Long-distance acoustic emission monitoring device and method for PE gas pipeline leakage

Abstract

The present invention discloses a long-distance acoustic emission monitoring device and method for PE gas pipeline leakage, which relates to the technical field of gas pipeline monitoring, including: S1, collecting weak acoustic emission signal data under the leakage state of the pipeline to be tested; S2, optimizing the control parameters of the VMD algorithm by using a genetic algorithm and inputting them into the VMD algorithm, decomposing the collected signal, selecting effective IMF components, reconstructing the signal to complete the noise removal, extracting the energy entropy eigenvalue of the reconstructed signal to train the SVM support vector machine prediction model; S3, continuously monitoring the pipeline to be tested and saving the data, performing signal denoising processing by using step S2, extracting the energy entropy eigenvalue of the reconstructed signal and inputting it into the trained SVM support vector machine prediction model; S4, using the SVM support vector machine prediction model to identify the signal category collected in real time and perform subsequent processing. The present invention can realize long-distance PE pipeline leakage monitoring, has a simple device, does not need to lay wires along the pipeline, has high monitoring accuracy, low cost, good real-time performance, and is widely used.

Description

Technical Field

[0001] The invention relates to an acoustic emission monitoring technology for PE gas pipeline leakage, and in particular to a long-distance acoustic emission monitoring device and method for PE gas pipeline leakage. Background Art

[0002] As the most economical and environmentally friendly natural gas transportation equipment, pipeline transportation has many advantages, such as small footprint, short construction period, high transportation efficiency, safe and easy operation, less medium loss, low cost and high benefit, and is very suitable for long-distance transportation. With the vigorous development of polymer materials, polyethylene (PE) pipes are gradually replacing metal pipes and widely used in gas pipeline networks due to their excellent performance, such as high quality and low price, corrosion resistance, easy construction, good earthquake resistance, economic and environmental protection, and long service life.

[0003] At present, for reasons such as safety, reducing natural and man-made damage, and saving ground space, most of my country's PE gas pipelines are buried underground. Aging, corrosion, defects, and cumulative deformation can easily lead to leakage. Early leakage can easily develop into major safety accidents due to the lack of effective monitoring methods. Pipeline leakage monitoring is very necessary. It can monitor pipeline operation in a timely and effective manner, prevent the occurrence of safety accidents, and reduce economic losses, casualties, and environmental damage caused by leakage accidents. It has important social value and economic benefits.

[0004] For example, the Chinese patent with publication number CN217542288U designs a pipeline leakage acoustic emission detector, which can move on the outer wall of the pipeline, so that the detection is no longer limited to a fixed position, and the pipeline can be blocked by a blocking block to prevent further leakage. However, in actual situations, most gas pipelines are buried underground and it is difficult to move the detection device. Therefore, this patent is not suitable for long-distance detection of buried gas pipelines.

[0005] PE material is a viscoelastic polymer material, and the acoustic emission signal will experience severe acoustic attenuation when propagating in it. The monitoring distance when using the acoustic emission method is less than 1m. Therefore, a new type of long-distance acoustic emission monitoring device and method for PE gas pipeline leakage is urgently needed. Summary of the invention

[0006] The present invention overcomes the shortcomings of the prior art and provides a long-distance acoustic emission monitoring device and method for PE gas pipeline leakage, aiming to solve the problem of severe attenuation of PE pipeline acoustic emission signals, realize long-distance PE pipeline leakage acoustic emission monitoring, and achieve the purpose of accurate identification and positioning of pipeline leakage.

[0007] In view of the above problems in the prior art, according to one aspect of the present invention, in order to solve the above technical problems, the present invention adopts the following technical solutions:

[0008] A long-distance acoustic emission monitoring device for PE gas pipeline leakage, characterized by comprising:

[0009] Leakage acoustic emission source access devices, which are in several quantities and are respectively arranged at the leak-prone parts of the pipeline to be tested;

[0010] Buckles, the number of which corresponds to the leakage acoustic emission source access device, and each buckle is correspondingly arranged on both sides of a leakage acoustic emission source access device;

[0011] A metal wire, the number of which is one, the metal wire is connected to the buckle; a node acoustic emission sensor is connected to the metal wire between each two leakage acoustic emission source access devices; each of the node acoustic emission sensors is further connected to an acoustic emission preamplifier;

[0012] An acoustic emission instrument connected to the acoustic emission preamplifier;

[0013] A data storage and display device is connected to the acoustic emission instrument.

