A method, device and storage medium for detecting defects of a cable

Abstract

This invention discloses a method, device, and storage medium for cable defect detection. The method includes: acquiring time-domain reflection signals, frequency-domain reflection signals, and original discharge spectra of the cable; constructing a joint discharge spectrum based on the time-domain reflection signals, frequency-domain reflection signals, and original discharge spectra; extracting time-domain and frequency-domain feature information from the joint discharge spectrum; mapping the joint discharge spectrum to a bispectral space to obtain bispectral features; calculating the correlation between the feature information and the bispectral features; if the correlation is less than or equal to a preset threshold, fusing the bispectral features and feature information into the cable's fingerprint data; and identifying the cable's defect type based on the fingerprint data. This embodiment constructs a joint discharge spectrum at the first level based on the cable's information in the time domain, frequency domain, and spectrum, and extracts features from time-frequency and high-dimensional data at the second level. This effectively increases the amount of information in the features, improves the accuracy of the fingerprint data, and thus improves the accuracy of defect type identification.

Description

Technical Field

[0001] This invention relates to the field of power technology, and in particular to a method, equipment and storage medium for detecting defects in cables. Background Technology

[0002] Due to limitations in cable manufacturing processes and cable laying conditions, cables may have defects and there is a certain probability of malfunction.

[0003] Currently, the main method for detecting cable defects is to collect the electrical signals of the cable, extract spectral features from the electrical signals, construct fingerprint data, and then detect the type of defect based on the fingerprint data.

[0004] However, the actual defects in cables are quite complex, and the amount of information in their spectral characteristics is relatively small, resulting in low accuracy in identifying cable defects. Summary of the Invention

[0005] In view of this, the present invention provides a method, device and storage medium for detecting cable defects, so as to improve the accuracy of detecting cable defects.

[0006] A first aspect of the present invention provides a method for detecting defects in a cable, comprising:

[0007] Time-domain reflection signal, frequency-domain reflection signal, and raw discharge spectrum were collected from the cable.

[0008] A joint discharge spectrum is constructed based on the time-domain reflection signal, the frequency-domain reflection signal, and the original discharge spectrum.

[0009] Extract time-domain and frequency-domain feature information from the joint discharge spectrum;

[0010] The combined discharge spectrum is mapped to a bispectral space to obtain bispectral features;

[0011] Calculate the correlation between the feature information and the bispectral features;

[0012] If the correlation is less than or equal to a preset threshold, the bispectral features and the feature information are fused together to form the fingerprint data of the cable.

[0013] The defect type of the cable is identified based on the fingerprint data.

[0014] A second aspect of the present invention provides a cable defect detection device, comprising:

[0015] The cable information acquisition module is used to acquire time-domain reflection signals, frequency-domain reflection signals, and raw discharge spectra of cables.

[0016] A joint discharge spectrum construction module is used to construct a joint discharge spectrum based on the time-domain reflection signal, the frequency-domain reflection signal, and the original discharge spectrum.

[0017] The feature information extraction module is used to extract time-domain and frequency-domain feature information from the joint discharge spectrum;

[0018] The bispectral feature mapping module is used to map the joint discharge spectrum to a bispectral space to obtain bispectral features;

[0019] A correlation calculation module is used to calculate the correlation between the feature information and the bispectral features;

[0020] The fingerprint data generation module is used to fuse the bispectral features and the feature information into fingerprint data of the cable if the correlation is less than or equal to a preset threshold.

[0021] The defect type identification module is used to identify the defect type of the cable based on the fingerprint data.

[0022] A third aspect of the present invention provides an electronic device, the electronic device comprising:

[0023] At least one processor; and

[0024] A memory communicatively connected to the at least one processor; wherein,

[0025] The memory stores a computer program that can be executed by the at least one processor to enable the at least one processor to perform the cable defect detection method as described in the first aspect above.

[0026] A fourth aspect of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the cable defect detection method as described in the first aspect above.

[0027] A fifth aspect of the present invention provides a computer program product comprising a computer program that, when executed by a processor, implements the cable defect detection method as described in the first aspect above.

