A method, device and equipment for online positioning and identification of pipeline leakage crack morphology

Abstract

The present invention discloses a method, device, and equipment for online positioning and identification of pipeline leakage crack morphology. The method comprises pre-emphasis processing on the downsampled acoustic emission signal and performing a frame short-time Fourier transform on the pre-emphasized acoustic emission signal to obtain the acoustic emission signal spectrum characteristics. A set number of Mel filter banks are designed based on the maximum frequency of the acoustic emission signal spectrum characteristics to filter the acoustic emission signal spectrum. A logarithmic operation is performed on the filtered acoustic emission signal spectrum to obtain FBank features. The GI index of the FBank features is calculated, and the first n-dimensional features in the GI index are selected, where n = 30 to 60. The first n-dimensional features are input into a pre-trained pipeline leakage crack morphology recognition model to obtain the pipeline leakage crack morphology. The present invention can utilize the acoustic emission phenomenon inside the material during the leakage process to perform defect detection, reducing the computational workload and preserving the original signal information to the greatest extent.

Description

Technical Field

[0001] The present invention belongs to the field of pipeline leakage detection, and in particular relates to a method, device and equipment for online positioning and identification of pipeline leakage crack morphology. Background Art

[0002] In pipeline leakage monitoring, especially in special areas such as nuclear power water pipelines, early warning of pipeline leakage is very important. Currently, the main method for diagnosing leakage faults is the leakage threshold. However, different operating conditions (different cracks) have different times when the threshold is reached. Therefore, this method has great limitations, which will significantly increase the risk of water loss faced by equipment during service. In addition, leakage data is very rare in engineering, and occasional abnormal data is insufficient to support operation and maintenance analysis. At the same time, the stability of threshold warning is poor, and it is impossible to accurately find the abnormal leakage state of the pipeline or predict the time when the threshold is reached.

[0003] Currently, no research has been found on pipeline leak crack morphology recognition. Acoustic emission technology is usually used to detect pipeline leaks. For example, Dong Xingliang et al. proposed an underwater pipeline leak acoustic emission detection method based on convolutional neural networks. This method converts the leakage waveform signal and the background noise leakage waveform signal into two-dimensional image samples. By optimizing the network structure and hyperparameters, a convolutional neural network underwater pipeline leak detection model is constructed. However, the extracted two-dimensional image samples have a large amount of information, high redundancy, and slow computation speed, making online analysis difficult. Hao Yongmei et al. proposed a method for processing acquired signals by combining modal acoustic emission technology and wavelet transform. This method introduces independent component analysis technology and uses the FastICA toolbox to separate leakage source signals. This method aims to improve the leakage information in acoustic emission signals. However, the problem is that using modal analysis to directly separate and analyze leakage signals will inevitably lose some leakage information, which not only significantly affects subsequent feature extraction but also makes it difficult to accurately locate the leak location. Summary of the Invention

[0004] In response to the problems existing in the prior art, the present invention provides a method, device and equipment for online positioning and identification of pipeline leakage crack morphology, which can utilize the acoustic emission phenomenon inside the material during the leakage process to perform defect detection, thereby reducing the workload of calculations and retaining the information of the original signal to the greatest extent.

[0005] In order to solve the above technical problems, the present invention is implemented through the following technical solutions:

[0006] A method for online positioning and identification of pipeline leakage crack morphology, comprising:

[0007] Acquiring an acoustic emission signal generated when a pipeline leaks, and locating the leak location on the pipeline based on the acoustic emission signal;

[0008] downsampling the acoustic emission signal to obtain a downsampled acoustic emission signal;

[0009] Performing pre-emphasis processing on the down-sampled acoustic emission signal, and performing frame-based short-time Fourier transform on the pre-emphasized acoustic emission signal to obtain a spectrum feature of the acoustic emission signal;

[0010] Design a set number of Mel filter banks according to the maximum frequency of the acoustic emission signal spectrum feature to filter the acoustic emission signal spectrum, and perform a logarithmic operation on the filtered acoustic emission signal spectrum to obtain an FBank feature;

[0011] Calculate the GI index of the FBank feature and select the first n-dimensional features in the GI index, where n=30~60;

[0012] The first n-dimensional features are input into a pre-trained pipeline leakage crack morphology recognition model to obtain the pipeline leakage crack morphology.

