Hydrogen transmission pipeline defect detection method, system and device, and storage medium

Abstract

The application provides a hydrogen pipeline defect detection method and system, equipment and a storage medium, and belongs to the technical field of defect detection. The method comprises the following steps: acquiring phased array ultrasonic detection data, digital radiographic imaging data and hydrogen concentration data of a hydrogen pipeline; extracting defect boundary features based on the phased array ultrasonic detection data, extracting defect microstructure features based on the digital radiographic imaging data, and extracting hydrogen concentration spatiotemporal distribution features based on the hydrogen concentration data; constructing a modal correlation graph by taking the defect boundary features, the defect microstructure features and the hydrogen concentration spatiotemporal distribution features as nodes, and determining the edge weight between every two nodes in the modal correlation graph according to the mutual information of the two nodes; generating a multi-modal correlation feature vector based on the modal correlation graph; and obtaining a defect detection result of the hydrogen pipeline based on the multi-modal correlation feature vector. The application can improve the accuracy of hydrogen pipeline defect detection.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of defect detection, and more particularly relates to a hydrogen pipeline defect detection method and system, equipment and a storage medium. BACKGROUND

[0002] As a core facility for hydrogen energy storage and transportation, the safe operation of a hydrogen pipeline is directly related to the reliability of the hydrogen energy industry chain. Hydrogen has strong permeability and hydrogen embrittlement effect, which can easily lead to the deterioration of the microstructure of the pipeline material and cause defects such as crack propagation. However, the existing hydrogen pipeline defect detection technology mainly relies on a single detection method, and the limited information dimension can easily lead to defect misjudgment or missed detection, which is difficult to meet the demand for comprehensive and accurate defect detection of hydrogen pipelines, and a more accurate and comprehensive hydrogen pipeline defect detection method is urgently needed. SUMMARY

[0003] The application aims to provide a hydrogen pipeline defect detection method and system, equipment and a storage medium to improve the accuracy of hydrogen pipeline defect detection.

[0004] The first aspect of the application provides a hydrogen pipeline defect detection method, comprising:

[0005] Obtaining phased array ultrasonic detection data, digital radiographic imaging data and hydrogen concentration data of the hydrogen pipeline;

[0006] Extracting defect boundary features based on the phased array ultrasonic detection data, defect microstructure features based on the digital radiographic imaging data, and hydrogen concentration spatiotemporal distribution features based on the hydrogen concentration data;

[0007] Constructing a modal correlation graph by taking the defect boundary features, the defect microstructure features and the hydrogen concentration spatiotemporal distribution features as nodes, and determining the edge weight between each two nodes in the modal correlation graph by the mutual information of the two nodes; generating a multi-modal correlation feature vector based on the modal correlation graph;

[0008] Obtaining a defect detection result of the hydrogen pipeline based on the multi-modal correlation feature vector; the defect detection result includes a defect type and a defect position.

[0009] The second aspect of the application provides a hydrogen pipeline defect detection system, comprising:

[0010] A data acquisition module for acquiring phased array ultrasonic detection data, digital radiographic imaging data and hydrogen concentration data of the hydrogen pipeline;

[0011] A feature extraction module for extracting defect boundary features based on the phased array ultrasonic detection data, defect microstructure features based on the digital radiographic imaging data, and hydrogen concentration spatiotemporal distribution features based on the hydrogen concentration data;

[0012] a feature fusion module configured to construct a modality correlation graph by taking the defect boundary feature, the defect microstructure feature, and the hydrogen concentration spatiotemporal distribution feature as nodes, and determine the edge weight between each two nodes in the modality correlation graph according to the mutual information of the two nodes; and generate a multi-modal correlation feature vector based on the modality correlation graph;

[0013] a defect detection module configured to obtain a defect detection result of the hydrogen pipeline based on the multi-modal correlation feature vector, wherein the defect detection result comprises a defect type and a defect location.

[0014] In a third aspect, an electronic device is provided, which comprises a memory, a processor, and a computer program stored in the memory and running on the processor, and the processor implements the steps of the hydrogen pipeline defect detection method described above when executing the computer program.

[0015] In a fourth aspect, a computer readable storage medium is provided, which stores a computer program, and the computer program implements the steps of the hydrogen pipeline defect detection method described above when executed by a processor.

[0016] The hydrogen pipeline defect detection method and system, device, and storage medium provided by the embodiments of the present application have the following advantages: the traditional hydrogen pipeline defect detection method relies on a single detection means and has the problem of one-sided feature information. The embodiments of the present application achieve the all-around characterization of defects from macro to micro and from static form to dynamic evolution by extracting the defect boundary feature, microstructure feature, and hydrogen concentration spatiotemporal distribution feature respectively. The embodiments of the present application construct a modality correlation graph and quantify the correlation between features by mutual information, solve the problem of multi-source data heterogeneity, make the scattered features form an organic whole, and improve the completeness of feature representation. The multi-modal correlation feature vector of the embodiments of the present application fuses the complementary information of three types of sample data, has a more comprehensive learning basis than a single modal model, can effectively reduce the single data noise interference, and improve the accuracy of defect type and location judgment.

[0017] In summary, the present application significantly improves the comprehensiveness and accuracy of hydrogen pipeline defect detection through multi-modal feature fusion and correlation modeling, and provides more reliable technical support for hydrogen safety transportation. BRIEF DESCRIPTION OF DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort based on these drawings.

[0019] Figure 1A flowchart of a hydrogen pipeline defect detection method provided by an embodiment of the present application is shown in FIG. 1.

[0020] Figure 2 A sample pipe diagram of an artificial preset defect provided by an embodiment of the present application is shown in FIG. 2.

[0021] Figure 3 A structural block diagram of a hydrogen pipeline defect detection system provided by an embodiment of the present application is shown in FIG. 3.

[0022] Figure 4 A schematic block diagram of an electronic device provided by an embodiment of the present application is shown in FIG. 4. DETAILED DESCRIPTION

[0023] In the following description, for the purpose of explanation and not limitation, specific details are set forth, such as particular system configurations, techniques, etc., in order to provide a thorough understanding of the embodiments of the present application. However, it will be apparent to those skilled in the art that the present application can be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present application with unnecessary detail.

[0024] In order to make the objects, technical solutions and advantages of the present application clearer, the following will be described in conjunction with the accompanying drawings and specific embodiments.

[0025] Reference will be made to Figure 1 , Figure 1 A flowchart of a hydrogen pipeline defect detection method provided by an embodiment of the present application is shown in FIG. 1. The method can be executed by an electronic device, and specifically, the method can include S101-S104.

[0026] S101: Obtain phased array ultrasonic detection data, digital radiographic imaging data, and hydrogen concentration data of a hydrogen pipeline.

[0027] In the present embodiment, the phased array ultrasonic detection data refers to a structured echo signal matrix set containing pipeline matrix and defect reflection information collected by multi-angle and multi-array collaborative scanning of a hydrogen pipeline by a phased array ultrasonic probe array, wherein each matrix element corresponds to defect / matrix reflection signal amplitude and timing information under a specific array element, scanning angle, and depth. The digital radiographic imaging data refers to grayscale image data received by a detector after high-energy rays penetrate the pipeline. The hydrogen concentration data refers to a hydrogen concentration value set of the hydrogen pipeline along the line collected by a sensor array.

[0028] In this embodiment, considering that the hydrogen transmission pipeline is prone to hydrogen embrittlement due to strong hydrogen permeability, and existing single detection technology is difficult to fully characterize defects (such as macro morphology, micro damage and hydrogen-induced correlation), this embodiment obtains multi-modal data to capture defect features from different dimensions. For the acquisition of phased array ultrasonic testing data, specifically, phased array ultrasonic testing can generate a controllable acoustic beam by controlling the probe element delay excitation. When the sound wave propagates inside the pipeline, the defect interface will reflect the echo. After the echo (i.e. phased array ultrasonic testing data) is obtained in this embodiment, the defect position and morphology can be obtained by analyzing the time and amplitude of the echo. For the acquisition of digital radiographic imaging data, specifically, digital radiographic imaging utilizes the attenuation difference of different materials (defects and matrix) when the ray penetrates the pipeline to form a gray scale contrast image on the detector. Because the ray can penetrate the material, the microstructure such as grain boundary and inclusion can be presented. For the acquisition of hydrogen concentration data, specifically, hydrogen concentration detection can be realized by pre-installed sensor array adsorbing hydrogen to produce physical signals (such as resistance change). Because hydrogen is easy to permeate and related to defect evolution, this embodiment can correlate the pipeline defects by analyzing the hydrogen concentration data.

[0029] For example, this embodiment can periodically obtain phased array ultrasonic testing data through a phased array probe installed along the circumferential direction of the pipeline. The probe can be coupled with the outer wall of the pipeline, and the scanning parameters (such as frequency 5 MHz, array element 128, and angle step 0.5°) are set in advance. The probe can collect echo signals (i.e. phased array ultrasonic testing data) in real time and store them during the movement along the axial direction of the pipeline.