[0028] In this embodiment, time-domain reflection signals, frequency-domain reflection signals, and the original discharge spectrum of the cable are collected. A joint discharge spectrum is constructed based on the time-domain reflection signals, frequency-domain reflection signals, and the original discharge spectrum. Feature information in the time and frequency domains is extracted from the joint discharge spectrum. The joint discharge spectrum is mapped to a bispectral space to obtain bispectral features. The correlation between the feature information and the bispectral features is calculated. If the correlation is less than or equal to a preset threshold, the bispectral features and feature information are fused into the cable's fingerprint data. The defect type of the cable is identified based on the fingerprint data. This embodiment constructs a joint discharge spectrum at the first level based on the cable's information in the time domain, frequency domain, and spectrum. At the second level, features are extracted from time and frequency and high dimensions. Constructing features at different levels and dimensions can effectively increase the amount of information in the features, improve the accuracy of the fingerprint data, and thus improve the accuracy of defect type identification.

[0029] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0030] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0031] Figure 1 This is a flowchart of a cable defect detection method provided in Embodiment 1 of the present invention.

[0032] Figure 2 This is a schematic diagram of the structure of a cable defect detection device provided in Embodiment 2 of the present invention.

[0033] Figure 3 This is a schematic diagram of the structure of an electronic device provided in Embodiment 3 of the present invention. Detailed Implementation

[0034] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0035] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be used interchangeably where appropriate so that the embodiments of the invention described herein can cover implementations in sequences other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0036] Example 1

[0037] See Figure 1 The diagram illustrates a flowchart of a cable defect detection method according to Embodiment 1 of the present invention. This method can be executed by a cable defect detection device, which can be implemented in hardware and / or software and can be configured in an electronic device. Figure 1 As shown, the method includes:

[0038] Step 101: Collect time-domain reflection signal, frequency-domain reflection signal, and original discharge spectrum of the cable.

[0039] When a cable fails, time-domain reflection signals, frequency-domain reflection signals, and original discharge spectra can be collected from the cable.

[0040] The time-domain reflectometry (TDR) signal is acquired using this technique. TDR works by observing that as a signal travels along a path, impedance changes occur, causing some signal to be reflected while the rest continues along the path. TDR calculates the impedance change by measuring the voltage amplitude of the reflected wave and the time from the reflection point to the signal output point, thus determining the location of the impedance change point in the transmission path. In this embodiment, a time-domain signal is emitted from the cable using a time-domain reflectometer, and the reflected signal is received to obtain the time-domain reflectometry signal.

[0041] Frequency domain reflection signals are signals acquired through Frequency Domain Reflectometry (FDR). FDR is a method for measuring and evaluating the reflection characteristics of fiber optic transmission systems. It assesses these characteristics by measuring the reflected signals within the system. Its principle is based on signal spectral analysis; by converting a time-domain signal to a frequency-domain signal, the intensity of the reflected signal at different frequencies is obtained. By analyzing the spectrum of the reflected signal, the reflection situation in the fiber optic transmission system can be understood, thereby determining its performance and quality. In this embodiment, a frequency-domain signal is emitted from the power cable using a frequency-domain reflectometer, and the reflected signal is received from the frequency-domain reflectometer to obtain the frequency-domain reflection signal.

[0042] The original discharge spectrum is usually obtained through partial discharge tests on the cable.

[0043] Step 102: Construct a joint discharge spectrum based on the time-domain reflection signal, the frequency-domain reflection signal, and the original discharge spectrum.

[0044] In this embodiment, the time-domain reflection signal, the frequency-domain reflection signal, and the original discharge spectrum can be fused into a joint discharge spectrum, thereby increasing the amount of information.

[0045] In one embodiment of the present invention, step 102 may include the following steps:

[0046] Step 1021: Use an interactive function to interact the time-domain reflection signal and the frequency-domain reflection signal to obtain an interactive time-domain signal and an interactive frequency-domain signal.