[0014] Furthermore, the plurality of node acoustic emission sensors have unique numbers.

[0015] Furthermore, the node acoustic emission sensors connected to each of the metal wires are spaced apart by 50-100 meters.

[0016] Furthermore, the metal wire is a copper wire.

[0017] Furthermore, the leakage acoustic emission source access device is a pipeline metal wrapping mesh.

[0018] Furthermore, an acoustic emission signal analysis system is embedded in the acoustic emission instrument, and the acoustic emission signal analysis system includes using a SVM support vector machine to identify abnormal signal categories and locate node acoustic emission sensors corresponding to leakage signals.

[0019] Furthermore, the training of the SVM support vector machine includes:

[0020] The genetic algorithm is used to optimize the VMD algorithm control parameters [K, α], where K is the number of decomposition layers and α is the penalty factor; the optimal parameters are input into the VMD algorithm to decompose the three types of collected signals, and the effective IMF components are selected through the correlation coefficient. The signal is reconstructed to remove the noise, and the energy entropy eigenvalue of the reconstructed signal is extracted to train the SVM support vector machine.

[0021] Furthermore, the range of the control parameters [K, α] is set to [3:20, 1000:4000].

[0022] Furthermore, the correlation coefficient: , where: Cov (X, Y) is the covariance of random variables X and Y; , The standard deviation of random variables X and Y respectively;

[0023] The IMF components with correlation coefficients greater than or equal to 0.3 are selected as effective components for signal reconstruction.

[0024] A long-distance acoustic emission monitoring method for PE gas pipeline leakage, comprising:

[0025] S1: Put the leakage acoustic emission source access device on the leak-prone part of the monitored pipeline, pass the signal transmission wire through the buckles on both sides of the leakage acoustic emission source access device and connect it to the buckles on both sides of the next adjacent leakage acoustic emission source access device, and also connect the wire to the node acoustic emission sensor, and then amplify the signal through the acoustic emission preamplifier and transmit it to the acoustic emission instrument;

[0026] S2: The acoustic emission signal analysis system uses a genetic algorithm to optimize the VMD algorithm control parameters [K, α], inputs the obtained optimal parameters [K, α] into the VMD algorithm, decomposes the three types of signals collected by the acoustic emission instrument in step S1, selects the effective IMF components through the correlation coefficient, reconstructs the signal to remove the noise, and extracts the energy entropy eigenvalue of the reconstructed signal to train the SVM support vector machine prediction model;

[0027] S3: Continuously monitor the operation status of multiple pipelines to be tested using multiple node acoustic emission sensors, save the abnormal signals found by the data storage and display device, use the VMD algorithm optimized in step S2 combined with the correlation coefficient to reconstruct the signal to complete the denoising process of the abnormal signal, extract the energy entropy eigenvalue of the reconstructed signal and input it into the SVM support vector machine trained in step S2 to complete the classification and recognition of the abnormal signal;

[0028] S4: After using the SVM support vector machine prediction model to identify the signal category collected in real time, issue a warning signal and guide processing based on the identification result; if it is a normal operation signal of the pipeline, continue monitoring; if a long-term interference signal is found, arrange staff to find the cause of the abnormal signal emitted by the pipeline near the corresponding sensor according to the number of the acoustic emission sensors at multiple nodes; if it is a leakage signal, close the valve immediately, and quickly and accurately determine the specific location of the pipeline leakage based on the acoustic emission sensor number and the distance between the leakage source and the sensor, and evacuate nearby personnel to repair the pipeline.