[0030] This embodiment can obtain digital radiographic imaging data by arranging a ray source (such as an X-ray machine) on one side of the pipeline and a flat panel detector installed at the corresponding position on the other side. The energy of the ray source can be 200 keV, and the exposure time can be 1 s-5 s to ensure clear image gray scale. The ray source flat panel detector can collect a cross-sectional image every 0.5 m along the axial direction of the hydrogen transmission pipeline.

[0031] This embodiment can obtain hydrogen concentration data in real time at a fixed frequency by installing hydrogen concentration sensors every 1 m along the axial direction of the outer wall of the pipeline. The hydrogen concentration sensors can be in close contact with the surface of the pipeline, and continuously record the hydrogen concentration values at each position. This embodiment can align the acquisition times of the three types of equipment through a time marker synchronizer to ensure that the multi-modal data at the same time and the same position can be correlated.

[0032] S102: Extract defect boundary features based on phased array ultrasonic testing data, extract defect microstructure features based on digital radiographic imaging data, and extract hydrogen concentration spatiotemporal distribution features based on hydrogen concentration data.

[0033] In the embodiment, there is no strict sequence limitation among the steps of extracting defect boundary features based on phased array ultrasonic detection data, extracting defect microstructure features based on digital radiographic imaging data, and extracting hydrogen concentration spatiotemporal distribution features based on hydrogen concentration data. For example, in the embodiment, the defect boundary features, the defect microstructure features, and the hydrogen concentration spatiotemporal distribution features can be extracted simultaneously.

[0034] In the embodiment, the phased array ultrasonic detection data includes a raw echo signal matrix obtained by circumferential scanning of an ultrasonic probe along a hydrogen conveying pipeline; the defect boundary features are extracted based on the phased array ultrasonic detection data, specifically including: performing noise reduction processing on the raw echo signal matrix to obtain a denoised echo signal matrix; generating a binary segmentation map based on the echo signal matrix, extracting a closed defect boundary based on the binary segmentation map; calculating defect geometric features, defect topological features, and defect texture features based on the closed defect boundary, and splicing the defect geometric features, the defect topological features, and the defect texture features to obtain the defect boundary features.

[0035] In the embodiment, the raw echo signal matrix refers to a set of reflection signals received by an array element of a phased array probe. The binary segmentation map refers to a black-and-white binary image converted from the echo signal matrix. The closed defect boundary refers to a continuous and unbroken defect contour line. The defect geometric features refer to parameters describing the shape of the defect, which can include length, maximum width, and aspect ratio, etc. The defect topological features refer to boundary spatial distribution features, which can include the number of concave-convex points and average curvature, etc. The defect texture features refer to the gray scale distribution features of the boundary region, which can include the mean gray value and the gradient standard deviation, etc.

[0036] In the embodiment, the phased array ultrasonic detection utilizes the reflection characteristics of sound waves at the defect interface. The raw echo signal matrix contains the reflection differences between the defect and the background, but due to the influence of electronic noise and material scattering, there will be interference in the raw echo signal matrix. The noise reduction processing of the embodiment can suppress irrelevant signals in the raw echo signal matrix, retain the true amplitude and timing characteristics of the defect echo, and provide more accurate and reliable data for subsequent segmentation. The embodiment converts the continuous signal into a black-and-white binary image through binary segmentation. The combination of the defect geometric features, the defect topological features, and the defect texture features can comprehensively describe the spatial properties of the defect and support the differentiation of defect types.

[0037] For example, the extraction process of the defect boundary features can include:

[0038] Noise reduction processing: The embodiment can use a wavelet noise reduction algorithm to decompose the original echo signal matrix by 3-5 layers, use soft threshold processing (for example, the threshold can be set to 1.5 times the noise standard deviation) on the high-frequency coefficients obtained by decomposition, and reconstruct the echo signal matrix after noise reduction to improve the signal-to-noise ratio to more than 20dB.

[0039] Generating a binary segmentation map: The embodiment can use an adaptive threshold method on the noise-reduced echo signal matrix to calculate a local threshold with a 5x5 pixel window, mark the pixels in the window whose echo amplitude exceeds the threshold as 1 (defects), and the rest as 0 (no defects), to obtain a binary segmentation map.

[0040] Extracting closed defect boundaries: The embodiment can perform 8-neighbor contour tracking on the segmentation map, use morphological closing operation (3x3 structure element) to connect broken edges, delete short boundaries with a length less than 0.5mm, and retain contour lines with a closure degree higher than 95% as closed defect boundaries.

[0041] Calculate features and splice: The embodiment can calculate geometric features through boundary pixel coordinates, for example, length is the straight-line distance between the starting point and the endpoint of the boundary, and the maximum width is the maximum distance perpendicular to the length direction, and the length and the maximum width are taken as geometric features; The embodiment can calculate topological features through the curvature formula, for example, calculate the curvature every 10 pixels, and take the mean and standard deviation as topological features; The embodiment can obtain texture features through the gray histogram statistics of the boundary region, for example, extract the pixel proportion of 10 gray intervals, and take the pixel proportion data of the 10 gray intervals as texture features. The embodiment can splice the three types of features in the order of geometry, topology and texture to form defect boundary features.

[0042] In the embodiment, the digital radiographic imaging data includes a gray-scale image of a hydrogen transmission pipeline cross section; based on the digital radiographic imaging data, microstructure features of defects are extracted, specifically including: dividing the gray-scale image into a plurality of pixel groups, and extracting gray-scale gradient data corresponding to each pixel group respectively; based on the gray-scale gradient data of the plurality of pixel groups, grain distribution features of the hydrogen transmission pipeline are extracted; a high gray abnormal discrete area with a gray value exceeding a first gray threshold in the gray-scale image is determined, and based on the high gray abnormal discrete area, a defect density feature is determined; the grain distribution features and the defect density features are taken as the microstructure features of the defects.

[0043] In the embodiment, the gray-scale image refers to a two-dimensional image generated after the ray penetrates the pipeline. A pixel group corresponds to a local area after the image is divided. The gray-scale gradient data refers to the gray-scale change rate of adjacent pixels in the pixel group, and can include gradient direction and amplitude. The grain distribution feature refers to the spatial distribution parameters of the material grains, which can include grain equivalent diameter, orientation angle, and distribution uniformity. The first gray-scale threshold is a gray-scale critical value for distinguishing normal areas from abnormal areas. The high gray-scale abnormal discrete area refers to an isolated pixel group with a gray-scale value exceeding the first gray-scale threshold. The parameters of the high gray-scale abnormal discrete area can include area and circularity. The defect density feature refers to the number of high gray-scale abnormal areas per unit area and the equivalent area ratio.

[0044] In the embodiment, digital radiography utilizes the difference in attenuation of different materials to rays. The attenuation coefficients of grains and defects in the gray-scale image are different, which is manifested as a difference in gray-scale values. The image is divided into pixel groups in the embodiment, which can reduce the computational complexity. The grain boundaries (i.e., gradient sudden change areas) can be identified by analyzing the gray-scale gradient in the group, and the distribution features such as grain size and orientation can be extracted. The high gray-scale abnormal discrete area corresponds to dense defects (such as inclusions and hydrogen bubbles) that are difficult for rays to penetrate. The defect density feature corresponding to the high gray-scale abnormal discrete area is used to represent the degree of micro-damage of the material (hydrogen-induced defects are usually manifested as high-density distribution). The correlation analysis of grain distribution and defect density in the embodiment can distinguish hydrogen-induced micro-damage from conventional material defects, providing a basis for judging the cause of the defects.

[0045] For example, the grain distribution feature is extracted, specifically including: the gray-scale image can be divided into non-overlapping pixel groups according to 5x5 pixels in the embodiment, and the gray-scale difference between adjacent pixels in each group is calculated to obtain the gray-scale gradient data (reflecting the degree of gray-scale change between pixels). Further, the global gray-scale gradient data of the entire image can be uniformly clustered, and the pixels with a gradient value greater than 50 are marked as grain boundaries (the atomic arrangement at the grain boundary is irregular, and the gray-scale mutation is significant); the region growing algorithm can be used to merge adjacent non-grain boundary pixels to form independent grain regions, and then the equivalent diameter (area equivalent circle diameter) and orientation angle (the angle between the longest axis of the grain boundary and the horizontal direction) of each grain are calculated, and the proportion of grains in different size intervals is counted, and finally the grain distribution feature is formed to comprehensively describe the size, arrangement, and distribution of normal grains.

