Pipeline weld defect analysis method and device based on magnetic flux leakage detection

Abstract

The invention relates to the technical field of pipeline detection, and discloses a pipeline weld defect analysis method and device based on magnetic flux leakage detection.The method comprises the steps that a composite excitation signal is generated and optimized through a multi-frequency excitation source, a magnetic flux leakage attenuation coefficient and scattering characteristics are calculated in combination with pipeline parameters, and if the attenuation coefficient exceeds a preset attenuation threshold value, the pipeline weld defect is detected; if yes, adjusting the excitation signal for optimization, extracting high-frequency transient and low-frequency stable features of the optimized magnetic flux leakage signal, and generating a preliminary feature vector of the weld defect; mapping the preliminary feature vector to a three-dimensional defect space position, generating a space distribution diagram containing the defect position, optimizing the preliminary feature vector to make a fuzzy region clear, and obtaining the depth and direction of the defect; and according to the depth and direction of the defect, analyzing the spatial distribution characteristics of the preliminary feature vectors, determining the type of the defect, dynamically adjusting excitation signal parameters for optimization, recalculating the attenuation coefficient of the signal, extracting the optimized scattering features, and obtaining a high-precision weld defect positioning result.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of pipeline detection, and in particular to a pipeline weld defect analysis method and device based on magnetic flux leakage detection. BACKGROUND

[0002] Pipeline transportation is the core infrastructure in the field of energy and chemical industry, and its safety and reliability are directly related to the national economic lifeline and public safety. Among them, the quality of pipeline weld is particularly critical, and weld defects may cause leakage, rupture and even catastrophic accidents. Magnetic flux leakage detection technology has become the mainstream method for pipeline weld detection due to its non-destructive characteristics and high sensitivity to internal defects of metal.

[0003] In an existing magnetic flux leakage detection method, a strong magnetic field is applied to the surface of the pipe wall by a direct current / low frequency alternating current excitation. A magnetic sensor array scans along the weld, capturing the surface magnetic flux field strength and gradient. Finally, the signal is processed for feature extraction, classification and evaluation, and the signal pattern is matched against the defect database.

[0004] The traditional method relies on a single excitation method, which is difficult to adapt to complex pipeline structures and variable working environments, resulting in insufficient detection accuracy. Especially when facing deep or small defects, the signal capture and analysis capability is significantly reduced, making it difficult to accurately distinguish the type and spatial location of the defect. Therefore, the existing technology leads to insufficient precision in pipeline weld defect detection, which may cause misjudgment or missed detection. SUMMARY

[0005] The present application provides a pipeline weld defect analysis method and device based on magnetic flux leakage detection to solve the problem of insufficient precision in pipeline weld defect detection caused by the existing technology.

[0006] In the first aspect, to solve the above technical problems, the present application provides a pipeline weld defect analysis method based on magnetic flux leakage detection, comprising: generating a first composite excitation signal by a multi-frequency excitation source and adjusting and optimizing it to obtain an initial magnetic flux leakage signal set; According to the initial magnetic flux leakage signal set, combined with the calibrated pipeline material parameters, the attenuation coefficient of the magnetic flux leakage signal is calculated to determine the scattering characteristics of the signal; If the attenuation coefficient of the magnetic flux leakage signal exceeds the preset attenuation threshold, adjust the excitation signal and generate an optimized magnetic flux leakage signal set; Extract the high-frequency transient features and low-frequency stable features in the signal time sequence and spatial distribution characteristics in the optimized magnetic flux leakage signal set to generate a preliminary feature vector of the weld defect; Map the preliminary feature vector to a three-dimensional defect space location through a pre-established signal propagation path model to generate a spatial distribution map containing the defect location; If there is a fuzzy area in the spatial distribution map, the scattering feature and the attenuation coefficient are combined to optimize the preliminary feature vector to determine the depth and direction of the defect; According to the depth and direction of the defect, the spatial distribution characteristics of the preliminary feature vector are analyzed to determine the defect type, generate a detection report containing the defect type and spatial position, and dynamically adjust the excitation signal parameters to generate a second composite excitation signal; According to the second composite excitation signal, the attenuation coefficient of the signal is recalculated, the optimized scattering feature is extracted, and a high-precision weld defect positioning result is obtained.

[0007] In a second aspect, the present application provides a pipeline weld defect analysis device based on magnetic flux leakage detection, comprising: An initial magnetic flux leakage signal generation module generates a first composite excitation signal through a multi-frequency excitation source and performs adjustment and optimization to obtain an initial magnetic flux leakage signal set; A magnetic flux leakage signal feature determination module calculates the attenuation coefficient of the magnetic flux leakage signal and determines the scattering feature of the signal according to the initial magnetic flux leakage signal set in combination with the calibrated pipeline material parameters; A magnetic flux leakage signal attenuation optimization module adjusts the excitation signal and generates an optimized magnetic flux leakage signal set if the attenuation coefficient of the signal exceeds a preset attenuation threshold; A weld defect feature extraction module extracts high-frequency transient features and low-frequency stable features in the signal time sequence and spatial distribution characteristics in the optimized magnetic flux leakage signal set to generate a preliminary feature vector of the weld defect; A defect position spatial distribution module maps the preliminary feature vector to a three-dimensional defect spatial position through a pre-established signal propagation path model to generate a spatial distribution map containing the defect position; A defect detection shape determination module optimizes the preliminary feature vector to determine the depth and direction of the defect if there is a fuzzy area in the spatial distribution map, in combination with the scattering feature and the attenuation coefficient; A second composite excitation signal module analyzes the spatial distribution characteristics of the preliminary feature vector to determine the defect type according to the depth and direction of the defect, generates a detection report containing the defect type and spatial position, and dynamically adjusts the excitation signal parameters to generate a second composite excitation signal; A weld defect report generation module recalculates the attenuation coefficient of the signal, extracts the optimized scattering feature, and obtains a high-precision weld defect positioning result according to the second composite excitation signal.

[0008] In a third aspect, the present application provides an electronic device comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor implements the pipeline weld defect analysis method according to any one of the preceding method embodiments when executing the computer program.

[0009] In a fourth aspect, the present application provides a computer readable storage medium comprising a stored computer program, wherein the computer readable storage medium controls a device in which the computer readable storage medium is located to execute the pipeline weld defect analysis method according to any one of the preceding method embodiments when the computer program is executed.

[0010] Compared with the prior art, the present application has the following beneficial effects: (1) The present application successfully separates high-frequency transient features and low-frequency stable features by performing multi-layer convolution processing on the time series data and spatial distribution data collected by magnetic flux leakage detection, effectively improving the resolution and analysis depth of the defect signal. The design of multi-core convolution kernel enables the detection system to simultaneously capture the dynamic changes and overall morphological features of the weld defects, significantly enhancing the accuracy and robustness of the detection, and meeting the application requirements under complex working conditions; (2) The present application organically combines high-frequency and low-frequency features to generate a unified feature vector, and effectively reduces the dimensionality through principal component analysis or autoencoder data reconstruction algorithm, which not only greatly reduces the consumption of computing resources, but also ensures the complete preservation of key defect information; (3) The multi-layer data processing and mapping mechanism constructed by the present application not only realizes the structured management and efficient retrieval of weld defect features, but also provides a solid foundation for subsequent defect classification, positioning and evaluation. This method effectively distinguishes different types of defects by integrating time and spatial feature information, enhances the recognition ability of the system for cracks, corrosion and other defects, and at the same time, intuitively displays the defect situation through a three-dimensional distribution map, greatly improving the precision and reliability of pipeline weld safety detection. BRIEF DESCRIPTION OF DRAWINGS

[0011] Figure 1 is a pipeline weld defect analysis method flowchart based on magnetic flux leakage detection provided by the first embodiment of the present application; Figure 2 is a pipeline weld defect analysis device structure diagram based on magnetic flux leakage detection provided by the second embodiment of the present application. DETAILED DESCRIPTION

[0012] With reference to the accompanying drawings, the technical solutions in the embodiments of the present application will be clearly and completely described below, obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of the present application.