[0013] Furthermore, the downsampled acoustic emission signal is pre-emphasized, and the calculation formula is:

[0014]

[0015] Where, For time, is the acoustic emission signal after pre-emphasis processing, for The acoustic emission signal after downsampling at each moment, for The acoustic emission signal after downsampling at each moment, is the pre-emphasis coefficient.

[0016] Furthermore, the acoustic emission signal after pre-emphasis processing is subjected to frame short-time Fourier transform to obtain the spectrum characteristics of the acoustic emission signal, and the calculation formula is:

[0017]

[0018] Where, is the window function, is the center of the window function, is an imaginary number, is the spectrum characteristic of the acoustic emission signal.

[0019] Furthermore, the Mel filter banks of a set number are designed according to the maximum frequency of the spectrum characteristics of the acoustic emission signal, and the calculation formula is:

[0020]

[0021] Where, For the filters, It corresponds to The center frequency of the filter, For the The value after the Mel scale of the filter is converted to frequency, For the The value after the Mel scale of the filter is converted to frequency, For the The value after the Mel scale of the filter is converted to frequency.

[0022] Furthermore, the logarithmic operation is performed on the spectrum of the filtered acoustic emission signal to obtain the FBank feature, and the calculation formula is:

[0023]

[0024] Where, is the characteristic length of the spectrum, It is through the The logarithmic energy of the output of the filter is the FBank feature.

[0025] Furthermore, the GI index of the FBank feature is calculated using the following formula:

[0026]

[0027] Where, For the Logarithmic energy norm, for Rearrange the data from small to large, yes No. Features, yes No. Features, is the GI index.

[0028] Furthermore, the training method of the pipeline leakage crack morphology recognition model is as follows:

[0029] Acquire the acoustic emission signals generated when pipelines with different leakage crack morphologies leak;

[0030] Downsampling the acoustic emission signals corresponding to different leakage crack morphologies is performed respectively to obtain downsampled acoustic emission signals corresponding to different leakage crack morphologies;

[0031] The downsampled acoustic emission signals corresponding to different leakage crack morphologies are pre-emphasized, and the pre-emphasized acoustic emission signals are subjected to frame-by-frame short-time Fourier transform to obtain the spectrum characteristics of the acoustic emission signals corresponding to different leakage crack morphologies.

[0032] According to the maximum frequency of the acoustic emission signal spectrum characteristics corresponding to different leakage crack morphologies, a set number of Mel filter banks are designed to filter the respective acoustic emission signal spectra. Logarithmic operation is performed on the filtered acoustic emission signal spectra to obtain the FBank characteristics corresponding to different leakage crack morphologies.

[0033] The GI index of the FBank features corresponding to different leakage crack morphologies is calculated respectively, and the first n-dimensional features in each GI index are selected, where n = 30~60;

[0034] The first n-dimensional features of each GI index and the corresponding leakage crack morphology are used as training samples, and a one-dimensional convolutional neural network based on DenseNet is trained to obtain the pipeline leakage crack morphology recognition model.

[0035] Furthermore, locating the leakage position on the pipeline according to the acoustic emission signal includes:

[0036] The time difference method is used to locate the leak position on the pipeline based on the acoustic emission signal. The calculation formula is as follows:

[0037]

[0038] Where, is the distance from the leak point to the sensor, is the sensor spacing, is the time difference between the acoustic emission wave and the two sensors, is the acoustic emission wave velocity, which is calculated through the commonly used lead breaking experiment.

[0039] An online positioning and identification device for pipeline leakage crack morphology, comprising:

[0040] An acquisition module is used to acquire an acoustic emission signal generated when a pipeline leaks, and locate the leak position on the pipeline according to the acoustic emission signal;

[0041] A downsampling module is used to downsample the acoustic emission signal to obtain a downsampled acoustic emission signal;

[0042] A spectrum feature module is used to perform pre-emphasis processing on the downsampled acoustic emission signal and perform frame short-time Fourier transform on the pre-emphasized acoustic emission signal to obtain a spectrum feature of the acoustic emission signal;

[0043] An FBank feature module is used to design a set number of Mel filter banks according to the maximum frequency of the acoustic emission signal spectrum feature to filter the acoustic emission signal spectrum, and perform logarithmic operation on the filtered acoustic emission signal spectrum to obtain FBank features;

[0044] GI index module, used to calculate the GI index of the FBank feature and select the first n-dimensional features in the GI index, where n=30~60;

[0045] The recognition module is used to input the first n-dimensional features into a pre-trained pipeline leakage crack morphology recognition model to obtain the pipeline leakage crack morphology.