[0046] Exemplarily, the defect density feature is extracted, specifically including: in the same gray image, for the abnormal area which has significant difference with the normal grain, the embodiment can use a first gray threshold to screen high gray pixels (defect area has stronger attenuation to the ray, and has higher gray value); the embodiment can merge the discrete high gray pixels by using a region growing algorithm (adjacent pixel gray difference is less than 30) to obtain a high gray abnormal discrete area with an area greater than 5 pixels; the embodiment can count the number of such areas in a unit area and the average distance between the areas to form a defect density feature, which is used to quantify the distribution density of micro defects.

[0047] After obtaining the grain distribution feature and the defect density feature, the embodiment can splice the grain distribution feature and the defect density feature in order to form a defect microstructure feature vector. The grain distribution feature reflects the microstructure state of the material matrix, and the defect density feature describes the abnormal damage in the matrix. The two features are complementary and can comprehensively represent the microstructure health state of the hydrogen transport pipeline, which provides a key basis for distinguishing hydrogen-induced defects (often accompanied by grain refinement and high-density micro defects) and conventional material defects.

[0048] In the embodiment, the hydrogen concentration spatiotemporal distribution feature is extracted based on the dynamic diffusion law of hydrogen leakage. Hydrogen-induced defects can cause local hydrogen concentration to fluctuate over time (time dimension) and diffuse along the pipeline axis to form a spatial concentration gradient (space dimension). By capturing the trend changes of the time series and the gradient differences of the spatial distribution, and combining the coupling relationship between the two, the embodiment can completely represent the dynamic evolution law of hydrogen concentration, and provide a basis for correlating the defect position (concentration gradient peak) and the activity (diffusion intensity).

[0049] Exemplarily, for extracting the time sequence feature of hydrogen concentration data at a single position, specifically including: the embodiment can fit the hydrogen concentration data at each pipeline position into a hydrogen concentration fitting curve (trend) and a change rate fitting curve (fluctuation speed), and fuse to form a time sequence feature. Then, according to the order of the pipeline axial position, the time sequence features of the positions are arranged into a hydrogen concentration spatial distribution sequence, and the position identification is associated. The embodiment can calculate the difference value of the time sequence features of adjacent positions, divide by the interval to obtain a spatial hydrogen concentration gradient sequence (quantifying spatial difference); then, the spatial hydrogen concentration gradient sequence is divided into multiple subsequences, and the concentration change rate of adjacent subsequences at the same position is calculated to obtain a spatial propagation feature; finally, the embodiment can fuse the spatial hydrogen concentration gradient sequence and the spatial propagation feature to form a hydrogen concentration spatiotemporal distribution feature.

[0050] S103: Defect boundary features, defect microstructure features, and hydrogen concentration spatiotemporal distribution features are used as nodes to construct a modal correlation graph, and the edge weight between each two nodes in the modal correlation graph is determined by the mutual information of the two nodes; a multi-modal correlation feature vector is generated based on the modal correlation graph.

[0051] In the embodiment, the modal correlation graph refers to a correlation network graph structure with three types of features as nodes, including defect boundary features, defect microstructure features, and hydrogen concentration spatiotemporal distribution features. The edge weight refers to the correlation strength value between nodes, which is calculated by mutual information, and ranges from 0 to 1, where 0 represents no correlation and 1 represents complete correlation. The multi-modal correlation feature vector refers to a comprehensive vector that fuses three types of features and correlation relationships, with a dimension equal to the sum of the dimensions of the three types of features and the sum of the number of edge weights.

[0052] In the embodiment, the calculation process of mutual information of two nodes specifically includes: determining the node feature vectors corresponding to the two nodes respectively, and calculating the mutual information of the two nodes through the joint probability distribution based on the node feature vectors corresponding to the two nodes (any two of the defect boundary features, the defect microstructure features, and the hydrogen concentration spatiotemporal distribution features) respectively.

[0053] In the embodiment, the formation and evolution of the hydrogen transmission pipeline defect are jointly affected by the macroscopic morphology, the microstructure, and the hydrogen concentration change, and the three types of features are internally correlated, such as the correlation between the hydrogen concentration gradient and the crack boundary expansion. The embodiment can quantify the dependency relationship between the features through mutual information, and the higher the mutual information value, the tighter the correlation. The embodiment constructs the modal correlation graph by taking mutual information as the edge weight, which can intuitively present the coupling law between the features. Based on the graph structure, the features can be fused while retaining the original information of each modality and incorporating the correlation relationship, so that the multi-modal correlation feature vector is closer to the nature of the defect, and the problem of ignoring the correlation of features in traditional simple splicing is solved, providing a more comprehensive input for subsequent model recognition.

[0054] For example, in the first step, the embodiment can determine the node features, specifically including: taking the defect boundary feature vector, the defect microstructure feature vector, and the hydrogen concentration spatiotemporal distribution feature vector as three nodes.

[0055] In the second step, the embodiment can calculate the mutual information, specifically including: the embodiment first determines the feature vectors corresponding to the two nodes to be calculated. For example, the defect boundary feature vector X (dimension M) and the defect microstructure feature vector Y (dimension N) are selected, where the elements of X are defect geometry, topology, and texture feature parameters, and the elements of Y are grain distribution and defect density feature parameters.

[0056] The embodiment performs discretization processing on the feature vectors corresponding to the two nodes determined, specifically: each dimension of X is divided into 5-10 intervals according to the numerical range (such as 0-1 mm, 1-3 mm, etc.), and similarly, each dimension of Y is divided into intervals, so that continuous features are converted into discrete symbols (such as using 0-9 to represent different intervals), ensuring the feasibility of subsequent probability calculation.

[0057] The embodiment can calculate the edge probability distribution and the joint probability distribution of the two node data after discretization processing, specifically including: counting the frequency of each discrete symbol appearing in X to obtain the edge probability distribution P(X) of X; similarly, the edge probability distribution P(Y) of Y is obtained. Counting the frequency of the combination of discrete symbols appearing in X and Y to obtain the joint probability distribution P(X, Y).

[0058] The embodiment can calculate the mutual information value based on the obtained joint probability distribution, specifically including: according to the definition of mutual information, taking the logarithm of the ratio of P(X, Y) to P(X)P(Y) of all discrete symbol combinations, multiplying P(X, Y) and accumulating to obtain the original mutual information value.

[0059] Finally, the embodiment can normalize the obtained original mutual information value: divide the original mutual information value by max(H(X), H(Y)) (H is the information entropy) to obtain a normalized mutual information value in the range of 0-1, which is the edge weight between the two nodes, representing the correlation strength between the features.

[0060] Through the above steps, the edge weight between the defect boundary feature vector and the defect microstructure feature vector, the defect microstructure feature vector and the hydrogen concentration spatiotemporal distribution feature vector, and the defect boundary feature vector and the hydrogen concentration spatiotemporal distribution feature vector can be finally obtained.

[0061] Finally, the embodiment can construct a modal correlation graph according to the determined nodes and the calculated edge weights, specifically including: taking the nodes as vertices and the edge weights as connection strengths to generate a graph structure containing node feature values and edge weights. The embodiment can generate a multi-modal correlation feature vector, specifically including: concatenating the three types of feature vectors in node order, and then merging the three groups of edge weight values to form a fusion feature vector.

[0062] S104: obtaining a defect detection result of the hydrogen transmission pipeline based on the multi-modal correlation feature vector; the defect detection result includes a defect type and a defect position.

[0063] In the embodiment, the defect detection result of the hydrogen transmission pipeline can be generated based on the phased array ultrasonic detection data, digital radiographic imaging data and hydrogen concentration data of the hydrogen transmission pipeline through a target defect recognition model; the target defect recognition model is trained based on a target training sample set. Each training sample in the target training sample set includes phased array ultrasonic detection sample data, digital radiographic imaging sample data, hydrogen concentration sample data and a defect label corresponding to the sample data.

[0064] In this embodiment, the target defect recognition model is a machine learning model for hydrogen pipeline defect detection, and the model parameters of the target defect recognition model can include 3-5 layers of network layers, 64-256 hidden layer nodes, activation function type, and 100-500 iterations. The defect types can include hydrogen-induced cracks, hydrogen bubbling, and mechanical damage, etc., and each defect type corresponds to a unique identifier. The defect location refers to the three-dimensional coordinates of the defect on the pipeline, and the parameters include axial mileage 0-1000m, circumferential angle 0-360°, and radial depth 0-50mm, etc. The target training sample set refers to a data set used for training the target defect recognition model, which contains multiple training samples, each training sample containing phased array ultrasonic, digital ray, and hydrogen concentration data and corresponding labels. The defect label refers to the defect information corresponding to the sample data, which can include defect type labels 1-5 (corresponding to different defects) and defect location coordinates.