[0013] In the field of energy and chemical industry, the safety and reliability of pipeline transportation are directly related to the national economic lifeline and public safety. Among them, the quality of pipeline weld is particularly critical, and weld defects may cause leakage, rupture and even catastrophic accidents.

[0014] The traditional method relies on a single excitation mode, which is difficult to adapt to complex pipeline structures and variable working environments, resulting in insufficient detection accuracy. Especially when facing deep or small defects, the signal capture and analysis capability is significantly reduced, and it is difficult to accurately distinguish the type and spatial position of the defect. Therefore, the existing technology leads to insufficient accuracy of pipeline weld defect detection, which is prone to misjudgment or missed detection.

[0015] In order to solve the above problems, the following specific embodiments will be used to introduce and explain a pipeline weld defect analysis method based on magnetic flux leakage detection provided by the embodiments of the present application.

[0016] Reference Figure 1 The first embodiment of the present application provides a pipeline weld defect analysis method based on magnetic flux leakage detection, comprising the following steps: S101, generating a first composite excitation signal by a multi-frequency excitation source, and adjusting and optimizing to obtain an initial magnetic flux leakage signal set; S102, calculating the attenuation coefficient of the magnetic flux leakage signal according to the initial magnetic flux leakage signal set, and determining the scattering characteristics of the signal in combination with the calibrated pipeline material parameters; S103, if the attenuation coefficient of the magnetic flux leakage signal exceeds a preset attenuation threshold, adjusting the excitation signal and generating an optimized magnetic flux leakage signal set; S104, extracting high-frequency transient features and low-frequency stable features in the signal time sequence and spatial distribution characteristics in the optimized magnetic flux leakage signal set to generate a preliminary feature vector of the weld defect; S105, mapping the preliminary feature vector to a three-dimensional defect space position through a pre-established signal propagation path model to generate a spatial distribution map containing the defect position; S106, if there is a fuzzy area in the spatial distribution map, combining the scattering characteristics and the attenuation coefficient to optimize the preliminary feature vector and determine the depth and direction of the defect; S107, according to the depth and direction of the defect, analyze the spatial distribution characteristics of the preliminary feature vector, determine the defect type, generate a detection report containing the defect type and spatial position, and dynamically adjust the excitation signal parameters to generate a second composite excitation signal; S108, according to the second composite excitation signal, recompute the signal attenuation coefficient, extract the optimized scattering characteristics, and obtain high-precision weld defect positioning results.

[0017] In step S101, a first composite excitation signal is generated by a multi-frequency excitation source, and is adjusted and optimized to obtain an initial magnetic flux leakage signal set, including: S1011, generate high-frequency signals and low-frequency signals through a preset frequency range, and synthesize a composite excitation signal containing high-frequency and low-frequency signals; S1012, according to the composite excitation signal, combine the pipe wall diameter and wall thickness parameters to calculate and adjust the amplitude adjustment coefficient, and obtain the amplitude-adjusted excitation signal data; S1013, calculate the signal propagation direction according to the amplitude-adjusted excitation signal data, and perform emission angle optimization to obtain optimized excitation signal data; S1014, obtain the magnetic flux leakage signals of different depth regions of the pipeline weld from the optimized excitation signal data, convert the collected analog signals into digital signals, and obtain the initial magnetic flux leakage signal set data.

[0018] In step S1011, high-frequency signals and low-frequency signals are generated through a preset frequency range, and a composite excitation signal containing high-frequency and low-frequency signals is synthesized; It should be noted that Fourier transform can represent signals in the time domain as superposition of different frequencies, reveal the frequency structure and amplitude, phase information of each frequency component, obtain a frequency domain feature vector, and describe the response characteristics of the signal at different frequencies, providing basic data for defect recognition and classification.

[0019] In an implementation manner, a composite signal can be designed through a preset frequency range. Assuming that the pipeline detection needs to cover a frequency range of 10Hz to 100kHz, first, a multi-frequency excitation source generates a signal covering the frequency range of 10Hz to 100kHz, then uses fast Fourier transform (FFT) to extract low-frequency signals (10Hz to 1kHz) and high-frequency signals (10kHz to 100kHz) from the generated signal, then weights and superimposes the two frequency components according to a certain amplitude ratio, and finally synthesizes a composite excitation signal covering a wide frequency band, wherein the low-frequency component is mainly used for deep defect detection to ensure penetration, and the high-frequency component is mainly used for surface crack detection to improve resolution, thereby realizing effective generation and synthesis of signals.

[0020] In step S1012, according to the composite excitation signal, the pipe wall diameter and the wall thickness parameters are combined to calculate the amplitude adjustment coefficient and adjust to obtain the amplitude-adjusted excitation signal data; It should be noted that the amplitude adjustment coefficient is calculated in combination with the pipe diameter and wall thickness parameters. The specific content is: based on the diameter and wall thickness of the pipe, the preset threshold value , is compared, and the amplitude adjustment coefficient is calculated. The adjustment coefficient can be calculated by the formula: , wherein , are the adjusted amplitude ratios of high-frequency and low-frequency signals, respectively, , are parameters representing high-frequency and low-frequency modes in the excitation signal, such as amplitude, frequency or corresponding filter characteristic parameters, which are used to distinguish and process signals of different frequency components, , is a gain parameter, and only when the parameter exceeds the threshold value, the corresponding amplitude is increased. The amplitude adjustment is performed separately for different frequency bands of the composite signal, that is, the amplitude ratio of the high-frequency signal is increased according to the pipe diameter, and the amplitude ratio of the low-frequency signal is increased according to the wall thickness, which meets the technical requirements of low-frequency detection of deep defects and high-frequency detection of surface defects. The adjustment operation adopts a multiplication factor form. After the signal is separated into different frequency bands by Fourier transform in the frequency domain, the amplitude of each frequency band is multiplied by the corresponding adjustment coefficient, and then inverse Fourier transform is performed to synthesize the time-domain composite signal, thereby ensuring the integrity and phase relationship of the signal. Coordination to avoid waveform distortion. The signal sampling rate and parameter calibration need to meet the system design requirements to ensure the effectiveness and stability of the adjusted signal in detection.

[0021] In one implementation, for amplitude adjustment, the pipe diameter is assumed to be 500 mm, the wall thickness is 10 mm, the preset diameter threshold is 300 mm, and the wall thickness threshold is 8 mm. Since the diameter is greater than the threshold, the high-frequency signal amplitude ratio is increased to 60% to enhance the surface defect detection capability; the wall thickness is greater than the threshold, and the low-frequency signal amplitude ratio is increased to 40% to improve the deep defect response. The adjusted signal is more suitable for the physical characteristics of the pipe, reducing the risk of false detection and missed detection.

[0022] In step S1013, the signal propagation direction is calculated according to the amplitude-adjusted excitation signal data, and the emission angle is optimized to obtain the optimized excitation signal data. It should be noted that the transmission angle of the excitation source is adjusted according to the pipe weld depth, and the propagation efficiency of the high-frequency signal and the low-frequency signal in the excitation signal is changed, which is specifically manifested as follows: if the pipe weld depth is greater than the preset threshold, the transmission angle of the excitation source is adjusted to enhance the propagation efficiency of the low-frequency signal; if the pipe weld depth is less than the preset threshold, the transmission angle is adjusted to enhance the propagation efficiency of the high-frequency signal, and the optimized excitation signal data is obtained. The signal propagation direction is derived based on the amplitude and phase space distribution data of the excitation signal, combined with finite element simulation by analyzing the phase gradient and amplitude distribution, calculated in the frequency domain, and then the time domain propagation path is obtained by inverse transformation, to realize accurate direction positioning and optimization.