[0047] In this embodiment, an interaction function can be preset. The interaction function is used to exchange some information between two different signals. When detecting defects in a cable, the time-domain reflection signal and the frequency-domain reflection signal can be substituted into the interaction function so that the time-domain reflection signal and the frequency-domain reflection signal exchange some information with each other, thereby obtaining the interaction time-domain signal and the interaction frequency-domain signal.

[0048] For example, the interaction function is represented as:

[0049] y1 = W t G(x1,θ1)+(1-W t )x2

[0050] y2=(1-W t )x1+W t G -1 (x2,θ2)

[0051] W t =S t +P t

[0052] S t=Softmax(x 1，t )

[0053] P t =Softmax(x 2，t )

[0054] Where x1 is the time-domain reflected signal, x2 is the frequency-domain reflected signal, and x 1，t Let x be the time-domain reflected signal at time t. 2，t Let y1 be the frequency domain reflected signal at time t, y2 be the interactive time domain signal, and W be the interactive frequency domain signal. t S is the interaction kernel at time t. t P is the time-domain component of the interaction function. t Let G(x1,θ1) be the frequency domain component of the interaction function, representing the first Gaussian mixture model with x1 as input and θ1 as parameter. -1 (x2,θ2) represents the second Gaussian mixture model with x2 as input and θ2 as parameter, and Softmax is the normalized exponential function.

[0055] Furthermore, in, This represents the first Gaussian mixture function with x1 as input and θ1 as parameter. Let F(x1) represent the second Gaussian mixture function with x2 as input and θ2 as parameter, and let F(x2) represent the weights of the Gaussian distribution of x1 and F(x2) represent the weights of the Gaussian distribution of x2.

[0056] In addition, θ1 includes the mean and standard deviation of x1, and θ2 includes the mean and standard deviation of x2.

[0057] Step 1022: Use a fusion function to combine the interactive time-domain signal and the interactive frequency-domain signal to obtain a time-frequency fused signal.

[0058] In this embodiment, a fusion function can be preset to fuse two different signals into a new signal. When detecting defects in a cable, the interactive time-domain signal and the interactive frequency-domain signal can be substituted into the fusion function to generate a time-frequency fused signal.

[0059] For example, the fusion function is represented as:

[0060] Y = y1Wy2

[0061] W = Softmax(W t )

[0062] Where Y is the time-frequency fusion signal, y1 is the interactive time-domain signal, y2 is the interactive frequency-domain signal, and W is the fusion function. t Let be the interaction kernel at time t, and Softmax be the normalized exponential function.

[0063] Step 1023: Adjust the original discharge spectrum based on the time-frequency fusion signal to obtain the joint discharge spectrum.

[0064] In this embodiment, the original discharge spectrum can be adjusted using the time-frequency fusion signal as a reference to obtain the joint discharge spectrum.

[0065] In the specific implementation, the parameters θ1 of the first Gaussian mixture model and θ2 of the second Gaussian mixture model are substituted into the fusion function for calculation, and the parameters θ1 of the first Gaussian mixture model and θ2 of the second Gaussian mixture model are fused to form the parameters θ3 of the third Gaussian mixture model, that is, θ3 = θ1Wθ2.

[0066] A Gaussian transform is applied to the time-frequency fusion signal Y using a third Gaussian mixture model that includes parameter θ3. The resulting time-frequency fusion signal R is expressed as follows:

[0067] If the Gaussian transform is completed, functions such as Softmax are used to normalize the time-frequency fusion signal.

[0068] If normalization is completed, the time-frequency fusion signal is used as the convolution kernel H (the convolution kernel H is a square matrix with R as the main diagonal, and the optional convolution kernel is a 3*3 square matrix). The convolution kernel H is used to perform a convolution operation on the original discharge spectrum to obtain the joint discharge spectrum.

[0069] In this embodiment, the correlation between time-domain and frequency-domain information in the reflected signal of cable faults was explored. Based on this, the discharge spectrum was adjusted, and the time-domain and frequency-domain features of the cable reflection signal were integrated into the discharge spectrum, effectively increasing the information content of the features and thus improving the accuracy of defect fingerprints. This lays the foundation for accurate identification of subsequent cable defects and improves the accuracy of power cable defect identification.