[0029] Compared with the prior art, one of the beneficial effects of the present invention is:

[0030] In the technical scheme of the present invention, 1) the leakage acoustic emission source access device has the function of accessing and enhancing the leakage signal. The designed leakage acoustic emission source access device and the analysis system with the leakage acoustic emission signal analysis algorithm are used, and the metal copper wire is used to transmit the acoustic emission signal, which overcomes the problems of weak acoustic emission signal and serious attenuation of the acoustic emission signal by the polymer material of the PE pipeline itself, greatly improves the monitoring distance, and can realize long-distance pipeline leakage monitoring; it has good engineering practicality and the device cost is low and has the prospect of large-scale application; 2) the genetic algorithm is used to optimize the VMD algorithm control parameters and combine the correlation coefficient to select the effective IMF component to remove the noise in the monitoring signal, which can realize the better removal of the noise of the collected signal. The denoising algorithm provided by the present invention ensures that the signal characteristics do not have excessive loss while removing the noise; 3) the SVM support vector machine is introduced to realize the classification of different signals, and the energy entropy characteristic value of the denoised signal is extracted as the SVM input parameter. The trained SVM can accurately classify different states. The classification and recognition method provided by the present invention can achieve 100% leakage signal monitoring and recognition accuracy within the monitoring range. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the embodiments of the present application document or the technical solutions in the prior art, the drawings required for use in the description of the embodiments or the prior art are briefly introduced below. Obviously, the drawings described below are only references to some embodiments in the present application document. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0032] Figure 1 This is a schematic diagram of the installation structure of the long-distance acoustic emission monitoring device proposed in the present invention.

[0033] Figure 2 This is a schematic diagram of the leakage monitoring process framework proposed in the present invention.

[0034] Figure 3 a. Figure 3 b and Figure 3 c is a schematic diagram of the results of optimizing VMD algorithm parameters using the genetic algorithm proposed in the present invention.

[0035] Figure 4 This is a schematic diagram of the original waveform of the pipeline leakage acoustic emission signal proposed by the present invention.

[0036] Figure 5 This is a schematic diagram of the denoising waveform of the pipeline leakage acoustic emission signal proposed by the present invention.

[0037] Figure 6 This is a schematic diagram of the energy entropy characteristic values ​​of various pipeline signals proposed by the present invention.

[0038] Figure 7This is a schematic diagram of the pipeline SVM classification and recognition results proposed in the present invention.

[0039] The names corresponding to the reference numerals in the accompanying drawings are:

[0040] 1-outlet gas tank, 2-check valve, 3-pump, 4-first control valve, 5-three-way auxiliary pipeline, 6-second control valve, 7-ball valve, 8-pressure gauge, 9-leakage acoustic emission source access device, 10-clip, 11-pipeline to be tested, 12-metal wire, 13-No. 2 ball valve, 14-inlet gas tank, 15-first node acoustic emission sensor, 16-second node acoustic emission sensor, 17-third node acoustic emission sensor, 18-fourth node acoustic emission sensor, 19-first acoustic emission preamplifier, 20-second acoustic emission preamplifier, 21-third acoustic emission preamplifier, 22-fourth acoustic emission preamplifier, 23-acoustic emission instrument, 24-data storage and display equipment. DETAILED DESCRIPTION

[0041] The present invention is further described in detail below in conjunction with examples, but the embodiments of the present invention are not limited thereto.

[0042] like Figure 1 As shown, Figure 1 The structure of a long-distance acoustic emission monitoring device for PE gas pipeline leakage in one embodiment is shown, which includes an outbound gas storage tank 1, a three-way auxiliary pipeline 5, a pipeline to be tested 11, an inbound gas storage tank 14, a leakage acoustic emission source access device 9, a buckle 10, a metal wire 12, a node acoustic emission sensor, an acoustic emission preamplifier, an acoustic emission instrument 23 and a data storage and display device 24; the outbound gas storage tank 1, the three-way auxiliary pipeline 5, the pipeline to be tested 11 and the inbound gas storage tank 14 are connected in sequence; the check valve 2, the pump 3, the first control valve 4 and the three-way auxiliary pipeline 5 are arranged in sequence along the pipeline of the outbound gas storage tank 1. A second control valve 6 is arranged on one side of the three-way auxiliary pipeline 5; a ball valve 7, a pressure gauge 8 and the pipeline to be tested 11 are arranged in sequence on the other side. There are several leakage acoustic emission source access devices 9, which are respectively arranged at the leakage-prone parts of the pipeline 11 to be tested; the number of buckles 10 corresponds to the leakage acoustic emission source access devices 9, and each buckle 10 is correspondingly arranged on both sides of a leakage acoustic emission source access device 9; there is one metal wire 12, and the metal wire 12 is connected to the buckle 10; a node acoustic emission sensor is connected to the metal wire 12 between every two leakage acoustic emission source access devices 9; each of the node acoustic emission sensors is further connected to an acoustic emission preamplifier; an acoustic emission instrument 23 is connected to the acoustic emission preamplifier; a data storage and display device 24 is connected to the acoustic emission instrument 23, and the data storage and display device 24 is used to store acoustic emission monitoring data and display module waveform diagrams to complete the long-distance acoustic emission monitoring process of the leakage of the pipeline 11 to be tested.