[0023] In an implementation manner, the transmission angle is optimized according to the weld depth. Assuming that the weld depth is 5 mm and the preset threshold is 3 mm. If the depth is greater than the threshold, the transmission angle of the excitation source is adjusted to 30 degrees to enhance the propagation efficiency of the low-frequency signal along the pipe axial direction, which is beneficial to the detection of deep weld defects. If the depth is less than the threshold, it is adjusted to 45 degrees to enhance the radial propagation of the high-frequency signal, which is beneficial to the identification of surface defects.

[0024] In step S1014, the magnetic flux leakage signals of different depth regions of the pipe weld are obtained from the optimized excitation signal data, the collected analog signals are converted into digital signals, and the initial magnetic flux leakage signal set data is obtained; It should be noted that the specific steps of converting the collected analog signals into digital signals are as follows: using a magnetic flux leakage signal collection device to obtain analog magnetic flux leakage signals located at the pipe weld which can reflect defects, then periodically measuring the analog signals at a certain sampling frequency to ensure that the high-frequency transient characteristics and low-frequency stable parts of the signals can be captured, and finally mapping the analog amplitude values obtained by sampling to a limited number of digital levels to form discrete digital values, thereby obtaining digital magnetic flux leakage signals.

[0025] In an implementation manner, a high-sensitivity magnetic sensor is used to collect the magnetic flux leakage signals, and the magnetic flux leakage data excited by the excitation signal is optimized. Assuming that the sampling rate of the collection device is 1 MHz, the signal is converted into a digital signal through an analog-to-digital converter, and the magnetic flux leakage signal intensity of a certain weld region is 0.1 mT. After being converted into a digital signal, the defect position and depth can be accurately analyzed, for example, according to the collected signal set, a 1 mm deep surface crack and a 5 mm deep internal defect can be clearly distinguished, and the defect identification sensitivity and reliability are improved.

[0026] In step S102, according to the initial magnetic flux leakage signal set, the attenuation coefficient of the magnetic flux leakage signal is calculated in combination with the calibrated pipe material parameters, and the scattering characteristics of the signal are determined, including: S1021, according to the initial magnetic flux leakage signal set, in combination with the pre-input pipe diameter, wall thickness and weld width, a high-precision three-dimensional finite element grid is generated, and three-dimensional grid data is obtained; S1022, according to the three-dimensional grid data, and the preset pipe material magnetic permeability and conductivity parameters, simulating the electromagnetic interaction boundary of the excitation source and the pipe surface, calculating the electromagnetic field distribution, obtaining the initial data of the magnetic flux leakage signal; S1023, according to the initial data of the magnetic flux leakage signal, calculating the propagation path of the signal in the complex pipe structure, combining the path geometric characteristics, determining the attenuation coefficient and scattering characteristics.

[0027] In step S1021, according to the initial magnetic flux leakage signal set, combining the pre-input pipe diameter, wall thickness and weld width, a high-precision three-dimensional finite element grid is generated, and three-dimensional grid data is obtained; It should be noted that the finite element grid is to divide the geometric region of the calculation model into many small and simple units (such as triangles, tetrahedrons, hexahedrons, etc.), which are called "element cells", and the nodes are the vertices of the elements. Through this division, a complex physical problem can be converted into a numerical calculation problem on discrete elements.

[0028] In an implementation manner, based on the initial magnetic flux leakage signal set, a finite element simulation method is adopted. Taking ANSYS software as an example, a high-precision three-dimensional finite element grid is generated, taking the pipe diameter 0.5m, the wall thickness 0.01m and the weld width 0.005m as an example, setting the grid unit size to 0.002m, ensuring that the weld area is refined to 0.001m, generating about 5 million tetrahedral elements, and ensuring the geometric accuracy.

[0029] In step S1022, according to the three-dimensional grid data, and the preset pipe material magnetic permeability and conductivity parameters, simulating the electromagnetic interaction boundary of the excitation source and the pipe surface, calculating the electromagnetic field distribution, obtaining the initial data of the magnetic flux leakage signal; It should be noted that the main function of the Maxwell equation set is to describe the electromagnetic properties of the pipe material, such as magnetic permeability and conductivity, to establish a mathematical model of the actual electromagnetic field environment, and to accurately simulate the electromagnetic field distribution and signal propagation process in the pipe weld area.

[0030] In an implementation manner, taking ANSYS software as an example, if the three-dimensional grid data is generated, according to the relative magnetic permeability μ r = 200 and the conductivity σ = 5×10 6 S / m parameters of the pipe material, the material electromagnetic properties are given to the corresponding grid area in the software, and the excitation source such as frequency 50Hz, alternating current density 1×10 6A / m2, and Dirichlet boundary conditions, the Maxwell equations are solved by the finite element method to simulate the electromagnetic interaction boundary between the excitation source and the pipe surface. Then the numerical solution of Maxwell equations is obtained, and the spatial distribution of electromagnetic field and the initial data of magnetic flux leakage signal are obtained by combining the frequency domain analysis, which can realize the high-precision detection and evaluation of the pipe weld defects.

[0031] In step S1023, according to the initial data of the magnetic flux leakage signal, the propagation path of the signal in the complex pipe structure is calculated, and the attenuation coefficient and the scattering characteristic are determined according to the path geometry characteristic; It should be noted that the scattering characteristic is specifically manifested as the amplitude and direction change of the magnetic flux leakage signal in the pipe defect area, which reflects the scattering mode of the signal at the defect. These characteristics are obtained by finite element simulation, and are usually manifested as scattering intensity distribution at different grid points, and contain information such as signal intensity and angle deviation in high-dimensional feature space. After dimension reduction by principal component analysis, the significant scattering characteristics caused by defects can be highlighted to accurately locate the spatial position and shape of the defects.

[0032] In an implementation manner, the propagation path is calculated based on the finite element simulation and the material geometry information, and a three-dimensional path with spatial refraction, reflection and scattering characteristics is generated. The geometry characteristics include path length, refraction bending angle and scattering angle. Taking ANSYS software as an example, according to the electromagnetic field distribution of the pipe weld area and the initial data of the magnetic flux leakage signal, the attenuation process of the magnetic flux density is analyzed combined with the geometry characteristic of the path, and finally the attenuation coefficient (such as 0.15 / m) of the signal and the scattering characteristic extracted by the magnetic field gradient and Fourier transform are determined, and the comprehensive quantitative analysis of the detection signal is completed.

[0033] In step S103, if the attenuation coefficient of the magnetic flux leakage signal exceeds the preset attenuation threshold, the excitation signal is adjusted and an optimized magnetic flux leakage signal set is generated, including: S1031, if the attenuation coefficient of the magnetic flux leakage signal exceeds the preset attenuation threshold, the amplitude and emission angle of the signal are adjusted, an adjusted parameter set is generated, the amplitude change and angle adjustment are recorded, and a first optimized parameter set is obtained; S1032, according to the first optimized parameter set, a magnetic flux leakage signal set is generated, high-frequency sampling is performed to capture transient characteristics, and a first magnetic flux leakage signal set is obtained; S1033, time series data is extracted from the first magnetic flux leakage signal set, transient characteristics and spatial distribution characteristics are separated, and a first feature data set is obtained; S1034, the first feature data set is optimized to generate optimized time series and spatial distribution characteristics, and the optimized magnetic flux leakage signal set is obtained.

[0034] In step S1031, if the attenuation coefficient of the magnetic flux leakage signal exceeds the preset attenuation threshold, the amplitude and emission angle of the signal are adjusted, an adjusted parameter set is generated, the amplitude change and angle adjustment are recorded, and a first optimized parameter set is obtained. In an implementation manner, if the attenuation coefficient of the magnetic flux leakage signal exceeds the preset threshold (for example, 0.8), the amplitude (for example, increased from 100 mT to 120 mT) and the emission angle (for example, adjusted from 45° to 50°) of the excitation signal are adjusted, a parameter set containing the amplitude change of 20 mT and the angle adjustment of 5° is generated, and then a first optimized parameter set is formed. The process is iteratively updated and recorded on the signal parameters, to ensure the accuracy and sensitivity of subsequent magnetic flux leakage signal acquisition.