[0046] A device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the method for online positioning and identifying pipeline leakage crack morphology are implemented.

[0047] Compared with the prior art, the present invention has at least the following beneficial effects:

[0048] The present invention provides an online positioning and identification method for pipeline leakage crack morphology. It uses the acoustic emission phenomenon inside the material during the leakage process to perform defect detection, which helps to detect small defects in the material. Pre-emphasis processing is performed on the acoustic emission signal to balance the spectral information and improve the signal-to-noise ratio. Since the deep learning network is insensitive to highly correlated information, the introduction of FBank features has better effects than the commonly used MFCC features, which helps to improve the accuracy of recognition. Finally, the introduction of the GI index reduces the workload of calculations while retaining the information of the original signal to the greatest extent. The calculation method of the present invention is simple and fast, with high feature discrimination, strong robustness, and high engineering practicality. It provides an effective technical implementation path for the identification of pipeline leakage with cracks of different morphologies and can accurately locate the source of the leakage.

[0049] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the technical solutions in the specific embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the specific embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0051] Figure 1The present invention is a flow chart of a method for online positioning and identification of pipeline leakage crack morphology.

[0052] Figure 2 Schematic diagram of the acoustic emission signal acquisition system in the embodiment.

[0053] Figure 3 Figure 2 is a pipe diagram of cracks with different morphologies in the embodiment.

[0054] Figure 4a This is the waveform of the acoustic emission signal after downsampling in the embodiment. Figure 4b 4 is a spectrum diagram after downsampling in the embodiment.

[0055] Figure 5 This is a spectrum diagram of the acoustic emission signal of the axial crack pipeline leakage after pre-emphasis processing in the embodiment.

[0056] Figure 6a This is the FBank characteristic diagram of the axial crack pipeline leakage in the embodiment, Figure 6b This is the FBank characteristic diagram of the zigzag crack pipeline leakage in the embodiment, Figure 6c This is the FBank characteristic diagram of circumferential crack pipeline leakage in the embodiment.

[0057] Figure 7 It is a broken line graph of the 50 GI characteristics in the axial, flexural and circumferential directions in the embodiment.

[0058] Figure 8a This is a one-dimensional convolutional neural network structure diagram based on DenseNet in the embodiment. Figure 8b This is a block structure diagram of the Dense Block in the DenseNet network in the embodiment. DETAILED DESCRIPTION

[0059] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of them. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0060] As a specific embodiment of the present invention, Figure 1 As shown, this embodiment provides a method for online positioning and identification of pipeline leakage crack morphology, which specifically includes the following steps:

[0061] S1. Acquire an acoustic emission signal generated when a pipeline leaks, and locate the leak position on the pipeline according to the acoustic emission signal.

[0062] Specifically, an acoustic emission sensor is used to obtain the acoustic emission signal generated when the pipeline leaks.

[0063] Preferably, the time difference method is used to locate the leakage position on the pipeline according to the acoustic emission signal, and the calculation formula is as follows:

[0064]

[0065] Where, is the distance from the leak point to the sensor, is the sensor distance, and the time difference between the acoustic emission wave and the two sensors is , is the acoustic emission wave velocity, which is calculated through the commonly used lead breaking experiment.

[0066] S2. Downsampling the acoustic emission signal to obtain a downsampled acoustic emission signal.

[0067] Preferably, downsampling is performed according to Shannon's theorem to obtain an undistorted acoustic emission signal, that is, a downsampled acoustic emission signal.

[0068] S3. Performing pre-emphasis processing on the down-sampled acoustic emission signal, and performing frame-based short-time Fourier transform on the pre-emphasized acoustic emission signal to obtain a spectrum feature of the acoustic emission signal.

[0069] Preferably, the downsampled acoustic emission signal is pre-emphasized, and the calculation formula is:

[0070]

[0071] Where, For time, is the acoustic emission signal after pre-emphasis processing, for The acoustic emission signal after downsampling at each moment, for The acoustic emission signal after downsampling at each moment, is the pre-emphasis coefficient.