[0043] Figure 1 In the embodiment, the node acoustic emission sensor may include a first node acoustic emission sensor 15, a second node acoustic emission sensor 16, a third node acoustic emission sensor 17 and a fourth node acoustic emission sensor 18; the acoustic emission preamplifier may include a first acoustic emission preamplifier 19, a second acoustic emission preamplifier 20, a third acoustic emission preamplifier 21 and a fourth acoustic emission preamplifier 22. The node acoustic emission sensors located at each node are respectively connected to an acoustic emission preamplifier for amplifying the acoustic emission signal. The number of the above devices is determined according to the actual situation, and the present invention is not limited thereto.

[0044] The above-mentioned node acoustic emission sensors are arranged above the corresponding metal wires 12 for collecting acoustic emission signals. The acoustic emission sensor of each node has a unique number, and the acoustic emission sensor is used to conveniently locate the leakage point, that is, to quickly locate the node acoustic emission sensor with a certain number.

[0045] A preferred solution is that the spacing distance between the node acoustic emission sensors connected to each of the metal wires 12 is 50-100 meters. Among them, the metal wire 12 is preferably a copper wire. The connection of the metal copper wire can overcome the serious attenuation problem of the acoustic emission signal transmitted by the PE polymer material, extend the propagation distance of the acoustic emission signal, and avoid the situation of missed detection due to signal attenuation when a leak occurs. It has a simple structure, simple signal transmission, low cost, and a wide range of applications, but the present invention is not limited to this.

[0046] The leakage acoustic emission source access device 9 preferably adopts a pipeline metal wrapped mesh, which is arranged at the leak-prone part of the pipeline 11 to be tested, and the two sides of the pipeline leakage acoustic emission source access device 9 are positioned and fixed with buckles 10; the buckles 10 can make the copper wire more firmly fixed on the leakage acoustic emission source access device 9. The leakage acoustic emission source access device 9 can enhance the acoustic emission signal and avoid the problem of weak acoustic emission signal.

[0047] The leak-prone parts in the above embodiments include but are not limited to leak-prone welds, flange connections and valves.

[0048] The acoustic emission instrument 23 is embedded with an acoustic emission signal analysis system, which includes an acoustic emission sensor for identifying abnormal signal categories based on an SVM support vector machine and locating nodes corresponding to leakage signals. The signal category recognition algorithm based on the SVM support vector machine is used to identify normal working signals, interference signals and leakage signals.

[0049] Recombination Figure 2-7 As shown, a long-distance acoustic emission monitoring method for PE gas pipeline leakage includes:

[0050] S1: The leakage acoustic emission source access device 9 is put on the leak-prone part of the monitored pipeline, the signal transmission metal wire 12 is inserted into the buckle 10 on the side of the leakage acoustic emission source access device 9, and the metal wire 12 is connected to the node acoustic emission sensor, and then the signal is amplified by the acoustic emission preamplifier and transmitted to the acoustic emission instrument 23; that is, the device and software system designed by the present invention are used to obtain the collected working state acoustic emission data of the pipeline to be tested; its working data includes the acoustic emission signal during normal operation, the acoustic emission signal when leakage occurs, and the acoustic emission signal when interference exists but there is no leakage.

[0051] S2: The acoustic emission signal analysis system uses a genetic algorithm to optimize the VMD algorithm control parameters [K, α], inputs the obtained optimal parameters [K, α] into the VMD algorithm, decomposes the three types of signals (normal working signals, interference signals and leakage signals) collected by the acoustic emission instrument 23 in step S1, selects effective IMF components through correlation coefficients, reconstructs the signal to remove noise, and extracts the energy entropy eigenvalue of the reconstructed signal to train the SVM support vector machine.

[0052] Among them, the number of iterations of the genetic algorithm is 30, the number of populations is 20; the crossover and mutation probabilities are 0.8 and 0.2; the decomposition level K and the penalty factor α are set in the range of [3:20, 1000:4000]; the objective function is the local minimum of the envelope entropy; and the IMF component with a correlation coefficient greater than or equal to 0.3 is selected as the effective component; the kernel function used by the SVM support vector machine is the RBF kernel function.