[0070] Step 103: Extract time-domain and frequency-domain feature information from the joint discharge spectrum.

[0071] In this embodiment, feature engineering can be performed on the joint discharge spectrum to extract feature information that simultaneously contains both time and frequency domains.

[0072] In one embodiment of the present invention, step 103 may include the following steps:

[0073] Step 1031: Perform a Fourier transform on the combined discharge spectrum to obtain the transformed discharge spectrum.

[0074] In this embodiment, the joint discharge spectrum can be subjected to a Fourier transform (FT) to convert it from the time domain to the frequency domain, thereby obtaining the transformed discharge spectrum.

[0075] The Fourier transform can include variations such as the FFT (Fast Fourier Transform).

[0076] Step 1032: Fuse the transformed discharge spectrum and the combined discharge spectrum to obtain a fused discharge image.

[0077] In this embodiment, the transformed discharge spectrum and the joint discharge spectrum can be fused to obtain a fused discharge image containing time-domain information and frequency-domain information.

[0078] In the specific implementation, a predetermined clustering algorithm is used to cluster the pixels in the transformed discharge spectrum to divide the transformed discharge spectrum into multiple segmented transformed blocks; each segmented transformed block has a first block centroid.

[0079] The pixels in the joint discharge spectrum are clustered using a predetermined clustering algorithm to divide the joint discharge spectrum into multiple segmented discharge blocks; each segmented discharge block has a second block centroid.

[0080] Traverse each segmentation transformation block and each segmentation discharge block, and identify the positional relationship between the segmentation transformation block and the segmentation discharge block based on the centroid of the first block and the centroid of the second block.

[0081] Generally, if the distance between the centroids of the first block and the second block is less than or equal to a certain threshold, the positional relationship between the segmented transformation block and the segmented discharge block can be considered to be overlapping. If the distance between the centroids of the first block and the second block is greater than a certain threshold, the positional relationship between the segmented transformation block and the segmented discharge block can be considered to be non-overlapping.

[0082] If the positional relationship is overlapping, then the overlapping area of ​​the segmented transformation block and the segmented discharge block is determined.

[0083] The system iterates through each pixel in the segmented transformation block and each pixel in the overlapping segmented discharge block within the region. If the distance between the pixel in the segmented transformation block and the pixel in the overlapping segmented discharge block is the smallest, then the system performs a weighted sum of the pixel in the segmented transformation block and the pixel in the overlapping segmented discharge block to obtain the fused discharge image.

[0084] In the weighted summation, the weight of the pixel in the segmented transformation block is the variance of the pixel in the segmented transformation block, and the weight of the pixel in the overlapping segmented discharge block is the variance of the pixel in the segmented discharge block.

[0085] Step 1033: Input the fused discharge image into the preset feature extraction model to output feature information.

[0086] In this embodiment, the fused discharge image can be input into a preset feature extraction model for computation, and the feature extraction model outputs feature information.

[0087] The feature extraction model can be a machine learning model or a deep learning model; this embodiment does not impose any restrictions on it.

[0088] Step 104: Map the combined discharge spectrum to the bispectral space to obtain bispectral features.

[0089] In this embodiment, kernel functions or other methods can be used to map the joint discharge spectrum to a higher-dimensional bispectral space (vector space) to obtain bispectral features.

[0090] In a higher-dimensional bispectral space, it is easier to distinguish different features in a linear manner in the new vector space, thereby improving the accuracy of cable defect identification.

[0091] Step 105: Calculate the correlation between feature information and bispectral features.

[0092] In this embodiment, both the feature information and the bispectral features originate from the same joint discharge spectrum. Since they are features extracted from the joint discharge spectrum from different angles, the correlation between the feature information and the bispectral features can be calculated to amplify the differences between different angles.

[0093] In one embodiment of the present invention, step 105 may include the following steps:

[0094] Step 1051: Map the bispectral features and feature information into the same vector space.

[0095] In this embodiment, fully connected layers or similar methods can be used to map bispectral features and feature information into the same vector space, so as to facilitate the comparison of bispectral features and feature information.