[0072] Preferably, the acoustic emission signal after pre-emphasis processing is subjected to a frame-based short-time Fourier transform to obtain the spectrum characteristics of the acoustic emission signal, and the calculation formula is:

[0073]

[0074] Where, is the window function, is the center of the window function, is an imaginary number, is the spectrum characteristic of the acoustic emission signal.

[0075] S4. Design a set number of Mel filter banks according to the maximum frequency of the acoustic emission signal spectrum feature to filter the acoustic emission signal spectrum, perform a logarithmic operation on the filtered acoustic emission signal spectrum, and obtain an FBank feature.

[0076] Preferably, a set number of Mel filter banks are designed according to the maximum frequency of the acoustic emission signal spectrum characteristics, and the calculation formula is:

[0077]

[0078] Where, For the filters, It corresponds to The center frequency of the filter, For the The value after the Mel scale of the filter is converted to frequency, For the The value after the Mel scale of the filter is converted to frequency, For the The value after the Mel scale of the filter is converted to frequency.

[0079] Preferably, a logarithmic operation is performed on the spectrum of the filtered acoustic emission signal to obtain the FBank feature, and the calculation formula is:

[0080]

[0081] Where, is the characteristic length of the spectrum, It is through the The logarithmic energy of the output of the filter is the FBank feature.

[0082] S5. Calculate the GI index of the FBank feature and select the first n-dimensional features in the GI index, where n=30-60.

[0083] Preferably, the GI index of the FBank feature is calculated using the following formula:

[0084]

[0085] Where, For the Logarithmic energy norm, for Rearrange the data from small to large, yes No. Features, yes No. Features, is the GI index.

[0086] S6. Input the first n-dimensional features into a pre-trained pipeline leakage crack morphology recognition model to obtain the pipeline leakage crack morphology.

[0087] Preferably, the training method of the pipeline leakage crack morphology recognition model is as follows:

[0088] a. Obtain the acoustic emission signals generated when a pipeline with different leakage crack morphologies leaks.

[0089] Specifically, the sampling frequency of the pipeline leakage acoustic emission signal and the distance of the acoustic emission sensor are determined according to the acoustic emission signal type and attenuation characteristics, and the acoustic emission signal dataset of pipeline leakage with different crack morphologies is obtained by reasonably designing the sample length.

[0090] b. downsampling the acoustic emission signals corresponding to different leakage crack morphologies to obtain downsampled acoustic emission signals corresponding to different leakage crack morphologies;

[0091] c. Pre-emphasize the downsampled acoustic emission signals corresponding to different leakage crack morphologies, and perform frame-by-frame short-time Fourier transform on the pre-emphasized acoustic emission signals to obtain the spectral characteristics of the acoustic emission signals corresponding to different leakage crack morphologies;

[0092] d. Design a set number of Mel filter banks based on the maximum frequency of the acoustic emission signal spectrum characteristics corresponding to different leakage crack morphologies to filter the respective acoustic emission signal spectra, perform logarithmic operation on the respective filtered acoustic emission signal spectra, and obtain FBank features corresponding to different leakage crack morphologies;

[0093] e. Calculate the GI index of the FBank features corresponding to different leakage crack morphologies respectively, and select the first n-dimensional features in each GI index, where n = 30~60;

[0094] f. The first n-dimensional features of each GI index and the corresponding leakage crack morphology are used as training samples to train a one-dimensional convolutional neural network based on DenseNet, as shown in Figure 8, to obtain the pipeline leakage crack morphology recognition model.

[0095] The present invention can fully utilize the information of acoustic emission signals, extract the essential characteristics of pipeline leakage acoustic emission signals, combine with deep learning networks, and construct an accurate identification model, thereby realizing accurate identification of pipeline leakage acoustic emission signals under different working conditions, thereby providing an accurate and reliable method for online positioning and identification of pipeline leakage crack morphology.

[0096] Example:

[0097] like Figure 2 The figure shows a schematic diagram of the acoustic emission signal acquisition system for pipeline leakage in this embodiment, which mainly includes an industrial computer 1, a data acquisition card 2, a preamplifier 3, an acoustic emission sensor 4, a leaking fault pipeline 5 and a prefabricated crack 6. Among them, the industrial computer 1, the data acquisition card 2, the preamplifier 3 and the acoustic emission sensor 4 are connected in sequence to form an acoustic emission signal acquisition system. During the pipeline leakage process, the acoustic emission signal acquisition system collects acoustic emission signals in real time.