[0065] In this embodiment, the target defect recognition model learns the mapping relationship between the phased array ultrasonic detection sample data, digital ray imaging sample data, and hydrogen concentration sample data in the training sample and the defect label, and establishes a nonlinear mapping from the input data to the output result. During model training, the difference between the predicted result and the true label is minimized by optimizing the parameters, so that the model has the generalization ability to new samples. When performing defect detection on the hydrogen pipeline, the obtained phased array ultrasonic detection data, digital ray imaging data, and hydrogen concentration data are input into the model, the model can extract features from the input data, fuse the extracted features to obtain a multi-modal correlation feature vector, and then output the defect type and location based on the training learning and defect rules. The multi-modal correlation feature vector solves the problems of large recognition error and inaccurate positioning caused by incomplete information in traditional methods, and is especially suitable for complex defect detection of hydrogen-induced defects affected by multiple factors.

[0066] For example, the target training sample set can be prepared, specifically including: collecting phased array ultrasonic detection data, digital ray imaging data, and hydrogen concentration data sample data of the hydrogen pipeline with pre-set defects, labeling the corresponding defect type and location as labels, and dividing them into a training set and a validation set in a ratio of 8:2.

[0067] The model can be trained, specifically including: initializing the structure of the target defect recognition model, inputting the multi-modal correlation feature vector of the training set into the model, calculating the difference between the predicted label and the true label with the cross-entropy loss function, iteratively optimizing the model parameters with the gradient descent method, evaluating the accuracy with the validation set every round of iteration, until the accuracy improvement of the last 5 rounds is less than 0.5%, and saving the model parameters.

[0068] The embodiment can utilize a model to perform defect detection, specifically including: obtaining three types of data of a hydrogen conveying pipeline to be detected and inputting the trained model, the model outputs a defect type identifier and three-dimensional coordinates, the output result is verified (such as whether the position coordinates are within the pipeline range), and finally a defect detection result is output.

[0069] From the above, it can be concluded that the embodiment comprehensively captures defect characteristics from three dimensions of macroscopic morphology, microscopic structure and hydrogen-induced correlation by fusing phased array ultrasonic detection data, digital radiographic imaging data and hydrogen concentration data, solving the problem that a single detection technology cannot completely represent the defects of the hydrogen conveying pipeline. The defect boundary characteristics of the phased array ultrasonic, the microscopic structure characteristics of the digital radiographic and the spatiotemporal distribution characteristics of the hydrogen concentration complement each other, and can reflect the evolution law of the defects in all directions.

[0070] The embodiment constructs a modal correlation graph and generates a multi-modal correlation feature vector by mutual information, retains the internal correlation between different features, and avoids information loss caused by traditional feature simple splicing. This deep fusion method makes the feature representation closer to the nature of the defect, significantly improves the discrimination ability of the target defect recognition model for complex defects, and especially improves the recognition accuracy of hydrogen-induced cracks and mechanical damage and other easily confused defects.

[0071] The target defect recognition model is trained based on a multi-modal training sample set, and can accurately output the defect type and three-dimensional position by capturing the hydrogen embrittlement effect in combination with the spatiotemporal distribution characteristics of the hydrogen concentration, effectively meeting the high safety detection needs of the hydrogen conveying pipeline and providing reliable protection for hydrogen energy storage and transportation safety.

[0072] In an embodiment of the present application, the hydrogen concentration data includes hydrogen concentration data of multiple pipeline positions in the hydrogen conveying pipeline within a target time period; the hydrogen concentration spatiotemporal distribution feature is extracted based on the hydrogen concentration data, including: for each pipeline position in the hydrogen conveying pipeline, based on the hydrogen concentration data of the pipeline position within the target time period, a hydrogen concentration time sequence feature of the pipeline position within the target time period is extracted; based on all pipeline positions in the hydrogen conveying pipeline and the hydrogen concentration time sequence features within the target time period corresponding to each pipeline position, a hydrogen concentration spatial distribution sequence is generated; the hydrogen concentration spatial distribution sequence includes hydrogen concentration time sequence features arranged in order according to the spatial position relationship, and each hydrogen concentration time sequence feature corresponds to a position identifier; the hydrogen concentration spatiotemporal distribution feature is extracted based on the hydrogen concentration spatial distribution sequence.

[0073] In the embodiment, the target time period refers to a continuous time interval of hydrogen concentration monitoring. The pipeline position refers to a monitoring point arranged along the axis of the hydrogen conveying pipeline, and the position identifier is an axial mileage (such as 0m, 1m…10m, etc.) or a number (1-10, etc.). The hydrogen concentration data refers to a set of concentration values of each position within the target time period.

[0074] In the embodiment, the hydrogen concentration time sequence feature in the pipeline position target time period is extracted based on the hydrogen concentration data in the pipeline position target time period, specifically including: determining a first time window based on a data variation span of the hydrogen concentration data in the pipeline position target time period; the data variation span is determined by a maximum hydrogen concentration value and a minimum hydrogen concentration value in the hydrogen concentration data, and the data variation span is negatively correlated with the first time window; calculating a mean value sequence and a standard deviation sequence of the hydrogen concentration data of the pipeline position in the target time period according to the first time window; generating a hydrogen concentration fitting curve of the pipeline position in the target time period based on the mean value sequence, and generating a hydrogen concentration change rate fitting curve of the pipeline position in the target time period based on the standard deviation sequence; determining the hydrogen concentration time sequence feature in the pipeline position target time period based on the hydrogen concentration fitting curve and the hydrogen concentration change rate fitting curve.

[0075] In the embodiment, the mean value sequence and the standard deviation sequence of the hydrogen concentration data of the pipeline position in the target time period are calculated according to the first time window, specifically including: dividing the target time period into a plurality of time blocks according to the first time window; calculating the mean value of the hydrogen concentration data corresponding to each time block, and splicing the mean values of the hydrogen concentration data corresponding to each time block in time sequence as the mean value sequence of the hydrogen concentration data; calculating the standard deviation of the hydrogen concentration data corresponding to each time block, and splicing the standard deviations of the hydrogen concentration data corresponding to each time block in time sequence as the standard deviation sequence of the hydrogen concentration data.

[0076] In the embodiment, the data variation span refers to the concentration fluctuation range of a single position, which can be specifically the difference between the maximum concentration value and the minimum concentration value. The first time window refers to the time interval for calculating the statistical feature, and the parameters can include the window length such as 1 minute. The first time window is negatively correlated with the variation span, and the larger the span is, the smaller the window time is. The mean value sequence refers to the sequence composed of the concentration average values in each time window, and the length of the mean value sequence is the target time period / window length (such as 24 hours / 30 minutes=48 points). The standard deviation sequence refers to the sequence composed of the concentration standard deviations in each time window, and the length of the standard deviation sequence is consistent with that of the mean value sequence. The hydrogen concentration fitting curve refers to the curve obtained by smoothing fitting the mean value sequence. The hydrogen concentration change rate fitting curve refers to the curve obtained by fitting the change rate of the standard deviation sequence, wherein the negative value is the decline rate and the positive value is the rise rate. The hydrogen concentration time sequence feature refers to the feature vector fused with the fitting curve and the change rate curve.

[0077] For example, the process of extracting the hydrogen concentration time sequence feature based on the hydrogen concentration data can include:

[0078] (1) Determine the target time period and pipeline location: This embodiment can select 24 consecutive hours as the target time period, and set 50 monitoring points (position identifiers 1-50) every 1 meter along the pipeline axis, collect hydrogen concentration data at each position, wherein the sampling frequency is set to 5Hz, and the mean value is taken every 1 minute for downsampling.

[0079] (2) Calculate the data variation span and the first time window: For each position, this embodiment can calculate the maximum concentration value and the minimum concentration value in the target time period, and calculate the variation span (for example, the maximum concentration value of position 10 is 800ppm, and the minimum concentration value is 300ppm, so the span is 500ppm); this embodiment can determine the first time window according to the negative correlation rule, for example, when the span is greater than 300ppm, the first time window is set to 10 minutes, when the span is between 100-300ppm, the first time window is set to 20 minutes, and when the span is less than 100ppm, the first time window is set to 30 minutes.

[0080] (3) Generate mean value sequence and standard deviation sequence: This embodiment can divide the concentration data of the position according to the first time window, calculate the concentration mean value and standard deviation of each window, and form the mean value sequence and the standard deviation sequence.

[0081] (4) Fit the curve and extract the time sequence features: This embodiment can perform a second-order polynomial fitting on the mean value sequence to obtain a hydrogen concentration fitting curve, and extract the peak value, valley value and average slope of the curve; this embodiment can calculate the difference (divided by the window interval) of adjacent windows for the standard deviation sequence, perform linear fitting to obtain a variation rate curve, and extract the maximum value and minimum value of the variation rate; this embodiment can concatenate the two types of features such as the peak value, valley value and average slope of the curve, and the maximum value and minimum value of the variation rate into a hydrogen concentration time sequence feature vector.