[0053] S3: Utilize multiple node acoustic emission sensors to continuously monitor the operating conditions of multiple pipelines to be tested, and the data storage and display device 24 saves the abnormal signals found. The VMD algorithm optimized in step S2 is combined with the correlation coefficient to reconstruct the signal to complete the denoising process of the abnormal signal. The energy entropy eigenvalue of the reconstructed signal is extracted and input into the SVM support vector machine prediction model trained in step S2 to complete the classification and identification of the abnormal signal.

[0054] S4: After using the SVM support vector machine prediction model to identify the signal category collected in real time, an early warning signal is issued and processing is guided according to the identification results. If it is a normal operation signal of the pipeline, continue monitoring; if a long-term interference signal is found, arrange staff to find the cause of the abnormal signal emitted by the pipeline near the corresponding sensor according to the number of the acoustic emission sensor of multiple nodes; if it is a leakage signal, close the valve immediately, quickly and accurately determine the specific location of the pipeline leakage according to the acoustic emission sensor number and the distance between the leakage source and the sensor, and evacuate nearby personnel to repair the pipeline.

[0055] In one embodiment, for example, there is a 2mm leak hole near the weld of the pipeline 11 to be tested, and the gas impacts the pipeline. The leakage acoustic emission source access device 9 transmits the vibration acoustic emission signal to the metal copper wire 12. The acoustic emission sensor monitors the metal copper wire 12 to obtain the leakage signal time domain information, converts it into data form for storage, and the acoustic emission sensor collects and stores the time domain information of the normal operation signal of the pipeline. The acoustic emission sensor also collects and stores the time domain information of the interference signal under various interferences, such as knocking on the pipeline, the environment near the pipeline making sounds, etc., to obtain the acoustic emission data under the typical operation state of the pipeline.

[0056] In order to remove various noises contained in the acoustic emission signal, a genetic algorithm is proposed to optimize the VMD algorithm parameters and combine the correlation coefficient to denoise the signal. Genetic algorithm: It can effectively solve the global nonlinear optimization problem and quickly and accurately obtain the optimal parameters of the VMD decomposition algorithm.

[0057] VMD algorithm: An adaptive quasi-orthogonal nonlinear processing algorithm that can effectively avoid modal aliasing and endpoint effects, has good robustness, and is very suitable for analyzing non-stationary characteristic signals.

[0058] Correlation coefficient: , where: Cov (X, Y) is the covariance of random variables X and Y; , The standard deviation of the random variables X and Y respectively.

[0059] The IMF components with correlation coefficients greater than or equal to 0.3 are selected as effective components for signal reconstruction, that is, to remove the noise in the original signal.

[0060] After denoising the signal, it is proposed to use SVM support vector machine to classify and identify the signal.

[0061] The kernel function used by SVM is RBF kernel function:

[0062] In the formula, is the feature vector of two sample points, is the dot product of two sample points, is the kernel parameter ( ).

[0063] SVM requires characteristic parameters of the input signal. Effective characteristic parameters can reduce the probability of false alarms and missed alarms, which plays a vital role in the monitoring accuracy of pipeline leakage. It is proposed to use the energy entropy characteristic values ​​of various pipeline signals after denoising for SVM training. Among them, the energy entropy characteristic values ​​of various signals are used for SVM training, and the trained SVM can be used to classify and identify unknown pipeline signals.

[0064] Specifically, the working principle of the long-distance acoustic emission monitoring device is:

[0065] 1) Determine the need to monitor the pipeline working conditions, refer to Figure 2 The process first collects acoustic emission signals of various operating conditions of the monitored pipeline, and uses genetic algorithm to optimize VMD algorithm operation. Figure 3 This is the optimization result of the genetic algorithm for the VMD decomposition algorithm of the pipeline leakage signal to be tested. Then, the optimized VMD algorithm is used to decompose the signals of each working condition, and the effective IMF components are selected through the correlation coefficient method to reconstruct the signal to complete the noise removal. Figure 4 The original waveform of the acoustic emission signal of the pipeline leakage to be tested, reference Figure 5 This is the denoised waveform of the acoustic emission signal of the pipeline leakage to be tested. The noise contained in the original leakage signal is effectively removed, and the leakage signal characteristics are well preserved, with excellent denoising effect. Figure 6 This is a schematic diagram of the energy entropy characteristic value data of various denoised signals. It can be seen that the energy entropy of the three signals is quite different, and the entropy curves are obviously different. There is no intersection or close proximity between the normal signal, the knocking signal and the leakage signal. The three signals are input into the SVM support vector machine for training. The trained SVM support vector machine can complete the classification and recognition of unknown abnormal signals of the pipeline to be tested.

