Step 2: Determine the alarm level of the monitored object: Select the type of acoustic emission parameter data to be rated according to the application scenario of the monitored object; set the criteria for different real-time intensity levels for the selected type of acoustic emission parameter data according to national standards or industry standards; According to the criterion of real-time intensity level, all real-time intensity levels of the monitored object in the sampling time period are determined based on the values ​​of the acoustic emission parameters collected in real time in the sampling time period in units of the sensor sampling step length; According to all the real-time intensity levels in the above sampling period, the highest real-time intensity level is used as the statistical intensity level in the sampling period; According to the number of different real-time intensity levels in the sampling time period, the criteria for statistical activity levels of different intensities are set, and the statistical activity level with the largest number of occurrences is used as the statistical activity level in the sampling time period; The comprehensive level is determined based on the statistical intensity level and the statistical activity level during the sampling period; The alarm level includes one or more of the real-time intensity level, statistical intensity level, statistical activity level, and comprehensive level; Step 3: Push the alarm level of the monitored object to one or more of the cloud server, mobile phone and local client terminal through the communication network to realize unattended monitoring alarm push, and the user determines the performance or status of the monitored object according to the alarm level; In step 2, the national standards or industry standards at least include: "JT / T 1037-2022 Technical Specifications for Highway Bridge Structure Monitoring"; "JB / T 10764-2007 Nondestructive Testing of Atmospheric Pressure Metal Storage Tank Bottom Plate Acoustic Emission Testing and Evaluation"; "NB / T47013-2015 Nondestructive Testing of Pressure Equipment" Part 9: Acoustic Emission Monitoring; "GBT34370.8-2020 Nondestructive Testing of Amusement Facilities" Part 8: Acoustic Emission Testing; "NB∕T10859-2021 Technical Requirements for Online Monitoring System of Metal Structure Equipment Status in Hydropower Projects"; In step 2, the criteria for setting different real-time intensity levels for the type of selected acoustic emission parameter data according to national standards or industry standards include: determining a numerical range for each selected acoustic emission parameter data, and arranging them from small to large and dividing them into N numerical intervals, N≤4, the N numerical intervals are successively recorded as interval one, interval two, interval three and interval four, the interval one, interval two, interval three and interval four respectively correspond to the first level real-time intensity, interval two corresponds to the second level real-time intensity, interval three corresponds to the third level real-time intensity, and interval four corresponds to the fourth level real-time intensity; select any one or several intervals among the N numerical intervals as the real-time intensity level intervals participating in the rating; In step 2, according to the number of different real-time intensity levels within the sampling time period, the criteria for different statistical activity levels are set, including: the number of each level of real-time intensity level within the sampling time period is arranged from small to large and divided into M numerical intervals, M≤5, interval one corresponds to the first level statistical activity, interval two corresponds to the second level statistical activity, interval three corresponds to the third level statistical activity, interval four corresponds to the fourth level statistical activity, and interval five corresponds to the fifth level statistical activity.

2. The automatic monitoring and alarm method according to claim 1, characterized in that: Any one or several intervals among the M numerical intervals are selected as the statistical activity level intervals for participation in rating.

3. The automatic monitoring and alarm method according to claim 1, characterized in that: In step 2, N=4, M=5, the criterion for the comprehensive level is determined based on the statistical intensity level and the statistical activity level during the sampling period; including: The statistical activity level of the first-level statistical intensity is level one, or the statistical activity level of the second-level statistical intensity is level one, and the comprehensive level is level one; If the statistical activity level of the first-level statistical intensity is the second level, or the statistical activity level of the second-level statistical intensity is the second level, or the statistical activity level of the third-level statistical intensity is the first level, then the comprehensive level is the second level; If the statistical activity level of the first-level statistical intensity is level three, or the statistical activity level of the first-level statistical intensity is level four, or the statistical activity level of the second-level statistical intensity is level three, or the statistical activity level of the third-level statistical intensity is level two, or the statistical activity level of the fourth-level statistical intensity is level one, then the comprehensive level is level three; If the statistical activity level of the second-level statistical intensity is level four, or the statistical activity level of the third-level statistical intensity is level three, or the statistical activity level of the third-level statistical intensity is level four, or the statistical activity level of the first-level statistical intensity is level five, or the statistical activity level of the fourth-level statistical intensity is level two, or the statistical activity level of the fourth-level statistical intensity is level three, then the comprehensive level is level four; If the statistical activity level of the second-level statistical intensity is level five, or the statistical activity level of the third-level statistical intensity is level five, or the statistical activity level of the fourth-level statistical intensity is level four, or the statistical activity level of the fourth-level statistical intensity is level five, then the comprehensive level is level five.

4. The automatic monitoring and alarm method according to claim 1, characterized in that: The sampling time period is 1 minute to 100 days. The monitoring process includes multiple rounds of sampling. When the previous sampling time period ends, the next sampling time period begins, and steps 1 to 3 are repeated repeatedly.

5. The automatic monitoring and alarm method according to claim 1, characterized in that: The units of each acoustic emission parameter data are: The unit of amplitude is dB, The unit of ring count is times. The unit of duration is us. The unit of energy is kPj, The unit of rising count is times. The unit of rise time is us. The unit of root mean square RMS is mV. The unit of average signal level ASL is dB. The unit of centroid frequency is kHz, The unit of peak frequency is kHz, The unit of the number of collisions is times.

Images