[0082] In this embodiment, the hydrogen concentration spatiotemporal distribution features are extracted based on the hydrogen concentration spatial distribution sequence, which specifically includes: calculating the difference of the hydrogen concentration time sequence features corresponding to each two adjacent position identifiers in the hydrogen concentration spatial distribution sequence to obtain a spatial hydrogen concentration difference sequence; generating a spatial hydrogen concentration gradient sequence based on the spatial hydrogen concentration difference sequence; dividing the hydrogen concentration spatial distribution sequence into hydrogen concentration spatial distribution subsequences of multiple time periods according to a second time window; the second time window is greater than the first time window; calculating the hydrogen concentration variation rate of the same position identifier in each two adjacent hydrogen concentration spatial distribution subsequences; extracting the spatial propagation features of the hydrogen concentration variation rate based on the hydrogen concentration variation rates of the same position identifier in all hydrogen concentration spatial distribution subsequences of the time periods; the spatial propagation features include spatial propagation direction and spatial propagation intensity; the spatial hydrogen concentration gradient sequence and the spatial propagation features are taken as the hydrogen concentration spatiotemporal distribution features.

[0083] In the embodiment, the hydrogen concentration spatial distribution sequence refers to a time sequence feature set arranged in the order of position identification, the sequence length of the hydrogen concentration spatial distribution sequence is the total number of positions, and each element of the hydrogen concentration spatial distribution sequence is a time sequence feature vector of a corresponding position. The spatial hydrogen concentration difference sequence refers to a difference value set of time sequence features of adjacent positions, such as a difference value of a mean sequence of positions i and i+1. The spatial hydrogen concentration gradient sequence refers to a gradient value set obtained by dividing the difference sequence by the interval between positions. The second time window refers to a time interval for dividing the spatial distribution sequence. The second time window is an integer multiple of the first time window. The first time window is used for hydrogen concentration time sequence feature extraction of a single pipeline position, and is used for capturing fine-grained fluctuations of local concentration (such as concentration sudden rise / sudden drop in a short time). A small window is needed to achieve high time resolution to adapt to the scene of rapid change of hydrogen concentration (such as the initial stage of defect leakage).

[0084] In the embodiment, the second time window is used for dividing the hydrogen concentration spatial distribution sequence, and is used for analyzing the spatial propagation law of concentration change (such as the diffusion trend from a certain position to the adjacent region). A long enough time is needed to reflect the cumulative effect of spatial propagation. If the window is too small, the propagation feature cannot be shown due to insufficient time span. If it is an integer multiple of the first time window, a plurality of fine-grained time sequence features can be integrated into a spatial sub-sequence, which not only retains the time sequence details, but also ensures that the spatial analysis has enough time dimension support. For example, when the first time window is 5 minutes, the second time window can be 30 minutes, that is, it contains data of 6 first time windows. This size relationship design can make the fine-grained time sequence features of the first time window provide basic data for the spatial propagation analysis of the second time window, and the large span of the second time window ensures the recognizability of the spatial law, which adapts to the hierarchical needs of hydrogen concentration space-time distribution feature extraction.

[0085] In the embodiment, the hydrogen concentration spatial distribution sub-sequence refers to the spatial distribution sequence in a single second time window. The hydrogen concentration change rate refers to the ratio of the concentration change amount of the same position in adjacent sub-sequences to the time interval. The spatial propagation feature includes the propagation direction (axial positive direction / negative direction / undirectional) and the propagation intensity (cumulative sum of hydrogen concentration change rate). The hydrogen concentration space-time distribution feature refers to a vector that fuses the gradient sequence and the propagation feature.

[0086] In this embodiment, the hydrogen concentration of a single pipeline location dynamically changes over time, and the fluctuation law (such as the rising / falling trend, fluctuation amplitude) is closely related to the defect activity (such as hydrogen-induced defect leakage leading to continuous concentration rise). The data change span reflects the intensity of concentration fluctuation: when the fluctuation is large (such as in the early stage of leakage), a short time window (high time resolution) is needed to capture the details, and when the fluctuation is small (such as in the stable diffusion stage), a long time window (strong noise suppression) is used. This negative correlation design can balance the time resolution and noise suppression. The mean sequence and the standard deviation sequence are used to represent the average level of concentration and the stability of fluctuation, respectively. The fitting curve can smooth the random noise and highlight the trend change; the change rate curve quantifies the fluctuation speed, and the combination of the two forms a time sequence feature that can completely describe the time dynamic law of the unit location.

[0087] In this embodiment, the hydrogen leakage of the hydrogen delivery pipeline defect has a spatial propagation characteristic (such as axial diffusion), and needs to be combined with multi-location data to capture the spatial law. The spatial hydrogen concentration difference sequence reflects the concentration difference between adjacent locations, and the gradient sequence further quantifies the concentration change per unit distance (a sharp increase in gradient usually corresponds to the vicinity of the defect). By dividing the sub-sequences according to the second time window, the time-space problem can be transformed into an analysis mode that combines time slicing with spatial distribution, which facilitates tracking the spatial propagation process of concentration change. This embodiment can identify the propagation direction (such as continuous diffusion from location i to i+1) and the propagation intensity of the change trend by calculating the same location change rate of adjacent sub-sequences. The higher the cumulative value of the change rate, the more significant the hydrogen diffusion. This embodiment combines the spatial hydrogen concentration gradient sequence and the spatial propagation characteristics of the hydrogen concentration change rate, which can simultaneously represent the spatial non-uniformity and dynamic diffusion law of hydrogen concentration, and provides key evidence for correlating the defect location (gradient peak) and activity (high propagation intensity corresponds to high activity).

[0088] For example, the process of extracting hydrogen concentration spatio-temporal distribution features based on the hydrogen concentration spatial distribution sequence can include:

[0089] (1) Generate a hydrogen concentration spatial distribution sequence: In this embodiment, the time sequence feature vectors of each location are arranged in order according to the location identifiers 1-50 to form a spatial distribution sequence.

[0090] (2) Calculate the spatial hydrogen concentration difference and spatial hydrogen concentration gradient sequence: In this embodiment, the time sequence feature vectors of adjacent locations (such as 1 and 2, 2 and 3, …, 49 and 50) are calculated by dimension-by-dimension difference (such as the difference of the peak value of the mean sequence), to obtain a spatial hydrogen concentration difference sequence containing 49 elements; divide each difference by the location interval (1 meter) to generate a spatial hydrogen concentration gradient sequence.

[0091] (3) Subsequence of spatial distribution of hydrogen concentration is divided: the target time period (24 hours) can be divided into 24 time periods according to the second time window (for example, 1 hour), and each time period corresponds to a spatial distribution subsequence.

[0092] (4) Change rate of hydrogen concentration and spatial propagation characteristics are calculated: the change amount (for example, peak value change) of the time sequence characteristics of the same position of the adjacent time period subsequences (for example, the first and the second, the second and the third, and the eleventh and the twelfth) can be calculated, and the change rate is obtained by dividing the time interval (1 hour); the signs (positive / negative) of the change rates of the positions can be counted, if the change rates of the continuous three adjacent positions are all positive, the propagation direction is the positive direction of the axial direction (and vice versa); the cumulative sum of the change rates in the propagation direction can be taken as the propagation intensity.

[0093] (5) Spatial and temporal distribution characteristics are integrated: the spatial hydrogen concentration gradient sequence, the propagation direction (encoded as 1 / -1), and the propagation intensity are spliced to form a hydrogen concentration spatial and temporal distribution characteristic vector, which is used for subsequent modal correlation graph construction.

[0094] In this embodiment, the time sequence characteristics of hydrogen concentration are extracted by dynamically adapting the time window design, the first time window is adjusted according to the data change span, the details are captured at a high time resolution when the concentration fluctuates dramatically, and the noise is suppressed by a longer window in the stable stage, effectively balancing the accuracy and anti-interference ability of the time sequence characteristics, and the dynamic change law of the hydrogen concentration of the unit position is completely described.

[0095] Meanwhile, in combination with the spatial dimension analysis, the difference and gradient of the adjacent positions are calculated to accurately locate the concentration mutation area, and provide a key basis for defect position identification; the spatial distribution sequence is divided by using the second time window which is greater than the first time window and is an integer multiple of the first time window, which not only retains the fine-grained time sequence information, but also ensures the recognizability of the spatial propagation characteristics, and by analyzing the spatial propagation direction and intensity of the change rate, the diffusion trend of the hydrogen leakage can be effectively captured, and the defect activity can be quantified.

[0096] Overall, the spatial and temporal distribution characteristics of the embodiment fuse the spatial gradient and the propagation characteristics, which can comprehensively represent the spatial and temporal dynamic law of the hydrogen concentration, provide a reliable basis for accurate positioning and activity evaluation of the hydrogen transmission pipeline defects, significantly improve the scientificity and practicality of the hydrogen concentration monitoring, and ensure the safe operation of the hydrogen transmission system.