[0096] Step 1052: In the vector space, determine the first feature centroid of the bispectral feature and the second feature centroid of the feature information.

[0097] In the same vector space, the first feature centroid is determined in the bispectral features, and the second feature centroid is determined in the feature information.

[0098] Step 1053: Calculate the correlation between feature information and bispectral features using the first feature centroid and the second feature centroid.

[0099] In this embodiment, the relationship between the first feature centroid and the second feature centroid can be used to measure the correlation between feature information and bispectral features.

[0100] In the specific implementation, the distance (such as Euclidean distance) between the first feature centroid and the second feature centroid is calculated as the overall correlation quantity.

[0101] Algorithms such as K-means are used to cluster the bispectral features to obtain multiple bispectral clusters; each bispectral cluster has a third feature centroid.

[0102] The feature information is clustered using algorithms such as K-means to obtain multiple feature clusters; each feature cluster has a fourth feature centroid.

[0103] The spatial relationship between the bispectral cluster and the feature cluster is determined based on the third and fourth feature centroids.

[0104] Generally, when the distance between the third and fourth feature centroids (such as Euclidean distance) is less than or equal to a certain threshold, the spatial relationship between the bispectral cluster and the feature cluster can be determined to be overlapping; when the distance between the third and fourth feature centroids (such as Euclidean distance) is greater than a certain threshold, the spatial relationship between the bispectral cluster and the feature cluster can be determined to be non-overlapping.

[0105] If the spatial relationship is overlapping, then the distance between the third and fourth feature centroids (such as Euclidean distance) is set as the local correlation quantity.

[0106] Calculate the average value of all local correlation values, and use it as the average correlation value.

[0107] The correlation between the overall correlation and the average correlation is obtained by weighted summation.

[0108] In another embodiment of the present invention, step 105 may include the following steps:

[0109] Step 1054: Perform a convolution operation on the bispectral features to obtain the first convolutional feature.

[0110] In this embodiment, a preset convolution kernel can be used to perform a convolution operation on the linear information in the bispectral features to obtain the first convolutional feature.

[0111] Step 1055: Perform a convolution operation on the feature information to obtain the second convolution feature.

[0112] In this embodiment, a preset convolution kernel can be used to perform a convolution operation on the feature information to obtain a second convolution feature.

[0113] Step 1056: Calculate the cosine value between the first convolutional feature and the second convolutional feature, as the correlation between the feature information and the bispectral feature.

[0114] In this embodiment, the cosine value between the first convolutional feature and the second convolutional feature can be calculated, which represents the similarity between the feature information and the bispectral feature, and can be used as the correlation between the feature information and the bispectral feature.

[0115] Step 106: If the correlation is less than or equal to the preset threshold, the bispectral features and feature information are fused into the fingerprint data of the cable.

[0116] If the correlation between the feature information and the bispectral feature is less than or equal to the preset threshold, it indicates that the feature overlap between the feature information and the bispectral feature is low and the difference is large. In this case, the feature information can be adjusted according to the bispectral feature to obtain the fingerprint data of the cable.

[0117] In the specific implementation, if the correlation between the feature information and the bispectral feature is calculated first using the first feature centroid and the second feature centroid, then the cosine value between the first convolutional feature and the second convolutional feature can be calculated; if the cosine value between the first convolutional feature and the second convolutional feature has been calculated first, then the cosine value between the first convolutional feature and the second convolutional feature can be read directly.

[0118] If the phase information of the bispectral feature is less than the cosine value, the feature information is multiplied by the cosine value to obtain the fingerprint data of the cable.

[0119] In this embodiment, the fingerprint data of the cable integrates the characteristics of the cable in multiple dimensions during discharge, enabling the fingerprint data to accurately characterize the characteristics of cable discharge, improving the accuracy of the cable fingerprint data, and laying the foundation for subsequent cable fault analysis based on fingerprint data.

[0120] Step 107: Identify the type of cable defect based on fingerprint data.