[0035] In step S1032, a magnetic flux leakage signal set is generated according to the first optimized parameter set, high-frequency sampling is performed to capture transient characteristics, and a first magnetic flux leakage signal set is obtained. In an implementation manner, according to the first optimized parameter set, the signal generation module generates a magnetic flux leakage signal set through an electromagnetic exciter, and adopts a high-frequency sampling mode of 10 kHz to capture transient magnetic field changes (for example, signal peaks and mutations) caused by pipeline defects, and then obtains a first magnetic flux leakage signal set containing time sequence data and spatial distribution characteristics, to lay a foundation for subsequent defect feature extraction and analysis.

[0036] In step S1033, time sequence data is extracted from the first magnetic flux leakage signal set, transient characteristics and spatial distribution characteristics are separated, and a first feature data set is obtained. In an implementation manner, the first magnetic flux leakage signal set containing time sequence data is obtained through high-frequency real-time sampling, which can reflect the defect position and characteristics, and then feature extraction (for example, wavelet transform) is performed on the signal set, to separate transient characteristics (for example, rapid changes of defect edges) and spatial distribution characteristics (for example, defect depth and width).

[0037] In step S1034, the first feature data set is optimized, optimized time sequence and spatial distribution characteristics are generated, and an optimized magnetic flux leakage signal set is obtained. In an implementation manner, based on the first feature data set, dimension reduction and noise filtering processing are performed, the smoothness of the time sequence and the clarity of the transient peak value are optimized, and the accuracy and stability of the spatial distribution characteristics are enhanced, to generate an optimized magnetic flux leakage signal set containing more optimal time sequence and spatial distribution characteristics, and to provide high-quality input for subsequent weld defect identification and evaluation.

[0038] In step S104, high-frequency transient characteristics and low-frequency stable characteristics in the signal time sequence and the spatial distribution characteristics in the optimized magnetic flux leakage signal set are extracted, a preliminary feature vector of the weld defect is generated, including: S1041, from the optimized magnetic flux leakage signal set, obtain time series data and spatial distribution data, perform convolution operation on the time series data, extract high-frequency transient features, and obtain a first high-frequency feature set; S1042, according to the first high-frequency feature set, separate the low-frequency stable features of the spatial distribution data to obtain a first low-frequency feature set; S1043, if the feature dimensions of the first high-frequency feature set and the first low-frequency feature set satisfy a preset first dimension threshold, then the first high-frequency feature set and the first low-frequency feature set are merged to generate a first feature vector; S1044, for the first feature vector, dimension reduction processing is performed to generate a preliminary feature vector for the weld defect.

[0039] In step S1041, from the optimized magnetic flux leakage signal set, obtain time series data and spatial distribution data, perform convolution operation on the time series data, extract high-frequency transient features, and obtain a first high-frequency feature set; In an implementation manner, by uniformly arranging a sensor array on the weld surface with high spatial resolution (such as 2 mm) to collect time-space distribution data, and combining time series signals obtained with high sampling frequency (such as 100 kHz), an original data matrix containing signal changes over time and space is formed. When performing convolution operation on the time series data, first, a convolution kernel (for example, 3x3) of appropriate size is designed to capture high-frequency transient features; then the convolution kernel is slid along the time series data, and the convolution operation is calculated for each sliding position to extract local feature responses; through a multi-layer convolution network structure, different scale feature channels (such as 128 channels) are gradually extracted, and each channel corresponds to a transient mode; finally, after processing, a first high-frequency feature set reflecting the high-frequency changes of the signal is generated to represent the dynamic characteristics of the weld defect.

[0040] In step S1042, according to the first high-frequency feature set, separate the low-frequency stable features of the spatial distribution data to obtain a first low-frequency feature set; In an implementation manner, when detecting the weld of the same pipeline, the low-frequency features may correspond to the thickness changes of the weld surface or the uniform corrosion area of a large range of materials. First, Fourier transform is performed on the collected magnetic field spatial distribution data to convert it from the time domain or the spatial domain to the frequency domain; by setting a low-frequency threshold, the frequency components below the threshold in the frequency domain are screened out, and the features corresponding to these low-frequency components are the low-frequency features; then inverse Fourier transform is performed on the screened low-frequency components to convert them back to the spatial domain, thereby obtaining the first low-frequency feature set. The first low-frequency feature set obtained after processing may contain 64 feature channels, describing the stable distribution mode of the magnetic field in space.

[0041] In step S1043, if the feature dimensions of the first high-frequency feature set and the first low-frequency feature set satisfy a preset first dimension threshold, the first high-frequency feature set and the first low-frequency feature set are merged to generate a first feature vector; It should be noted that the specific steps of merging the first high-frequency feature set and the first low-frequency feature set include: first, judging whether the feature dimensions of the two satisfy a preset threshold (such as the total dimension after merging does not exceed 256); if yes, the first high-frequency feature set (such as 128 dimensions) and the first low-frequency feature set (such as 64 dimensions) are spliced in the feature dimension to form a unified first feature vector (for example, 192 dimensions); the fusion vector integrates the high-frequency features of the transient state and the low-frequency features of the stable state, and can more comprehensively describe the characteristics of the weld defect, facilitating subsequent dimension reduction processing and defect recognition.

[0042] In step S1044, dimension reduction processing is performed on the first feature vector to generate a preliminary feature vector for the weld defect. It should be noted that the specific steps of dimension reduction processing include: first, calculating the covariance matrix of the first feature data set to analyze the correlation between the dimensions of the features; then, by solving the eigenvalues and eigenvectors of the covariance matrix, the first few principal components that can retain most of the data variance (such as more than 90%) are selected; then, a low-dimensional subspace is constructed using these principal components, and the high-dimensional features are projected into the subspace to realize dimension compression, and finally, the second feature vector set after dimension reduction is obtained.

[0043] In an implementation manner, the 192-dimensional feature vector is reduced to a 32-dimensional preliminary feature vector, which retains the main defect information such as crack depth or corrosion area. Such dimension reduction operation can reduce the amount of calculation while retaining the key features, facilitating subsequent defect classification or positioning.

[0044] In step S105, the preliminary feature vector is mapped to a three-dimensional defect space position through a pre-established signal propagation path model to generate a spatial distribution map containing the defect position.

[0045] It should be noted that the training process of the pre-established signal propagation path model includes: S1051, initializing the initial propagation path data of the parameters of the signal propagation path model; S1052, performing finite element simulation and signal propagation simulation according to the initial propagation path data and the pre-stored geometric properties and material properties of the pipeline to obtain a preliminary propagation trajectory and a preliminary scattering feature; S1053, constructing a first scattering feature set based on the preliminary propagation trajectory and the preliminary scattering feature; S1054, dimension reduction processing is performed on the first scattering feature set to generate a second feature vector; S1055, according to the second feature vector, the defect space position is updated and the signal propagation path is adjusted; S1056, when the number of training times is greater than or equal to the preset maximum training times, it is determined that the training is completed, and an optimized signal propagation path model is obtained.

[0046] In step S1051, the parameters of the signal propagation path model and the initial propagation path data are initialized; It should be noted that the initialization of the parameters of the signal propagation path model mainly includes the geometric parameters of the pipe and the weld (such as pipe diameter, wall thickness, weld width and grid cell size), material parameters (such as relative permeability, electrical conductivity, sound speed and density), excitation source parameters (such as excitation current density, frequency, excitation position and boundary conditions), and attenuation coefficients and scattering features in signal propagation, which together constitute a finite element simulation environment to ensure accurate simulation of the propagation path and scattering behavior of the magnetic field signal in the weld and pipe material. The initial propagation path data is based on a set of composite excitation signals, combined with the geometric parameters, material properties and boundary conditions of the pipe and weld, to describe the preliminary data of the signal propagation path in the pipe weld material, i.e. the initial propagation path data.