[0066] 2) When monitoring the pipeline, the acoustic emission sensor continuously collects signals. When an abnormal signal appears, it immediately alarms and stores the abnormal signal. When processing abnormal signals, the genetic algorithm is no longer used to optimize the VMD algorithm. The optimized VMD algorithm in 1 can be used. After decomposing the abnormal signal, the effective IMF component is selected through the correlation coefficient to reconstruct the signal, and the energy entropy eigenvalue is extracted and input into the trained SVM support vector machine to complete the classification and identification of the abnormal signal. In one example, 100 groups of data of three types of signals, including normal signals, interference signals and leakage signals, were collected in advance through the pipeline leakage monitoring platform. Energy entropy was selected as the characteristic parameter, and the sample labels of the three types of signals were set to 1, 2, and 3 respectively. 60 groups of data from each type of signal were taken to participate in the training of the prediction model, and then part of the remaining data of the three types of signals were mixed together as test samples (20 groups of data for type 1, 40 groups of data for type 2, and 20 groups of data for type 3, a total of 100 groups of data), and input into the trained SVM prediction model. The predicted classification results are as follows: Figure 7 As shown, Figure 7 The symbols of "○ Actual test set type" and "* Predicted test set type" in Figure 7 The basic overlap is due to Figure 7 It can be seen that the gas pipeline has a 100% recognition accuracy rate for the three types of signals within the monitoring range of the acoustic emission sensor, which means that the prediction model can completely distinguish between normal working conditions (no leakage), external knocking interference and leakage signals.

[0067] 3) After identifying the abnormal signal, if it is a normal signal or an interference signal, the alarm will be lifted in time. If it is a leakage signal, the leakage site will be located by relying on the acoustic emission sensor numbers upstream and downstream of the leakage point. After the leakage point is located, staff will be dispatched to handle the pipeline on site, improve the efficiency of leakage emergency response, and monitor the pipeline operation in real time.

[0068] In summary, the long-distance acoustic emission monitoring device and method for PE gas pipeline leakage of the present invention realizes the long-distance monitoring process of non-metallic PE gas pipeline leakage accidents, analyzes and classifies abnormal signals that may appear in the pipeline, can accurately identify leakage signals from various typical signals, and can locate the leakage point, thereby achieving real-time and accurate monitoring.

[0069] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0070] When referred to in this specification, "one embodiment", "another embodiment", "an embodiment", etc., it means that the specific features, structures or characteristics described in conjunction with the embodiment are included in at least one embodiment generally described in this application. The same expression appearing in multiple places in the specification does not necessarily refer to the same embodiment. Further, when a specific feature, structure or characteristic is described in conjunction with any embodiment, it is claimed that the realization of such feature, structure or characteristic in conjunction with other embodiments also falls within the scope of the present invention.

[0071] Although the present invention is described herein with reference to a number of illustrative embodiments of the present invention, it will be appreciated that those skilled in the art may devise many other modifications and implementations that fall within the scope and spirit of the principles disclosed herein. More specifically, within the scope of the present disclosure, drawings, and claims, a variety of variations and improvements may be made to the components and / or layout of the subject combination layout. In addition to variations and improvements made to the components and / or layout, other uses will also be apparent to those skilled in the art.