[0121] In this embodiment, a fingerprint database can be pre-set, which records fingerprint data generated when the cable has various types of defects. At this time, the similarity between the current fingerprint data and the fingerprint data in the fingerprint database can be calculated. If the high similarity is greater than a certain threshold, the current fingerprint data can be considered to be the same as the fingerprint data in the fingerprint database. The defect type labeled on the fingerprint data in the fingerprint database is then extracted as the defect type of the current cable.

[0122] In this embodiment, time-domain reflection signals, frequency-domain reflection signals, and the original discharge spectrum of the cable are collected. A joint discharge spectrum is constructed based on the time-domain reflection signals, frequency-domain reflection signals, and the original discharge spectrum. Feature information in the time and frequency domains is extracted from the joint discharge spectrum. The joint discharge spectrum is mapped to a bispectral space to obtain bispectral features. The correlation between the feature information and the bispectral features is calculated. If the correlation is less than or equal to a preset threshold, the bispectral features and feature information are fused into the cable's fingerprint data. The defect type of the cable is identified based on the fingerprint data. This embodiment constructs a joint discharge spectrum at the first level based on the cable's information in the time domain, frequency domain, and spectrum. At the second level, features are extracted from time and frequency and high dimensions. Constructing features at different levels and dimensions can effectively increase the amount of information in the features, improve the accuracy of the fingerprint data, and thus improve the accuracy of defect type identification.

[0123] Example 2

[0124] See Figure 2 The diagram shows a structural schematic of a cable defect detection device provided in Embodiment 2 of the present invention. Figure 2 As shown, the device includes:

[0125] The cable information acquisition module 201 is used to acquire time-domain reflection signals, frequency-domain reflection signals, and raw discharge spectra of the cable.

[0126] The joint discharge spectrum construction module 202 is used to construct a joint discharge spectrum based on the time-domain reflection signal, the frequency-domain reflection signal, and the original discharge spectrum.

[0127] Feature information extraction module 203 is used to extract time-domain and frequency-domain feature information from the joint discharge spectrum;

[0128] The dual-spectrum feature mapping module 204 is used to map the joint discharge spectrum to a dual-spectrum space to obtain dual-spectrum features;

[0129] Correlation calculation module 205 is used to calculate the correlation between the feature information and the bispectral features;

[0130] The fingerprint data generation module 206 is used to fuse the bispectral features and the feature information into fingerprint data of the cable if the correlation is less than or equal to a preset threshold.

[0131] The defect type identification module 207 is used to identify the defect type of the cable based on the fingerprint data.

[0132] In one embodiment of the present invention, the joint discharge spectrum construction module 202 includes:

[0133] The signal interaction module is used to interact the time-domain reflected signal and the frequency-domain reflected signal using an interaction function to obtain an interactive time-domain signal and an interactive frequency-domain signal.

[0134] A time-frequency fusion signal combining module is used to combine the interactive time-domain signal and the interactive frequency-domain signal using a fusion function to obtain a time-frequency fusion signal;

[0135] The discharge spectrum adjustment module is used to adjust the original discharge spectrum according to the time-frequency fusion signal to obtain a joint discharge spectrum.

[0136] In one embodiment of the present invention, the interaction function is represented as:

[0137] y1 = W t G(x1,θ1)+(1-W t )x2

[0138] y2=(1-W t )x1+W t G -1 (x2,θ2)

[0139] W t =S t +P t

[0140] S t =Softmax(x 1，t )

[0141] P t =Softmax(x 2，t )

[0142] Where x1 is the time-domain reflection signal, x2 is the frequency-domain reflection signal, and x 1，t Let x be the time-domain reflected signal at time t. 2，t Let y1 be the frequency domain reflection signal at time t, y2 be the interactive time domain signal, and W be the interactive frequency domain signal. t S is the interaction kernel at time t. t P is the time-domain component of the interaction function. t Let G(x1,θ1) be the frequency domain component of the interaction function, and let G(x1,θ1) represent the first Gaussian mixture model with x1 as input and θ1 as parameter. -1 (x2,θ2) represents the second Gaussian mixture model with x2 as input and θ2 as parameter, and Softmax is the normalized exponential function;