[0097] In an embodiment of the present application, the target training sample set is obtained by: obtaining phased array ultrasonic detection data, digital radiographic imaging data and hydrogen concentration data of a preset defect hydrogen transmission pipeline; taking the phased array ultrasonic detection data, the digital radiographic imaging data, the hydrogen concentration data and the defect label of the preset defect hydrogen transmission pipeline as a first training sample set; the defect label includes a defect type and a defect position.

[0098] For each training sample in the first training sample set, at least one of the phased array ultrasonic detection data, the digital radiographic imaging data and the hydrogen concentration data in the training sample is data-augmented based on a defect label of the training sample, to obtain an augmented generated second training sample set;

[0099] The first training sample set and the second training sample set are taken as a target training sample set.

[0100] In the embodiment, the hydrogen delivery pipeline with a preset defect refers to a pipeline specimen artificially processed with a known defect, the defect types include hydrogen-induced cracks, hydrogen bubbling and mechanical damage, etc., the defect parameters include a length of 0.5-10 mm, a width of 0.1-2 mm and a depth of 0.2-5 mm, etc., the pipeline material is consistent with the actual hydrogen delivery pipeline (such as X65 steel). The first training sample set refers to a data set containing original detection data and labels, each sample contains phased array ultrasonic data, digital radiographic data and hydrogen concentration data and a defect label, and the sample amount can be 100-500 groups. Data augmentation refers to a process of generating new samples based on original samples. The second training sample set refers to an augmented generated sample set, and the target training sample set refers to a union set of the first sample set and the second sample set.

[0101] In the embodiment, the original preset defect sample can provide basic data, but the number is limited and it is difficult to cover all working conditions. The data augmentation based on the defect label can simulate the variation scene (such as probe angle deviation and ray energy fluctuation) in the actual detection under the premise of retaining the defect core features (such as crack direction, hydrogen concentration gradient and the relevance of the defect), and increase the sample diversity. For example, for a hydrogen-induced crack sample, the crack boundary topology and the spatial correlation of the high hydrogen concentration area are maintained during the augmentation, and the ultrasonic echo intensity is adjusted to simulate different detection distances, so that the model learns the essential features of the defect rather than accidental noise, and the generalization ability to the real scene is improved.

[0102] For example, the acquisition process of the target training sample set specifically includes:

[0103] (1) Constructing the first training sample set: in the embodiment, a phased array ultrasonic probe, a ray source and a hydrogen concentration sensor can be arranged on the preset defect pipeline to collect three types of data, and historical defect detection data of the hydrogen delivery pipeline can also be acquired; then the defect type and the defect position are manually labeled to form the first training sample set.

[0104] (2) Generating a second training sample set: For each sample, the embodiment can be augmented based on the defect label: for example, for ultrasonic data, the embodiment can adjust the echo amplitude according to the defect depth (the amplitude is attenuated by 0.5 dB for every 1 mm increase in depth); for radiographic data, the embodiment can adjust the grayscale according to the defect density (the grayscale value of the high-density defect area is ±8%); for hydrogen concentration data, the embodiment can simulate the diffusion rate according to the defect activity (the concentration of high-activity defects increases by 10%), and the augmented data and the initial sample data are randomly combined to form a second training sample set.

[0105] (3) Finally, the embodiment can combine the first training sample set and the second training sample set, remove duplicate or abnormal samples, and form a target training sample set for training of the target defect recognition model.

[0106] As shown in FIG. 1, Figure 2 Figure 2 is a schematic diagram of a sample pipe provided with artificial preset defects by the embodiment. In the figure, N5 longitudinal external notch refers to an N-type notch with a length direction parallel to the axis direction of the sample pipe and a depth of 5% of the nominal wall thickness of the sample pipe, A and C in the figure are external edge notches located on the base metal next to the edge of the weld, and B is an external middle notch located in the center of the weld. N5 longitudinal internal notch refers to an N-type notch with a length direction parallel to the axis direction of the sample pipe and a depth of 5% of the nominal wall thickness of the sample pipe, E and G in the figure are internal edge notches located on the base metal next to the edge of the weld, and F is an internal middle notch located in the center of the weld. Transverse internal and external notches refer to N-type notches with a length direction perpendicular to the axis direction of the sample pipe, crossing the weld, and a depth of 5% of the nominal wall thickness of the sample pipe. H and I in the figure are internal and external transverse notches, respectively. D is a through hole with a diameter of 1.6 mm, L, M, N, and O are four groups of flat-bottom holes with a diameter of 3.0 mm and perpendicular to the bevel face, J, K, P, and Q are four groups of flat-bottom holes with a diameter of 3.0 mm and perpendicular to the bevel face, and S and T are pipe end blind holes with a diameter of 3.2 mm, and U and V are hot zone layered holes with a diameter of 6.0 mm.

[0107] The first training sample set of the embodiment is constructed based on the real detection data of the preset defect pipeline, ensuring accurate association of the original features and the defect labels, and providing reliable basic samples for the model. The second training sample set is generated by the embodiment through targeted augmentation based on the defect label, which can cover more actual detection scene variations (such as probe deviation and energy fluctuation) while preserving the core features of the defects (such as the association of crack topology and hydrogen concentration), effectively solving the problems of insufficient original sample quantity and incomplete working condition coverage. The target training sample set obtained by combining the two can make the model fully learn the essential features of the defects and the laws of various scenes, significantly improve the recognition generalization ability of various defects in the hydrogen transmission pipeline, and provide strong support for the accuracy of the detection results.

[0108] Corresponding to the hydrogen transmission pipeline defect detection method of the above embodiment, Figure 3 ​A structural block diagram of a hydrogen pipeline defect detection system provided by an embodiment of the present application is shown. For ease of illustration, only parts related to the embodiments of the present application are shown. For reference Figure 3 The hydrogen pipeline defect detection system 20 includes a data acquisition module 21, a feature extraction module 22, a feature fusion module 23, and a defect detection module 24.

[0109] The data acquisition module 21 is configured to acquire phased array ultrasonic detection data, digital radiographic imaging data, and hydrogen concentration data of the hydrogen pipeline.

[0110] The feature extraction module 22 is configured to extract defect boundary features based on the phased array ultrasonic detection data, defect microstructure features based on the digital radiographic imaging data, and hydrogen concentration spatiotemporal distribution features based on the hydrogen concentration data.

[0111] The feature fusion module 23 is configured to construct a modality correlation graph by taking the defect boundary features, the defect microstructure features, and the hydrogen concentration spatiotemporal distribution features as nodes, and determine an edge weight between each two nodes in the modality correlation graph based on mutual information of the two nodes; and generate a multi-modal correlation feature vector based on the modality correlation graph.

[0112] The defect detection module 24 is configured to obtain a defect detection result of the hydrogen pipeline based on the multi-modal correlation feature vector; the defect detection result includes a defect type and a defect location.

[0113] In an embodiment of the present application, the phased array ultrasonic detection data includes an original echo signal matrix obtained by a ultrasonic probe scanning along a circumferential direction of the hydrogen pipeline; and the feature extraction module 22 is specifically configured to:

[0114] perform noise reduction processing on the original echo signal matrix to obtain a denoised echo signal matrix;

[0115] generate a binary segmentation map based on the echo signal matrix, and extract a closed defect boundary based on the binary segmentation map; the closed defect boundary is used to represent a spatial profile of the defect in the hydrogen pipeline, so as to reflect a spatial distribution and a morphological attribute of the defect;

[0116] calculate defect geometric features, defect topological features, and defect texture features based on the closed defect boundary, and splice the defect geometric features, the defect topological features, and the defect texture features to obtain the defect boundary features.

[0117] In an embodiment of the present application, the digital radiographic imaging data includes a gray-scale image of a cross section of the hydrogen pipeline; and the feature extraction module 22 is specifically further configured to:

[0118] divide the gray-scale image into a plurality of pixel groups, and extract gray-scale gradient data corresponding to each pixel group, respectively;

[0119] The grain distribution characteristics of the hydrogen transmission pipeline are extracted based on the gray gradient data of the plurality of pixel groups;

[0120] The high gray abnormal discrete area with the gray value exceeding the first gray threshold in the gray image is determined, and the defect density characteristics are determined based on the high gray abnormal discrete area. The grain distribution characteristics and the defect density characteristics are used as the defect microstructure characteristics.