Claims

1. A long-distance acoustic emission monitoring device for PE gas pipeline leakage, characterized in that: include: A plurality of leakage acoustic emission source access devices (9) are provided respectively at leak-prone locations of the pipeline to be tested (11); Buckles (10), the number of which corresponds to the leakage acoustic emission source access device (9), and each buckle is correspondingly arranged on both sides of a leakage acoustic emission source access device (9); A metal wire (12), the number of which is one, the metal wire (12) being connected to the buckle (10); a node acoustic emission sensor is connected to the metal wire (12) between every two leakage acoustic emission source access devices (9); and each of the node acoustic emission sensors is further connected to an acoustic emission preamplifier respectively; An acoustic emission instrument (23) is connected to an acoustic emission preamplifier; an acoustic emission signal analysis system is embedded in the acoustic emission instrument (23), the acoustic emission signal analysis system includes a signal category recognition algorithm based on an SVM support vector machine, and the prediction model based on the SVM support vector machine is used to identify normal working signals, interference signals and leakage signals; the training based on the SVM support vector machine includes: taking the collected normal signals, interference signals and leakage signals as original signals, optimizing the VMD algorithm control parameters [K, α] by using a genetic algorithm, wherein K is the number of decomposition layers and α is a penalty factor; inputting the obtained set of optimal [K, α] parameters into the VMD algorithm, decomposing the collected three types of signals, selecting effective IMF components through correlation coefficients, reconstructing the signal to complete noise removal, and extracting the energy entropy eigenvalue of the reconstructed signal to train the SVM support vector machine prediction model; the range of the control parameters [K, α] is set to [3:20, 1000:4000]; the correlation coefficient: Where: Cov(X,Y) is the covariance of random variables X and Y; σ X , σ Y The standard deviation of random variables X and Y respectively; IMF components with correlation coefficients greater than or equal to 0.3 are selected as effective components for signal reconstruction; A data storage and display device (24) is connected to the acoustic emission instrument (23).

2. The long-distance acoustic emission monitoring device for PE gas pipeline leakage according to claim 1 is characterized in that: The plurality of node acoustic emission sensors have unique numbers.

3. The long-distance acoustic emission monitoring device for PE gas pipeline leakage according to claim 1 is characterized in that: The node acoustic emission sensors connected to each of the metal wires (12) are spaced apart at a distance of 50-100 meters.

4. The long-distance acoustic emission monitoring device for PE gas pipeline leakage according to claim 1 is characterized in that: The metal wire (12) is a copper wire.

5. The long-distance acoustic emission monitoring device for PE gas pipeline leakage according to claim 1 is characterized in that: The leakage acoustic emission source access device (9) is a pipeline metal wrapping net.

6. A long-distance acoustic emission monitoring method for PE gas pipeline leakage, characterized in that: include: S1: The leakage acoustic emission source access device (9) is placed on the leak-prone part of the monitored pipeline, the signal transmission metal wire (12) is inserted into the buckle (10) on the side of the leakage acoustic emission source access device (9), and the metal wire (12) is connected to the node acoustic emission sensor, and then the signal is amplified by the acoustic emission preamplifier and transmitted to the acoustic emission instrument (23); S2: an acoustic emission signal analysis system is embedded in the acoustic emission instrument (23), and the acoustic emission signal analysis system optimizes the VMD algorithm control parameters [K, α] by using a genetic algorithm, and inputs the obtained optimal parameters [K, α] into the VMD algorithm, decomposing the three types of signals collected by the acoustic emission instrument (23) in step S1, and selecting effective IMF components through correlation coefficients, reconstructing the signal to remove noise, and extracting the energy entropy eigenvalue of the reconstructed signal to train the SVM support vector machine prediction model; S3: using multiple node acoustic emission sensors to continuously monitor the operating conditions of multiple pipelines to be tested, and using the data storage and display device (24) to save the abnormal signals found, and using the VMD algorithm optimized in step S2 combined with the correlation coefficient to reconstruct the signal to complete the denoising process of the abnormal signal, extracting the energy entropy feature value of the reconstructed signal and inputting it into the SVM support vector machine trained in step S2 to complete the classification and identification of the abnormal signal; S4: After using the SVM support vector machine prediction model to identify the signal category collected in real time, an early warning signal is issued and processing is guided according to the identification result; If it is a signal of normal operation of the pipeline, continue monitoring; if a long-term interference signal is found, arrange staff to find the cause of the abnormal signal emitted by the pipeline near the corresponding sensor according to the numbers of multiple node acoustic emission sensors; if it is a leakage signal, close the valve immediately, and quickly and accurately determine the specific location of the pipeline leakage based on the acoustic emission sensor number and the distance between the leakage source and the sensor, and evacuate nearby personnel to repair the pipeline.

Images