[0098] In this embodiment, the acoustic emission sensor 4 is 50 mm from the center of the leak. The leaking fault pipe 5 used in the embodiment is made by butt-welding two 50 cm galvanized steel pipes. The pipe diameter is 60 mm and the wall thickness is 4 mm. The two ends of the pipe are flanged. Industrial coupling agent is used to ensure a tight fit between the acoustic emission sensor 4 and the leaking fault pipe 5. Adhesive tape is used to ensure that the acoustic emission sensor 4 does not move during the leakage of the leaking fault pipe 5. The sampling rate is 1 MHz. In this embodiment, three pipes with different crack morphologies are used: axial cracks, zigzag cracks, and circumferential cracks. Figure 3 shown.

[0099] Step 1: The liquid medium is pressurized by the booster pump and flows into the water inlet of the leaking pipe. Due to the presence of cracks, an acoustic emission signal is generated inside the material. The acoustic emission sensor collects the signal synchronously, calculates the wave velocity of the acoustic emission signal based on the lead breaking experiment, and locates the leak source using the time difference method. Under the condition that the Shannon sampling theorem is satisfied, the original acoustic emission signal is downsampled by 2 times to obtain an undistorted downsampled signal, that is, the downsampled acoustic emission signal. The sampling rate is 500KHz. Figure 4a Figure 4 shows the waveform of the acoustic emission signal after downsampling preprocessing, and Figure 4b is the spectrum diagram after downsampling. It can be seen that the high-frequency information of the signal is weak at this time, while the pipeline leakage information is concentrated in the high-frequency band, so further processing is required.

[0100] Step 2: Pre-emphasize the downsampled acoustic emission signal, that is, apply a pre-emphasis filter to the downsampled acoustic emission signal to amplify the high-frequency part to balance the spectrum. This method can balance the high-frequency signal and improve the signal-to-noise ratio. Figure 5 This is the spectrum of the pre-emphasized acoustic emission signal of an axial crack pipeline leak. High-frequency information is amplified. The spectral characteristics of the pre-emphasized acoustic emission signal are obtained by performing a frame-by-frame short-time Fourier transform on the pre-emphasized acoustic emission signal.

[0101] Step 3: Design 128 Mel filter banks based on the maximum frequency of the acoustic emission signal spectrum characteristics, filter the acoustic emission signal spectrum, and perform logarithmic operation on the filtered acoustic emission signal spectrum to obtain the FBank characteristics of the acoustic emission signal. Figure 6a is the FBank characteristic diagram of axial crack pipeline leakage, Figure 6b is the FBank characteristic diagram of the zigzag crack pipeline leakage, Figure 6c These are the FBank characteristic diagrams of circumferential crack pipeline leakage. It can be seen from these three figures that different leakage forms have different frequency characteristics. Among them, circumferential cracks have multiple frequency centers, while axial cracks and flexural cracks have different frequency distribution widths.

[0102] Step 4: By calculating the GI index of the FBank feature of the pipeline leakage acoustic emission signal with different crack morphologies, the FBank feature corresponding to the largest GI index in the first 50 dimensions is selected to complete the dimensionality reduction, such as Figure 7 The figure shows the characteristic line graphs of 50 GI indices in the axial, flexural and circumferential directions. It can be seen that the three different crack forms have good distinction.

[0103] Step 5: Build Figure 8a 、 8b The one-dimensional convolutional neural network DenseNet shown in Figure 1 completes the classification and recognition of input features with an accuracy rate of up to 96.3%.

[0104] As demonstrated by the aforementioned experiments and examples, the present invention fully utilizes the acoustic information of acoustic emission signals and employs the GI index to reduce the dimensionality of FBank features, ultimately achieving feature extraction and identification of pipeline leaks with cracks of varying morphologies, with an accuracy rate of 96.3%. The proposed method is simple, has clear physical meaning for its features, and exhibits high reliability and robustness, providing an effective approach for online localization and identification of pipeline leak crack morphologies.