[0047] In step S1052, according to the initial propagation path data and the pre-stored pipe geometric properties and material properties, finite element simulation and signal propagation simulation are performed to obtain preliminary propagation trajectories and preliminary scattering features; It should be noted that according to the initial propagation path data, combined with the pre-stored pipe geometric properties (such as pipe diameter, wall thickness, weld width and grid division precision) and material properties (such as permeability, electrical conductivity, sound speed and density), a high-precision three-dimensional finite element model is constructed using finite element simulation software, the excitation source parameters and boundary conditions are set, the propagation process of the signal in the weld and pipe material is simulated, the propagation trajectory of the signal and the reflection and scattering phenomena caused by defects in the propagation process are calculated, and then the preliminary propagation trajectory and preliminary scattering feature are extracted.

[0048] In step S1053, a first scattering feature set is constructed based on the preliminary propagation trajectory and the preliminary scattering feature; It should be noted that the steps of constructing the first scattering feature set based on the preliminary propagation trajectory and the preliminary scattering feature are as follows: first, the weld area is divided into a plurality of small grid cells, the signal scattering intensity, direction and propagation path information corresponding to each cell are extracted, and a high-dimensional data set is formed; then these data are organized and coded according to the spatial position and signal characteristics to construct a first scattering feature set containing multi-channel, multi-dimensional information.

[0049] In step S1054, the first scattering feature set is dimensionally reduced to generate a second feature vector; In one solution, the principal component analysis (PCA) is used to reduce the dimension of the first scattering feature set. The specific steps include calculating the covariance matrix of the feature set, extracting the principal components, and selecting several principal components whose cumulative contribution rate reaches a preset threshold (such as 90% or more) to compress the high-dimensional first scattering feature set into a low-dimensional second feature vector. For example, the 256-dimensional first scattering feature set is reduced to a 48-dimensional second feature vector, which not only ensures the retention of the key information of the defect, but also significantly reduces the data dimension, facilitating subsequent defect positioning and classification analysis.

[0050] In step S1055, the second feature vector is used to update the spatial position of the defect and adjust the signal propagation path; It should be noted that the steps of updating the spatial position of the defect and adjusting the signal propagation path according to the second feature vector are as follows: first, the dimensionally reduced feature vector is mapped to the defect space through three-dimensional mapping to determine the specific position coordinate set of the defect; the three-dimensional distribution map containing the defect position is generated by using spatial interpolation based on the propagation trajectory and scattering feature obtained by finite element simulation, so as to update the spatial position of the defect; finally, according to the updated defect distribution information, the transmission parameters and refraction and reflection rules in the path are corrected to improve the propagation path accuracy of the subsequent signal and the accuracy of defect positioning, forming a closed-loop optimization process.

[0051] In step S1056, when the number of training times is greater than or equal to the preset maximum number of training times, it is determined that the training is completed, and an optimized signal propagation path model is obtained; It should be noted that when the number of training times is greater than or equal to the preset maximum number of training times, it is determined that the training process is completed. Based on the parameters updated and optimized during the training, the system uses the final composite excitation signal set to recalculate the signal propagation path and attenuation coefficient through finite element simulation, and extracts the final scattering feature, and then constructs an optimized signal propagation path model.

[0052] In an implementation, the preliminary feature vector is mapped to a defect space by the trained signal propagation path model to determine a set of three-dimensional coordinates of the defect, which can be expressed as: in detection, a 48-dimensional feature vector can be mapped to a pipe surface to generate a set of coordinate sets with a spatial resolution of 1 mm to mark the specific location of the crack or corrosion, based on the set of three-dimensional coordinates of the defect, the magnetic field intensity data of 50 position points covering the weld area are calculated by interpolation using the Kriging spatial interpolation algorithm, and a three-dimensional defect space distribution map with a size of 200 mm*100 mm*50 mm is generated with a spatial resolution of 1 mm*1 mm, clearly showing the spatial distribution characteristics of the crack extending 5 mm along the pipe axis and the depth varying in the range of 3 mm to 7 mm.

[0053] In step S106, if there is a fuzzy area in the spatial distribution map, the preliminary feature vector is optimized in combination with the scattering feature and the attenuation coefficient to determine the depth and direction of the defect, including: S1061, generating a preliminary feature vector for the fuzzy area in the spatial distribution map; S1062, updating the probability distribution according to the preliminary feature vector in combination with the scattering feature and the attenuation coefficient to obtain an optimized feature vector; S1063, performing three-dimensional reconstruction according to the optimized feature vector and determining the defect depth and defect direction in combination with the spatial positioning technology.

[0054] In step S1061, a preliminary feature vector is generated for the fuzzy area in the spatial distribution map; It should be noted that the existence of the fuzzy area is due to the inclusion of uncertain or interference information in the preliminary feature vector, for example, the rough pipe surface and the complex defect shape make it difficult to accurately extract the scattering feature and the attenuation coefficient of the ultrasonic signal, thereby causing the defect position distribution map to have unclear or fuzzy areas. Therefore, the Bayesian inference algorithm is used to solve the unclear or fuzzy areas. In combination with the scattering feature and the attenuation coefficient of the signal, the amplitude and phase information in the frequency domain is obtained by performing Fourier transform on the collected ultrasonic or magnetic flux leakage detection signal, thereby extracting the scattering feature reflecting the reflection intensity and direction of the signal in the defect area, and calculating the energy loss of the signal in the propagation process to form the attenuation coefficient; then, the scattering feature and the attenuation coefficient are combined to obtain a preliminary feature vector containing the properties of the defect.

[0055] In an implementation, for the fuzzy area in the spatial distribution map, the ultrasonic signal of the area is collected, and the 10MHz time domain signal is converted into frequency domain features by Fourier transform to extract a preliminary feature vector containing 128 dimensions as the basis for subsequent optimization of the defect depth and direction using the Bayesian inference algorithm.

[0056] In step S1062, according to the preliminary feature vector, the probability distribution is updated in combination with the scattering feature and the attenuation coefficient to obtain an optimized feature vector; It should be noted that the optimization process of updating the probability distribution in combination with the scattering feature and the attenuation coefficient is as follows: the Bayesian inference algorithm is used for continuous updating, so as to optimize the feature vector to accurately describe the depth and direction of the defect, and to obtain the probability distribution of the uncertainty of the defect position, depth and direction.

[0057] In an implementation manner, if there is a fuzzy area in the preliminary feature vector, for example, signal interference caused by rough pipe surface or complex defect shape, the Bayesian inference algorithm can be used for optimization. For example, assuming that some dimensions in the preliminary feature vector are fuzzy due to noise interference, the Bayesian inference can calculate the posterior probability by using the known acoustic characteristics of the steel material and the defect type distribution, and generate an optimized feature vector. This optimization process can refine the 128-dimensional vector to 64 dimensions, highlighting the key features related to the defect, and facilitating subsequent positioning.

[0058] In step S1063, according to the optimized feature vector, three-dimensional reconstruction is performed, and the spatial positioning technology is combined to determine the defect depth and the defect direction; It should be noted that the specific steps of the three-dimensional reconstruction include: based on the optimized preliminary feature vector (such as reduced to 64 dimensions by Bayesian inference), the specific position coordinates of the defect in the three-dimensional space are calculated by using the triangular positioning method in combination with the geometric information of the pipe.

[0059] In an implementation manner, based on the optimized feature vector, the spatial positioning technology is combined to perform three-dimensional reconstruction to determine the defect depth and the direction. For example, it is determined that the depth of a certain defect is 5 mm, and the direction is deviated by 15 degrees along the axial direction of the pipe, and the three-dimensional position of the defect is accurately described through spatial coordinate mapping. Further, through data fusion technology, the defect depth, direction and distribution map are integrated to generate a high-precision defect position distribution map. The distribution map usually adopts grid processing (such as dividing the pipe surface into 1 mm x 1 mm areas), and directly displays the specific position and spatial distribution of the crack or corrosion.