[0143] The fusion function is expressed as follows:

[0144] Y = y1Wy2

[0145] W = Softmax(W t )

[0146] Wherein, Y is the time-frequency fusion signal, y1 is the interactive time-domain signal, y2 is the interactive frequency-domain signal, and W is the fusion function. t Let be the interaction kernel at time t, and Softmax be the normalized exponential function;

[0147] The discharge spectrum adjustment module includes:

[0148] The parameter fusion module is used to substitute the parameters of the first Gaussian mixture model and the parameters of the second Gaussian mixture model into the fusion function and fuse them into the parameters of the third Gaussian mixture model;

[0149] The Gaussian transform module is used to perform a Gaussian transform on the time-frequency fusion signal using the third Gaussian mixture model;

[0150] The normalization module is used to normalize the time-frequency fused signal if a Gaussian transform is performed.

[0151] The convolution module is used to perform a convolution operation on the original discharge spectrum using the time-frequency fusion signal as the convolution kernel after normalization is completed, so as to obtain a joint discharge spectrum.

[0152] In one embodiment of the present invention, the feature information extraction module 203 includes:

[0153] The spectrum transformation module is used to perform a Fourier transform on the joint discharge spectrum to obtain a transformed discharge spectrum.

[0154] The spectrum fusion module is used to fuse the transformed discharge spectrum and the joint discharge spectrum to obtain a fused discharge image;

[0155] The model processing module is used to input the fused discharge image into a preset feature extraction model to output feature information.

[0156] In one embodiment of the present invention, the spectral fusion module includes:

[0157] The segmentation and transformation block division module is used to cluster the pixels in the transformation discharge spectrum to divide the transformation discharge spectrum into multiple segmentation and transformation blocks; each segmentation and transformation block has a first block centroid;

[0158] The segmented discharge block segmentation module is used to cluster the pixels in the joint discharge spectrum to segment the joint discharge spectrum into multiple segmented discharge blocks; each segmented discharge block has a second block centroid.

[0159] A positional relationship identification module is used to identify the positional relationship between the segmented transformation block and the segmented discharge block based on the centroid of the first block and the centroid of the second block;

[0160] The overlapping region determination module is used to determine the overlapping region between the segmented transformation block and the segmented discharge block if the positional relationship is overlapping;

[0161] A fused discharge image generation module is used to, in the region, if the distance between the pixels in the segmented transformation block and the pixels overlapping the segmented discharge block is the smallest, then to perform a weighted summation of the pixels in the segmented transformation block and the pixels overlapping the segmented discharge block to obtain a fused discharge image.

[0162] In the weighted summation, the weight of the pixel in the segmented transformation block is the variance of the pixel in the segmented transformation block, and the weight of the pixel in the overlapping segmented discharge block is the variance of the pixel in the segmented discharge block.

[0163] In one embodiment of the present invention, the correlation calculation module 205 includes:

[0164] A spatial mapping module is used to map the bispectral features and the feature information to the same vector space;

[0165] The feature centroid determination module is used to determine the first feature centroid of the bispectral feature and the second feature centroid of the feature information in the vector space.

[0166] The feature centroid calculation module is used to calculate the correlation between the feature information and the bispectral feature using the first feature centroid and the second feature centroid.

[0167] In one embodiment of the present invention, the feature centroid calculation module includes:

[0168] The overall correlation calculation module is used to calculate the distance between the first feature centroid and the second feature centroid as the overall correlation quantity.

[0169] A bispectral clustering module is used to cluster the bispectral features to obtain multiple bispectral clusters; each bispectral cluster has a third feature centroid;

[0170] A feature clustering module is used to cluster the feature information to obtain multiple feature clusters; each feature cluster has a fourth feature centroid.

[0171] A spatial relationship determination module is used to determine the spatial relationship between the bispectral cluster and the feature cluster based on the third feature centroid and the fourth feature centroid.

[0172] The local correlation quantity setting module is used to set the distance between the third feature centroid and the fourth feature centroid as a local correlation quantity if the spatial relationship is overlapping.