[0105] This embodiment further provides an online positioning and identification device for pipeline leakage crack morphology, which is used to implement an online positioning and identification method for pipeline leakage crack morphology, including:

[0106] An acquisition module is used to acquire an acoustic emission signal generated when a pipeline leaks, and locate the leak position on the pipeline according to the acoustic emission signal;

[0107] A downsampling module is used to downsample the acoustic emission signal to obtain a downsampled acoustic emission signal;

[0108] A spectrum feature module is used to perform pre-emphasis processing on the downsampled acoustic emission signal and perform frame short-time Fourier transform on the pre-emphasized acoustic emission signal to obtain a spectrum feature of the acoustic emission signal;

[0109] An FBank feature module is used to design a set number of Mel filter banks according to the maximum frequency of the acoustic emission signal spectrum feature to filter the acoustic emission signal spectrum, and perform logarithmic operation on the filtered acoustic emission signal spectrum to obtain FBank features;

[0110] GI index module, used to calculate the GI index of the FBank feature and select the first n-dimensional features in the GI index, where n=30~60;

[0111] The recognition module is used to input the first n-dimensional features into a pre-trained pipeline leakage crack morphology recognition model to obtain the pipeline leakage crack morphology.

[0112] In one embodiment of the present invention, a computer device is provided, comprising a processor and a memory, wherein the memory is used to store a computer program, the computer program including program instructions, and the processor is used to execute the program instructions stored in the computer storage medium. The processor may be a central processing unit (CPU), or may be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, which is suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions to implement the corresponding method flow or corresponding function; the processor described in the embodiment of the present invention can be used to implement the operation of a method for online positioning and identification of pipeline leakage crack morphology.

[0113] In one embodiment of the present invention, a method for online positioning and identification of pipeline leakage crack morphology, if implemented in the form of a software functional unit and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the process in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and when the computer program is executed by the processor, it can implement the steps of the above-mentioned method embodiments. The computer program includes computer program code, which can be in source code form, object code form, executable file or some intermediate form. Computer-readable storage media include permanent and non-permanent, removable and non-removable media, and information storage can be achieved by any method or technology. Information can be computer-readable instructions, data structures, program modules or other data.

[0114] The computer storage medium can be any available medium or data storage device that can be accessed by a computer, including but not limited to magnetic storage (such as floppy disks, hard disks, magnetic tapes, magneto-optical disks (MO)), optical storage (such as CDs, DVDs, BDs, HVDs, etc.), and semiconductor storage (such as ROMs, EPROMs, EEPROMs, non-volatile memories (NANDFLASH), solid-state drives (SSDs)), etc.

[0115] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0116] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products of the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0117] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0118] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 The steps for the function specified in one or more boxes.

[0119] Finally, it should be noted that the above-described embodiments are only specific implementations of the present invention, which are used to illustrate the technical solutions of the present invention, rather than to limit them. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the above-described embodiments, those skilled in the art should understand that any person skilled in the art can modify or easily conceive of changes to the technical solutions described in the above-described embodiments within the technical scope disclosed by the present invention, or replace some of the technical features therein with equivalents. Such modifications, changes, or replacements do not deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A method for online positioning and identification of pipeline leakage crack morphology, characterized in that: include: Acquiring an acoustic emission signal generated when a pipeline leaks, and locating the leak location on the pipeline based on the acoustic emission signal; downsampling the acoustic emission signal to obtain a downsampled acoustic emission signal; Performing pre-emphasis processing on the down-sampled acoustic emission signal, and performing frame-based short-time Fourier transform on the pre-emphasized acoustic emission signal to obtain a spectrum feature of the acoustic emission signal; Design a set number of Mel filter banks according to the maximum frequency of the acoustic emission signal spectrum feature to filter the acoustic emission signal spectrum, and perform a logarithmic operation on the filtered acoustic emission signal spectrum to obtain an FBank feature; Calculate the GI index of the FBank feature and select the first n-dimensional features in the GI index, where n=30~60; The first n-dimensional features are input into a pre-trained pipeline leakage crack morphology recognition model to obtain the pipeline leakage crack morphology.

2. The method for online positioning and identification of pipeline leakage crack morphology according to claim 1 is characterized in that: The pre-emphasis processing is performed on the down-sampled acoustic emission signal, and the calculation formula is: Where, For time, is the acoustic emission signal after pre-emphasis processing, for The acoustic emission signal after downsampling at each moment, for The acoustic emission signal after downsampling at each moment, is the pre-emphasis coefficient.

3. The method for online positioning and identification of pipeline leakage crack morphology according to claim 2 is characterized in that: The acoustic emission signal after pre-emphasis processing is subjected to frame short-time Fourier transform to obtain the spectrum characteristics of the acoustic emission signal, and the calculation formula is: Where, is the window function, is the center of the window function, is an imaginary number, is the spectrum characteristic of the acoustic emission signal.