[0060] In step S107, according to the depth and direction of the defect, the spatial distribution characteristics of the preliminary feature vector are analyzed, the defect type is determined, a detection report containing the defect type and the spatial position is generated, and the excitation signal parameters are dynamically adjusted to generate a second composite excitation signal, including: S1071, analyzing the detection report to obtain defect position distribution data, processing the defect coordinates and depth information in the defect position distribution data to obtain first defect distribution data; S1072, calculating an excitation signal frequency range according to the first defect distribution data, determining first frequency range data by analyzing material properties and signal attenuation characteristics in the first defect distribution data; S1073, performing signal propagation efficiency simulation on the first frequency range data, adjusting the amplitude of the excitation signal to obtain first amplitude adjustment parameters; S1074, calculating the signal propagation path in the first defect distribution data according to the first amplitude adjustment parameters and the first frequency range data, and performing angle calibration to generate a second composite excitation signal.

[0061] In step S1071, the detection report is analyzed to obtain defect position distribution data, and the defect coordinates and depth information in the defect position distribution data are processed to obtain first defect distribution data; It should be noted that the specific steps of generating the detection report include: combining defect classification results and spatial positioning technology to determine the type, depth and direction information of the defect; integrating multi-dimensional information of the defect to form a high-precision defect position distribution map; generating a heat map of defect distribution to intuitively display the spatial distribution of defects in the pipe weld, facilitating the quick positioning and judgment of problem areas by maintenance personnel; combining pipe operating state data (such as pressure, temperature) to analyze the safety risk of the defect, generating a comprehensive detection report containing defect type, location and impact assessment, providing a scientific basis for subsequent maintenance decision-making.

[0062] In addition, the process of processing the defect coordinates and depth information in the defect position distribution data includes: clustering and analyzing the defect coordinates and depth to determine the concentrated distribution of the defect depth region, generating a high-precision defect three-dimensional distribution map through spatial interpolation and smoothing processing, and optimizing defect positioning and visualization by combining material properties and signal attenuation characteristics, and finally forming the first defect distribution data for subsequent maintenance decision-making.

[0063] In one implementation method, after extracting a plurality of defect coordinates such as (10, 5, 3), (12, 6, 4), (15, 5, 2) and their corresponding depth information from the detection report, a K-means clustering algorithm is used to set the number of clusters to 2, and the defect coordinates and depth data are analyzed to determine the concentrated area of the defect at a depth of 2-3 mm and 4 mm, thereby forming the first defect distribution data to reflect the spatial distribution characteristics of the defect.

[0064] In step S1072, calculating an excitation signal frequency range according to the first defect distribution data, determining first frequency range data by analyzing material properties and signal attenuation characteristics in the first defect distribution data; In an implementation method, according to the first defect distribution data, the frequency range of the excitation signal is calculated by using a frequency response function, according to the first defect distribution data, it is known that the defects in the weld are mainly concentrated in the depth range of 2-4 mm by cluster analysis, combined with the acoustic impedance of the stainless steel material and the signal attenuation characteristics of the ultrasonic wave in the depth range, the signal propagation efficiency at different frequencies is calculated by using the frequency response function, and it is determined that the optimal frequency range of the excitation signal is 1-5 MHz; wherein, high frequency (such as 3-5 MHz) is used for shallow defects to improve resolution, and low frequency (such as 1-3 MHz) is used for deep defects to enhance penetration, so as to realize effective coverage and accurate detection of defects.

[0065] In step S1073, the signal propagation efficiency simulation is performed on the first frequency range data, the amplitude of the excitation signal is adjusted, and the first amplitude adjustment parameter is obtained. In an implementation method, based on the first frequency range data, the signal propagation and attenuation in the material and defects are simulated by using a finite element numerical method through signal propagation efficiency simulation, and the energy loss of the signal at different frequencies and paths is evaluated. It is found that about 20% of the energy is lost due to the reflection of the material boundary during the propagation of the excitation signal, and then the amplitude of the excitation signal is dynamically adjusted from the initial 100 mV to 120 mV, the energy loss is compensated, and the signal detection effect of the defect area is improved, so as to complete the determination of the first amplitude adjustment parameter.

[0066] In step S1074, according to the first amplitude adjustment parameter and the first frequency range data, the signal propagation path in the first defect distribution data is calculated, and angle calibration is performed to generate a second composite excitation signal. In an implementation method, according to the first amplitude adjustment parameter (such as adjusting the amplitude from 100 mV to 120 mV) and the first frequency range data (for example, 3 MHz), the defect coordinates (such as (10, 5, 3), (12, 6, 4) mm) obtained by clustering in the first defect distribution data are applied to the signal propagation path model based on finite element simulation, the propagation trajectory and attenuation characteristics of the signal in the material are numerically calculated, combined with the signal loss judged by the propagation efficiency simulation result, then the incident angle of the ultrasonic probe is adjusted from the initial 30 degrees to 45 degrees to optimize the signal coverage, and finally a second composite excitation signal set with higher detection efficiency and penetration for the defect depth area is generated.

[0067] In step S108, according to the second composite excitation signal, the attenuation coefficient of the signal is recalculated, the optimized scattering characteristics are extracted, and the high-precision weld defect positioning result is obtained, including: S1081, perform finite element simulation according to the second composite excitation signal set, in combination with a preset grid division precision and boundary condition setting, calculate a signal propagation path in the weld material, and obtain first propagation path data; S1082, calculate energy loss in a signal propagation process according to the first propagation path data in combination with material properties, and obtain first attenuation coefficient data; S1083, if the first attenuation coefficient data is lower than a preset attenuation threshold, extract scattering features of the first propagation path data, and obtain first scattering feature data; S1084, calculate accurate coordinates of defects in the weld according to the first scattering feature data in combination with distribution information of a defect depth region, and obtain the high-precision weld defect positioning result.

[0068] In step S1081, perform finite element simulation according to the second composite excitation signal set, in combination with a preset grid division precision and boundary condition setting, calculate a signal propagation path in the weld material, and obtain first propagation path data; In an implementation manner, in a weld defect detection scenario, when finite element simulation is performed based on a composite excitation signal set. Assuming that carbon steel material parameters (sound speed 5900 m / s, density 7850 kg / m³), and in combination with a preset grid division precision (such as 0.5 mm refined grid) and a free boundary condition, numerical simulation is performed by using a finite element software, a propagation path of an excitation signal in a weld is calculated. By solving a wave equation, a signal propagation, reflection and refraction process in a material is simulated, signal intensity and propagation time information is obtained, and finally first propagation path data reflecting a signal path and energy distribution is generated.

[0069] In step S1082, calculate energy loss in a signal propagation process according to the first propagation path data in combination with material properties, and obtain first attenuation coefficient data; In an implementation manner, according to the first propagation path data, in combination with properties such as acoustic impedance, density and elastic modulus of the material, a signal propagation process in the weld material is calculated by using finite element simulation, energy attenuation on different path segments is calculated by analyzing a signal intensity change curve with a propagation distance, for example, simulation shows that the signal intensity is attenuated from an initial 100% to 80% after 5 mm propagation, and accordingly an attenuation coefficient of 0.045 dB / mm is calculated by using a formula, and first attenuation coefficient data is formed.

[0070] In step S1083, if the first attenuation coefficient data is lower than a preset attenuation threshold, extract scattering features of the first propagation path data, and obtain first scattering feature data; In one implementation, if the first attenuation coefficient data is below a preset attenuation threshold (e.g., 0.3 dB / mm), the signal propagation trajectory is extracted from the first propagation path data, and its propagation and scattering characteristics in the material are calculated using finite element simulation. Then, a k-means clustering algorithm is used to extract scattering characteristics from the first propagation path data. Characteristics such as signal intensity and angular deviation in the propagation path can be used as input, and the number of clusters is set to three to distinguish different types of scattering patterns. For example, the clustering results may indicate that defects near the weld surface produce high-frequency scattering, while internal defects cause low-frequency scattering, thereby obtaining first scattering characteristic data reflecting the local characteristics of the defect.