[0121] In an embodiment of the present application, the hydrogen concentration data includes hydrogen concentration data of a plurality of pipeline positions in the hydrogen transmission pipeline within a target time period; the feature extraction module 22 is specifically further configured to:

[0122] For each pipeline position in the hydrogen transmission pipeline, the hydrogen concentration time sequence characteristics of the pipeline position within the target time period are extracted based on the hydrogen concentration data of the pipeline position within the target time period;

[0123] The hydrogen concentration spatial distribution sequence is generated based on the hydrogen concentration time sequence characteristics of all pipeline positions in the hydrogen transmission pipeline and the target time period corresponding to each pipeline position. The hydrogen concentration spatial distribution sequence includes hydrogen concentration time sequence characteristics arranged in order according to the spatial position relationship, and each hydrogen concentration time sequence characteristic corresponds to a position identifier;

[0124] The hydrogen concentration spatiotemporal distribution characteristics are extracted based on the hydrogen concentration spatial distribution sequence.

[0125] In an embodiment of the present application, the feature extraction module 22 is specifically further configured to:

[0126] A first time window is determined based on the data variation span of the hydrogen concentration data within the target time period of the pipeline position. The data variation span is determined by the maximum hydrogen concentration value and the minimum hydrogen concentration value in the hydrogen concentration data, and the data variation span is negatively correlated with the first time window;

[0127] The mean sequence and the standard deviation sequence of the hydrogen concentration data of the pipeline position within the target time period are calculated according to the first time window;

[0128] The hydrogen concentration fitting curve of the pipeline position within the target time period is generated based on the mean sequence, and the hydrogen concentration rate fitting curve of the pipeline position within the target time period is generated based on the standard deviation sequence;

[0129] The hydrogen concentration time sequence characteristics of the pipeline position within the target time period are determined based on the hydrogen concentration fitting curve and the hydrogen concentration rate fitting curve.

[0130] In an embodiment of the present application, the feature extraction module 22 is specifically further configured to:

[0131] The difference value calculation is performed on the hydrogen concentration time sequence characteristics corresponding to each two adjacent position identifiers in the hydrogen concentration spatial distribution sequence, and the spatial hydrogen concentration difference sequence is obtained;

[0132] generate a spatial hydrogen concentration gradient sequence based on the spatial hydrogen concentration difference sequence;

[0133] divide the hydrogen concentration spatial distribution sequence into a plurality of hydrogen concentration spatial distribution subsequences according to a second time window; the second time window is larger than the first time window;

[0134] calculate a hydrogen concentration change rate of the same position identifier in each two adjacent hydrogen concentration spatial distribution subsequences;

[0135] extract a spatial propagation feature of the hydrogen concentration change rate based on the hydrogen concentration change rate of the same position identifier in all hydrogen concentration spatial distribution subsequences; the spatial propagation feature includes a spatial propagation direction and a spatial propagation intensity;

[0136] use the spatial hydrogen concentration gradient sequence and the spatial propagation feature as the hydrogen concentration spatiotemporal distribution feature.

[0137] In an embodiment of the present application, the defect detection result of the hydrogen transmission pipeline is generated based on phased array ultrasonic detection data, digital radiographic imaging data and hydrogen concentration data of the hydrogen transmission pipeline, and through a target defect identification model; the target defect identification model is obtained based on a target training sample set;

[0138] The target training sample set is obtained in the following manner:

[0139] obtain phased array ultrasonic detection data, digital radiographic imaging data and hydrogen concentration data of the hydrogen transmission pipeline with a preset defect; use the phased array ultrasonic detection data, the digital radiographic imaging data, the hydrogen concentration data and a defect label of the hydrogen transmission pipeline with the preset defect as a first training sample set; the defect label includes a defect type and a defect position;

[0140] For each training sample in the first training sample set, perform data augmentation on at least one of the phased array ultrasonic detection data, the digital radiographic imaging data and the hydrogen concentration data in the training sample based on the defect label of the training sample, to obtain a second training sample set generated by the augmentation;

[0141] use the first training sample set and the second training sample set as the target training sample set.

[0142] Referring to Figure 4 , Figure 4 The schematic block diagram of the electronic device provided in an embodiment of the present application is shown in FIG. 1. As shown in FIG. 1, the electronic device includes a processor 10 and a memory 20. Figure 4The electronic device 300 in the embodiment shown can include one or more processors 301, one or more input devices 302, one or more output devices 303, and one or more memories 304. The above-mentioned processors 301, input devices 302, output devices 303, and memories 304 complete communication with each other through a communication bus 305. The memory 304 is configured to store a computer program, and the computer program includes program instructions. The processor 301 is configured to execute the program instructions stored in the memory 304. Specifically, the processor 301 is configured to invoke the program instructions to execute the functions of various modules in the above-mentioned system embodiments, for example Figure 3 The functions of the data acquisition module 21, the feature extraction module 22, the feature fusion module 23, and the defect detection module 24 shown are described.

[0143] It should be understood that, in the embodiments of the present application, the processor 301 can be a central processing unit (CPU), and the processor can also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor.

[0144] The input device 302 can include a touchpad, a fingerprint acquisition sensor (used to acquire fingerprint information and direction information of a fingerprint of a user), a microphone, etc., and the output device 303 can include a display (LCD, etc.), a loudspeaker, etc.

[0145] The memory 304 can include read-only memory and random access memory, and provide instructions and data for the processor 301. A portion of the memory 304 can also include non-volatile random access memory. For example, the memory 304 can also store information of hydrogen concentration data.

[0146] In specific implementations, the processor 301, the input device 302, and the output device 303 described in the embodiments of the present application can execute the implementation manners described in the embodiments of the hydrogen pipeline defect detection method provided by the embodiments of the present application, and can also execute the implementation manners of the electronic device 300 described in the embodiments of the present application, which will not be described here.

[0147] In another embodiment of the present application, a computer readable storage medium is provided, which stores a computer program. The computer program includes program instructions, which, when executed by a processor, implement all or part of the processes of the above-mentioned embodiment methods. The computer program can also instruct related hardware to complete the implementation. The computer program can be stored in a computer readable storage medium. When the computer program is executed by the processor, the steps of the above-mentioned various method embodiments can be implemented. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or some intermediate form. The computer readable medium can include any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc.

[0148] The computer readable storage medium can be an internal storage unit of the electronic device of any of the preceding embodiments, such as a hard disk or a memory of the electronic device. The computer readable storage medium can also be an external storage device of the electronic device, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. Further, the computer readable storage medium can include both the internal storage unit and the external storage device of the electronic device. The computer readable storage medium is used to store the computer program and other programs and data required by the electronic device. The computer readable storage medium can also be used to temporarily store data that has been output or will be output.

[0149] Those skilled in the art can appreciate that the modules / units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. In order to clearly illustrate the interchangeability of hardware and software, the components and steps of the examples have been described in general terms in the above description. Whether the functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. A person skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0150] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the electronic device and the units described above can refer to the corresponding processes in the above-mentioned method embodiments, which will not be described here.

[0151] In several embodiments provided in the present application, it should be understood that the disclosed electronic device and method can be implemented in other manners. For example, the division of the system embodiments described above is merely an example, and the division of the modules / units can be different, for example, a plurality of modules / units can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed coupling or direct coupling or communication connection between the modules / units can be indirect coupling or communication connection through some interfaces or modules / units, and can be electrical, mechanical or other forms of connection.

[0152] The modules / units described as separate components may or may not be physically separate, and the components shown as modules / units may or may not be physical modules / units, i.e., they can be located in one place or distributed on multiple network modules / units. Part or all of the modules / units can be selected according to actual needs to achieve the purpose of the embodiments of the present application.

[0153] In addition, the functional modules / units in each embodiment of the present application can be integrated into a processing module / unit, or each module / unit can exist physically, or two or more modules / units can be integrated into one module / unit. The integrated module / unit can be implemented in the form of hardware or in the form of a software functional module / unit.