4. The method for online positioning and identification of pipeline leakage crack morphology according to claim 3 is characterized in that: The Mel filter banks of a set number are designed according to the maximum frequency of the acoustic emission signal spectrum characteristics, and the calculation formula is: Where, For the filters, It corresponds to The center frequency of the filter, For the The value after the Mel scale of the filter is converted to frequency, For the The value after the Mel scale of the filter is converted to frequency, For the The value after the Mel scale of the filter is converted to frequency.

5. The method for online positioning and identification of pipeline leakage crack morphology according to claim 4 is characterized in that: The logarithmic operation is performed on the spectrum of the filtered acoustic emission signal to obtain the FBank feature, and the calculation formula is: Where, is the characteristic length of the spectrum, It is through the The logarithmic energy of the output of the filter is the FBank feature.

6. The method for online positioning and identification of pipeline leakage crack morphology according to claim 5 is characterized in that: The GI index of the FBank feature is calculated using the following formula: Where, For the Logarithmic energy norm, for Rearrange the data from small to large, yes No. Features, yes No. Features, is the GI index.

7. The method for online positioning and identification of pipeline leakage crack morphology according to claim 1 is characterized in that: The training method of the pipeline leakage crack morphology recognition model is as follows: Acquire the acoustic emission signals generated when pipelines with different leakage crack morphologies leak; Downsampling the acoustic emission signals corresponding to different leakage crack morphologies is performed respectively to obtain downsampled acoustic emission signals corresponding to different leakage crack morphologies; The downsampled acoustic emission signals corresponding to different leakage crack morphologies are pre-emphasized, and the pre-emphasized acoustic emission signals are subjected to frame-by-frame short-time Fourier transform to obtain the spectrum characteristics of the acoustic emission signals corresponding to different leakage crack morphologies. According to the maximum frequency of the acoustic emission signal spectrum characteristics corresponding to different leakage crack morphologies, a set number of Mel filter banks are designed to filter the respective acoustic emission signal spectra. Logarithmic operation is performed on the filtered acoustic emission signal spectra to obtain the FBank characteristics corresponding to different leakage crack morphologies. The GI index of the FBank features corresponding to different leakage crack morphologies is calculated respectively, and the first n-dimensional features in each GI index are selected, where n = 30~60; The first n-dimensional features of each GI index and the corresponding leakage crack morphology are used as training samples, and a one-dimensional convolutional neural network based on DenseNet is trained to obtain the pipeline leakage crack morphology recognition model.

8. The method for online positioning and identification of pipeline leakage crack morphology according to claim 1 is characterized in that: The step of locating the leakage position on the pipeline according to the acoustic emission signal comprises: The time difference method is used to locate the leak position on the pipeline based on the acoustic emission signal. The calculation formula is as follows: Where, is the distance from the leak point to the sensor, is the sensor spacing, is the time difference between the acoustic emission wave and the two sensors, is the acoustic emission wave velocity, which is calculated through the commonly used lead breaking experiment.

9. An online positioning and identification device for pipeline leakage crack morphology, characterized in that: include: An acquisition module is used to acquire an acoustic emission signal generated when a pipeline leaks, and locate the leak position on the pipeline according to the acoustic emission signal; A downsampling module is used to downsample the acoustic emission signal to obtain a downsampled acoustic emission signal; A spectrum feature module is used to perform pre-emphasis processing on the downsampled acoustic emission signal and perform frame short-time Fourier transform on the pre-emphasized acoustic emission signal to obtain a spectrum feature of the acoustic emission signal; An FBank feature module is used to design a set number of Mel filter banks according to the maximum frequency of the acoustic emission signal spectrum feature to filter the acoustic emission signal spectrum, and perform logarithmic operation on the filtered acoustic emission signal spectrum to obtain FBank features; GI index module, used to calculate the GI index of the FBank feature and select the first n-dimensional features in the GI index, where n=30~60; The recognition module is used to input the first n-dimensional features into a pre-trained pipeline leakage crack morphology recognition model to obtain the pipeline leakage crack morphology.

10. A device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the method for online positioning and identifying pipeline leakage crack morphology according to any one of claims 1 to 8 are implemented.

Images