[0071] In step S1084, the precise coordinates of the defect in the weld are calculated based on the first scattering characteristic data and the distribution information of the defect depth area, thereby obtaining the high-precision weld defect location result. In one implementation, defect location based on the first scattering signature data can be performed using geometric positioning, combined with information about the distribution of the defect depth. Assuming the inspection report indicates that the defect depth is concentrated between 1 and 4 mm, scattering signature data analysis can determine that the scattered signal from a particular defect primarily originates from a depth of 2 mm. Combining this with a geometric optical model, the signal's incident and reflection angles are calculated, allowing the precise coordinates of the defect to be derived, such as (11, 5, 2) mm. This ultimately yields the first weld defect location result.

[0072] In summary, the present invention discloses a pipeline weld defect analysis method based on magnetic flux leakage detection, comprising: Generate a first composite excitation signal through a multi-frequency excitation source, and adjust and optimize it to obtain an initial magnetic flux leakage signal set; Calculating the attenuation coefficient of the magnetic flux leakage signal based on the initial magnetic flux leakage signal set and combining it with the calibrated pipeline material parameters to determine the scattering characteristics of the signal; If the attenuation coefficient of the magnetic flux leakage signal exceeds a preset attenuation threshold, adjusting the excitation signal and generating an optimized magnetic flux leakage signal set; Extracting high-frequency transient features and low-frequency stable features from the signal time series and spatial distribution characteristics of the optimized magnetic flux leakage signal set to generate a preliminary feature vector of the weld defect; Mapping the preliminary feature vector to the three-dimensional defect spatial position through a pre-established signal propagation path model to generate a spatial distribution map containing the defect position; If there is a fuzzy area in the spatial distribution map, optimizing the preliminary feature vector by combining the scattering characteristics and the attenuation coefficient to determine the depth and direction of the defect; According to the depth and direction of the defect, the spatial distribution characteristics of the preliminary feature vector are analyzed, the defect type is determined, a detection report containing the defect type and the spatial position is generated, and the excitation signal parameters are dynamically adjusted to generate a second composite excitation signal; According to the second composite excitation signal, the attenuation coefficient of the signal is recalculated, the optimized scattering feature is extracted, and a high-precision weld defect positioning result is obtained.

[0073] Referring to Figure 2 The second embodiment of the present application provides a pipeline weld defect analysis device based on magnetic flux leakage detection, comprising: An initial magnetic flux leakage signal generation module: a first composite excitation signal is generated by a multi-frequency excitation source, and is adjusted and optimized to obtain an initial magnetic flux leakage signal set; A magnetic flux leakage signal feature determination module: according to the initial magnetic flux leakage signal set, the attenuation coefficient of the magnetic flux leakage signal is calculated in combination with the calibrated pipeline material parameters, and the scattering feature of the signal is determined; A magnetic flux leakage signal attenuation optimization module: if the attenuation coefficient of the signal exceeds a preset attenuation threshold, the excitation signal is adjusted and an optimized magnetic flux leakage signal set is generated; A weld defect feature extraction module: high-frequency transient features and low-frequency stable features in the signal time sequence and spatial distribution characteristics in the optimized magnetic flux leakage signal set are extracted to generate a preliminary feature vector of the weld defect; A defect position spatial distribution module: the preliminary feature vector is mapped to a three-dimensional defect spatial position through a pre-established signal propagation path model to generate a spatial distribution map containing the defect position; A defect detection shape determination module: if there is a fuzzy area in the spatial distribution map, the scattering feature and the attenuation coefficient are used to optimize the preliminary feature vector to determine the depth and direction of the defect; A second composite excitation signal module: according to the depth and direction of the defect, the spatial distribution characteristics of the preliminary feature vector are analyzed, the defect type is determined, a detection report containing the defect type and the spatial position is generated, and the excitation signal parameters are dynamically adjusted to generate a second composite excitation signal; A weld defect report generation module: according to the second composite excitation signal, the attenuation coefficient of the signal is recalculated, the optimized scattering feature is extracted, and a high-precision weld defect positioning result is obtained.

[0074] It should be noted that the embodiment of the present application provides a kind of based on deep learning multi-point touch signal processing based on magnetic flux leakage detection pipeline weld defect analysis device for executing all process steps of the above-mentioned embodiment based on magnetic flux leakage detection pipeline weld defect analysis method, the working principle and beneficial effects of the two are one-to-one correspondence, thus no longer tedious.

[0075] An embodiment of the present invention further provides an electronic device. The electronic device includes: a processor, a memory, and a computer program stored in the memory and executable on the processor, such as an initial magnetic flux leakage signal generation program. When the processor executes the computer program, the steps in the above-mentioned pipeline weld defect analysis method embodiments are implemented, such as Figure 1 Alternatively, when the processor executes the computer program, the functions of the modules / units in the above-mentioned device embodiments are realized, for example, Figure 2 The initial magnetic flux leakage signal generation module is shown.

[0076] The electronic device may be a computing device such as a desktop computer, notebook, PDA, or smart tablet. The electronic device may include, but is not limited to, a processor and memory. Those skilled in the art will appreciate that the aforementioned components are merely examples of electronic devices and do not constitute a limitation of the electronic device. The electronic device may include more or fewer components than those described above, or a combination of certain components, or different components. For example, the electronic device may also include input / output devices, network access devices, buses, and the like.

[0077] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor. The processor is the control center of the electronic device and connects various parts of the entire electronic device using various interfaces and lines.

[0078] The memory can be used to store the computer program and / or modules, and the processor realizes various functions of the electronic device by running or executing the computer program and / or modules stored in the memory, and calling data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application program required by a function (such as a sound playing function, an image playing function, etc.), etc.; and the data storage area can store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), etc. In addition, the memory can include a high-speed random access memory, and can also include a nonvolatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one disk storage device, a flash memory device, or other volatile solid-state memory devices.

[0079] It should be noted that the apparatus embodiments described above are merely illustrative, wherein the units described as separate components can or can not be physically separated, and the components displayed as units can or can not be physical units, i.e., can be located in one place, or can be distributed to multiple network units. Part or all of the modules can be selected according to actual needs to achieve the purpose of the present embodiment.

[0080] The above specific embodiments further illustrate the purpose, technical solutions and beneficial effects of the present application. It should be understood that the above description is only a specific embodiment of the present application, and does not limit the protection scope of the present application. It is particularly pointed out that any modification, equivalent replacement, improvement, etc. made by those skilled in the art within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A pipeline weld defect analysis method based on magnetic flux leakage detection, characterized in that: include: Generate a first composite excitation signal through a multi-frequency excitation source, and adjust and optimize it to obtain an initial magnetic flux leakage signal set; Calculating the attenuation coefficient of the magnetic flux leakage signal based on the initial magnetic flux leakage signal set and combining it with the calibrated pipeline material parameters to determine the scattering characteristics of the signal; If the attenuation coefficient of the magnetic flux leakage signal exceeds a preset attenuation threshold, adjusting the excitation signal and generating an optimized magnetic flux leakage signal set; Extracting high-frequency transient features and low-frequency stable features from the signal time series and spatial distribution characteristics of the optimized magnetic flux leakage signal set to generate a preliminary feature vector of the weld defect; Mapping the preliminary feature vector to the three-dimensional defect spatial position through a pre-established signal propagation path model to generate a spatial distribution map containing the defect position; If there is a fuzzy area in the spatial distribution map, optimizing the preliminary feature vector by combining the scattering characteristics and the attenuation coefficient to determine the depth and direction of the defect; Analyzing the spatial distribution characteristics of the preliminary feature vector according to the depth and direction of the defect, determining the defect type, generating a detection report including the defect type and spatial location, and dynamically adjusting the excitation signal parameters to generate a second composite excitation signal; According to the second composite excitation signal, the attenuation coefficient of the signal is recalculated, the optimized scattering characteristics are extracted, and a high-precision weld defect positioning result is obtained.