[0154] The above is merely a specific implementation of the present application, but the protection scope of the present application is not limited thereto, and any skilled person in the art can easily think of various equivalent modifications or replacements within the technical scope disclosed in the present application, and these modifications or replacements should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for detecting defects in hydrogen transportation pipelines, characterized in that, include: Acquire phased array ultrasonic testing data, digital X-ray imaging data, and hydrogen concentration data for hydrogen transportation pipelines; Defect boundary features are extracted based on the phased array ultrasonic testing data, defect microstructure features are extracted based on the digital X-ray imaging data, and spatiotemporal distribution features of hydrogen concentration are extracted based on the hydrogen concentration data. A modal correlation graph is constructed using the defect boundary features, the defect microstructure features, and the spatiotemporal distribution features of hydrogen concentration as nodes. The edge weight between each pair of nodes in the modal correlation graph is determined by the mutual information of the two nodes. A multimodal correlation feature vector is generated based on the modal correlation graph. The defect detection results of the hydrogen transport pipeline are obtained based on the multimodal correlation feature vector; the defect detection results include defect type and defect location; The hydrogen concentration data includes hydrogen concentration data at multiple pipeline locations within the hydrogen transport pipeline during the target time period; The extraction of spatiotemporal distribution features of hydrogen concentration based on the hydrogen concentration data includes: For each pipeline location in the hydrogen transport pipeline, extract the time series features of hydrogen concentration in the target time period based on the hydrogen concentration data of that pipeline location in that pipeline location; A spatial distribution sequence of hydrogen concentration is generated based on the hydrogen concentration time-series features of all pipeline locations in the hydrogen transportation pipeline and the corresponding target time periods for each pipeline location; the spatial distribution sequence of hydrogen concentration includes hydrogen concentration time-series features arranged in order of spatial location relationship, and each hydrogen concentration time-series feature corresponds to a location identifier; Based on the spatial distribution sequence of hydrogen concentration, the spatiotemporal distribution features of hydrogen concentration are extracted; The extraction of hydrogen concentration time-series features within the target time period based on hydrogen concentration data at the pipeline location includes: A first time window is determined based on the range of hydrogen concentration data changes within the target time period at the pipeline location; the range of data changes is determined by the maximum and minimum hydrogen concentration values ​​in the hydrogen concentration data, and the range of data changes is negatively correlated with the first time window; Calculate the mean and standard deviation sequences of hydrogen concentration data at the pipeline location within the target time period according to the first time window; Based on the mean sequence, a fitting curve for the hydrogen concentration at the pipeline location within the target time period is generated; based on the standard deviation sequence, a fitting curve for the rate of change of hydrogen concentration at the pipeline location within the target time period is generated. Based on the hydrogen concentration fitting curve and the hydrogen concentration change rate fitting curve, the temporal characteristics of hydrogen concentration within the target time period at the pipeline location are determined. The calculation of the mean and standard deviation sequences of hydrogen concentration data at the pipeline location within the target time period according to the first time window includes: dividing the target time period into multiple time blocks according to the first time window; calculating the mean of hydrogen concentration data corresponding to each time block, and concatenating the mean of hydrogen concentration data corresponding to each time block in chronological order to form a mean sequence of hydrogen concentration data; calculating the standard deviation of hydrogen concentration data corresponding to each time block, and concatenating the standard deviation of hydrogen concentration data corresponding to each time block in chronological order to form a standard deviation sequence of hydrogen concentration data.

2. The method for detecting defects in hydrogen pipelines as described in claim 1, characterized in that, The phased array ultrasonic detection data includes the original echo signal matrix obtained by scanning the hydrogen transport pipeline circumferentially with an ultrasonic probe. The extraction of defect boundary features based on the phased array ultrasonic testing data includes: The original echo signal matrix is ​​denoised to obtain the denoised echo signal matrix. A binarized segmentation map is generated based on the echo signal matrix, and closed defect boundaries are extracted based on the binarized segmentation map. The closed defect boundaries are used to characterize the spatial contour of defects in the hydrogen pipeline to reflect the spatial distribution and morphological properties of the defects. Based on the closed defect boundary, calculate the defect geometric features, defect topological features, and defect texture features, and then stitch the defect geometric features, defect topological features, and defect texture features together to obtain the defect boundary features.

3. The method for detecting defects in hydrogen pipelines as described in claim 1, characterized in that, The digital X-ray imaging data includes grayscale images of the cross-section of the hydrogen transport pipeline; The extraction of defect microstructure features based on the digital X-ray imaging data includes: The grayscale image is divided into multiple pixel groups, and grayscale gradient data corresponding to each pixel group is extracted. The grain distribution characteristics of the hydrogen transport pipeline are extracted based on the gray-level gradient data of the multiple pixel groups. Identify high gray-level abnormal discrete regions in the gray-level image whose gray-level values ​​exceed a first gray-level threshold, and determine defect density features based on the high gray-level abnormal discrete regions; The grain distribution characteristics and the defect density characteristics are used as the defect microstructure characteristics.

4. The method for detecting defects in hydrogen pipelines as described in claim 1, characterized in that, The extraction of spatiotemporal distribution features of hydrogen concentration based on the spatial distribution sequence of hydrogen concentration includes: The difference between the hydrogen concentration temporal features corresponding to each two adjacent positions in the spatial distribution sequence of hydrogen concentration is calculated to obtain the spatial hydrogen concentration difference sequence. A spatial hydrogen concentration gradient sequence is generated based on the spatial hydrogen concentration difference sequence; The hydrogen concentration spatial distribution sequence is divided into multiple time-segment hydrogen concentration spatial distribution sub-sequences according to the second time window; the second time window is larger than the first time window; Calculate the rate of change of hydrogen concentration at the same position in the spatial distribution subsequence of hydrogen concentration for every two adjacent time periods; The spatial propagation features of the hydrogen concentration change rate are extracted based on the hydrogen concentration change rate identified at the same position in the spatial distribution subsequence of hydrogen concentration across all time periods; the spatial propagation features include spatial propagation direction and spatial propagation intensity. The spatial hydrogen concentration gradient sequence and the spatial propagation characteristics are used as the spatiotemporal distribution characteristics of hydrogen concentration.

5. The method for detecting defects in hydrogen pipelines as described in claim 1, characterized in that, The defect detection results of the hydrogen pipeline are generated based on the phased array ultrasonic detection data, digital X-ray imaging data and hydrogen concentration data of the hydrogen pipeline, and through a target defect identification model. The target defect identification model is trained based on the target training sample set; The method for obtaining the target training sample set includes: Acquire phased array ultrasonic testing data, digital X-ray imaging data, and hydrogen concentration data of a hydrogen transport pipeline with pre-existing defects; use the phased array ultrasonic testing data, the digital X-ray imaging data, the hydrogen concentration data, and the defect labels of the hydrogen transport pipeline with pre-existing defects as a first training sample set; the defect labels include defect type and defect location; For each training sample in the first training sample set, at least one of the phased array ultrasonic testing data, the digital X-ray imaging data, and the hydrogen concentration data in the training sample is augmented based on the defect label of the training sample to obtain an augmented second training sample set. The first training sample set and the second training sample set are used as the target training sample set.

6. A defect detection system for hydrogen transportation pipelines, characterized in that, include: The data acquisition module is used to acquire phased array ultrasonic testing data, digital X-ray imaging data, and hydrogen concentration data for hydrogen transportation pipelines. The feature extraction module is used to extract defect boundary features based on the phased array ultrasonic detection data, extract defect microstructure features based on the digital X-ray imaging data, and extract spatiotemporal distribution features of hydrogen concentration based on the hydrogen concentration data; the hydrogen concentration data includes hydrogen concentration data at multiple pipeline locations in the hydrogen transport pipeline within the target time period; The feature extraction module is specifically used to extract the time-series features of hydrogen concentration at each pipeline location within a target time period based on the hydrogen concentration data at that pipeline location within that target time period. A spatial distribution sequence of hydrogen concentration is generated based on the hydrogen concentration time-series features of all pipeline locations in the hydrogen transportation pipeline and the corresponding target time periods for each pipeline location; the spatial distribution sequence of hydrogen concentration includes hydrogen concentration time-series features arranged in order of spatial location relationship, and each hydrogen concentration time-series feature corresponds to a location identifier; Based on the spatial distribution sequence of hydrogen concentration, the spatiotemporal distribution features of hydrogen concentration are extracted; The feature extraction module is further used to determine a first time window based on the data change span of hydrogen concentration data within a target time period at the pipeline location; the data change span is determined by the maximum and minimum hydrogen concentration values ​​in the hydrogen concentration data, and the data change span is negatively correlated with the first time window; Calculate the mean and standard deviation sequences of hydrogen concentration data at the pipeline location within the target time period according to the first time window; Based on the mean sequence, a fitting curve for the hydrogen concentration at the pipeline location within the target time period is generated; based on the standard deviation sequence, a fitting curve for the rate of change of hydrogen concentration at the pipeline location within the target time period is generated. Based on the hydrogen concentration fitting curve and the hydrogen concentration change rate fitting curve, the temporal characteristics of hydrogen concentration within the target time period at the pipeline location are determined. The feature extraction module is further used to divide the target time period into multiple time blocks according to the first time window; calculate the mean of the hydrogen concentration data corresponding to each time block, and concatenate the mean of the hydrogen concentration data corresponding to each time block in chronological order to form a mean sequence of hydrogen concentration data; calculate the standard deviation of the hydrogen concentration data corresponding to each time block, and concatenate the standard deviation of the hydrogen concentration data corresponding to each time block in chronological order to form a standard deviation sequence of hydrogen concentration data. The feature fusion module is used to construct a modal correlation graph by using the defect boundary features, the defect microstructure features, and the spatiotemporal distribution features of hydrogen concentration as nodes. The edge weight between each pair of nodes in the modal correlation graph is determined by the mutual information of the two nodes. A multimodal correlation feature vector is generated based on the modal correlation graph. The defect detection module is used to obtain the defect detection results of the hydrogen transport pipeline based on the multimodal correlation feature vector; the defect detection results include the defect type and defect location.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1 to 5.

8. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 5.

Images