2. The method according to claim 1, characterized in that The method of generating a first composite excitation signal by a multi-frequency excitation source and performing adjustment and optimization to obtain an initial magnetic flux leakage signal set includes: Generate high-frequency signals and low-frequency signals through a preset frequency range to synthesize a composite excitation signal containing high and low frequencies; According to the composite excitation signal, combined with the pipe wall diameter and wall thickness parameters, the amplitude adjustment coefficient is calculated and adjusted to obtain excitation signal data after amplitude adjustment; Calculating the signal propagation direction according to the amplitude-adjusted excitation signal data, and optimizing the emission angle to obtain optimized excitation signal data; The magnetic flux leakage signals of different depth areas of the pipeline weld are acquired from the optimized excitation signal data, and the acquired analog signals are converted into digital signals to obtain the initial magnetic flux leakage signal set data.

3. The method according to claim 1, characterized in that The step of calculating the attenuation coefficient of the magnetic flux leakage signal based on the initial magnetic flux leakage signal set and combining the calibrated pipeline material parameters to determine the scattering characteristics of the signal includes: Generate a high-precision three-dimensional finite element mesh based on the initial magnetic flux leakage signal set and in combination with the pre-input pipe diameter, wall thickness and weld width to obtain three-dimensional mesh data; Based on the three-dimensional grid data and the preset magnetic permeability and electrical conductivity parameters of the pipeline material, the electromagnetic interaction boundary between the excitation source and the pipeline surface is simulated, the electromagnetic field distribution is calculated, and the initial data of the magnetic flux leakage signal is obtained; Based on the initial data of the magnetic flux leakage signal, the propagation path of the signal in the complex pipeline structure is calculated, and the attenuation coefficient and scattering characteristics are determined in combination with the geometric characteristics of the path.

4. The method according to claim 1, wherein If the attenuation coefficient of the magnetic flux leakage signal exceeds a preset attenuation threshold, adjusting the excitation signal and generating an optimized magnetic flux leakage signal set includes: If the attenuation coefficient of the magnetic flux leakage signal exceeds a preset attenuation threshold, the amplitude and emission angle of the signal are adjusted to generate an adjusted parameter set, and the amplitude change and angle adjustment are recorded to obtain a first optimized parameter set; generating a magnetic flux leakage signal set according to the first optimized parameter set, performing high-frequency sampling to capture transient characteristics, and obtaining a first magnetic flux leakage signal set; Extracting time series data from the first magnetic flux leakage signal set, separating transient features and spatial distribution characteristics, and obtaining a first feature data set; The first feature data set is optimized to generate optimized time series and spatial distribution characteristics, thereby obtaining the optimized magnetic flux leakage signal set.

5. The method according to claim 1, wherein The step of extracting high-frequency transient features and low-frequency stable features from the signal time series and spatial distribution characteristics of the optimized magnetic flux leakage signal set to generate a preliminary feature vector of the weld defect includes: acquiring time series data and spatial distribution data from the optimized magnetic flux leakage signal set, performing a convolution operation on the time series data, extracting high-frequency transient features, and obtaining a first high-frequency feature set; Separating the low-frequency stable features of the spatial distribution data according to the first high-frequency feature set to obtain a first low-frequency feature set; If the feature dimensions of the first high-frequency feature set and the first low-frequency feature set meet a preset first dimension threshold, merging the first high-frequency feature set and the first low-frequency feature set to generate a first feature vector; A dimensionality reduction process is performed on the first eigenvector to generate a preliminary eigenvector for the weld defect.

6. The method according to claim 1, characterized in that The training process of the pre-established signal propagation path model includes: Initialize the parameters of the signal propagation path model and the initial propagation path data; Performing finite element simulation and signal propagation simulation based on the initial propagation path data and pre-stored pipeline geometric properties and material properties to obtain preliminary propagation trajectory and preliminary scattering characteristics; constructing a first scattering feature set based on the preliminary propagation trajectory and the preliminary scattering feature; performing dimensionality reduction processing on the first scattering feature set to generate a second feature vector; updating the defect spatial position and adjusting the signal propagation path according to the second eigenvector; When the number of training times is greater than or equal to the preset maximum number of training times, the training is determined to be completed, and an optimized signal propagation path model is obtained.

7. The method according to claim 1, characterized in that If there is a fuzzy area in the spatial distribution map, optimizing the preliminary feature vector by combining the scattering characteristics and the attenuation coefficient to determine the depth and direction of the defect includes: generating a preliminary feature vector for the fuzzy area in the spatial distribution map; According to the preliminary feature vector, combined with the scattering characteristics and the attenuation coefficient, the probability distribution is updated to obtain an optimized feature vector; Based on the optimized feature vector, three-dimensional reconstruction is performed, and the defect depth and defect direction are determined by combining with spatial positioning technology.

8. The method according to claim 1, characterized in that The dynamically adjusting the excitation signal parameters to generate the second composite excitation signal includes: Analyzing the inspection report to obtain defect location distribution data, and processing the defect coordinates and depth information in the defect location distribution data to obtain first defect distribution data; Calculating an excitation signal frequency range based on the first defect distribution data, and determining first frequency range data by analyzing material properties and signal attenuation characteristics in the first defect distribution data; Performing a signal propagation efficiency simulation on the first frequency range data, adjusting the amplitude of the excitation signal, and obtaining a first amplitude adjustment parameter; According to the first amplitude adjustment parameter and the first frequency range data, a signal propagation path in the first defect distribution data is calculated, and angle calibration is performed to generate a second composite excitation signal.

9. The method according to claim 1, characterized in that The method of recalculating the attenuation coefficient of the signal based on the second composite excitation signal, extracting the optimized scattering characteristics, and obtaining a high-precision weld defect positioning result includes: According to the second composite excitation signal set, combined with the preset meshing accuracy and boundary condition settings, finite element simulation is performed to calculate the propagation path of the signal in the weld material to obtain first propagation path data; Calculate the energy loss during signal propagation based on the first propagation path data and in combination with material properties to obtain first attenuation coefficient data; If the first attenuation coefficient data is lower than a preset attenuation threshold, extracting the scattering characteristics of the first propagation path data to obtain first scattering characteristic data; The precise coordinates of the defect in the weld are calculated based on the first scattering characteristic data and combined with the distribution information of the defect depth area to obtain the high-precision weld defect positioning result.

10. A pipeline weld defect analysis device, characterized in that: include: Initial magnetic flux leakage signal generation module: generates a first composite excitation signal through a multi-frequency excitation source, and performs adjustment and optimization to obtain an initial magnetic flux leakage signal set; Magnetic flux leakage signal feature determination module: based on the initial magnetic flux leakage signal set and combined with the calibrated pipeline material parameters, calculates the attenuation coefficient of the magnetic flux leakage signal and determines the scattering characteristics of the signal; Magnetic flux leakage signal attenuation optimization module: if the attenuation coefficient of the signal exceeds a preset attenuation threshold, adjust the excitation signal and generate an optimized magnetic flux leakage signal set; Weld defect feature extraction module: extracts high-frequency transient features and low-frequency stable features from the signal time series and spatial distribution characteristics of the optimized magnetic flux leakage signal set to generate a preliminary feature vector of the weld defect; Defect location spatial distribution module: maps the preliminary feature vector to the three-dimensional defect spatial position through a pre-established signal propagation path model, and generates a spatial distribution map containing the defect location; Defect detection shape determination module: if there is a fuzzy area in the spatial distribution map, the scattering characteristics and the attenuation coefficient are used to optimize the preliminary feature vector to determine the depth and direction of the defect; Second composite excitation signal module: Analyzes the spatial distribution characteristics of the preliminary feature vector according to the depth and direction of the defect, determines the defect type, generates a detection report including the defect type and spatial location, and dynamically adjusts the excitation signal parameters to generate a second composite excitation signal; Weld defect report generation module: Based on the second composite excitation signal, recalculate the signal attenuation coefficient, extract the optimized scattering characteristics, and obtain high-precision weld defect positioning results.

Images