[0173] The average correlation calculation module is used to calculate the average value of all the local correlation values ​​as the average correlation value.

[0174] The correlation quantity fusion module is used to weight and sum the overall correlation quantity and the average correlation quantity to obtain the correlation between the feature information and the bispectral features.

[0175] In another embodiment of the present invention, the correlation calculation module 205 includes:

[0176] The first convolution operation execution module is used to perform a convolution operation on the bispectral features to obtain the first convolution feature;

[0177] The second convolution operation execution module is used to perform convolution operations on the feature information to obtain the second convolution feature;

[0178] The cosine value calculation module is used to calculate the cosine value between the first convolutional feature and the second convolutional feature, as the correlation between the feature information and the bispectral feature;

[0179] The fingerprint data generation module 206 includes:

[0180] The feature adjustment module is used to multiply the feature information by the cosine value if the phase information of the bispectral feature is less than the cosine value, so as to obtain the fingerprint data of the cable.

[0181] The cable defect detection device provided in this embodiment of the invention can execute the cable defect detection method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects for executing the cable defect detection method.

[0182] Example 3

[0183] See Figure 3 This diagram illustrates a structural schematic of an electronic device according to an embodiment of the present invention. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0184] like Figure 3 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 may also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0185] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0186] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as cable defect detection methods.

[0187] In some embodiments, the cable defect detection method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or mounted on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the cable defect detection method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the cable defect detection method by any other suitable means (e.g., by means of firmware).

[0188] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0189] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0190] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0191] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0192] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0193] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0194] Example 4

[0195] This invention also provides a computer program product, which includes a computer program that, when executed by a processor, implements the cable defect detection method provided in any embodiment of this invention.

[0196] In implementing the computer program product, computer program code for performing the operations of this invention can be written in one or more programming languages ​​or a combination thereof. Programming languages ​​include object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0197] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0198] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for detecting defects in cables, characterized in that, include: Time-domain reflection signal, frequency-domain reflection signal, and raw discharge spectrum were collected from the cable. A joint discharge spectrum is constructed based on the time-domain reflection signal, the frequency-domain reflection signal, and the original discharge spectrum. Extract time-domain and frequency-domain feature information from the joint discharge spectrum; The combined discharge spectrum is mapped to a bispectral space to obtain bispectral features; Calculate the correlation between the feature information and the bispectral features; If the correlation is less than or equal to a preset threshold, the bispectral features and the feature information are fused together to form the fingerprint data of the cable. The defect type of the cable is identified based on the fingerprint data; The step of constructing a joint discharge spectrum based on the time-domain reflection signal, the frequency-domain reflection signal, and the original discharge spectrum includes: The time-domain reflected signal and the frequency-domain reflected signal are interacted using an interactive function to obtain an interactive time-domain signal and an interactive frequency-domain signal. The time-domain signal and the frequency-domain signal are combined using a fusion function to obtain a time-frequency fused signal; The original discharge spectrum is adjusted based on the time-frequency fusion signal to obtain a joint discharge spectrum; The interaction function is expressed as follows: y1=W t G(x1,θ1)+(1-W t )x2 y2=(1-W t )x1+W t G -1 (x2,θ2) W t =S t +P t S t =Softmax(x 1，t ) P t =Softmax(x 2，t ) Where x1 is the time-domain reflection signal, x2 is the frequency-domain reflection signal, and x 1，t Let x be the time-domain reflected signal at time t. 2，t Let y1 be the frequency domain reflection signal at time t, y2 be the interactive time domain signal, and W be the interactive frequency domain signal. t S is the interaction kernel at time t. t P is the time-domain component of the interaction function. t Let G(x1,θ1) be the frequency domain component of the interaction function, and let G(x1,θ1) represent the first Gaussian mixture model with x1 as input and θ1 as parameter. -1 (x2,θ2) represents the second Gaussian mixture model with x2 as input and θ2 as parameter, where Softmax is the normalized exponential function; The fusion function is expressed as follows: Y = y1Wy2 W=Softmax(W t ) Where Y is the time-frequency fusion signal, y1 is the interactive time-domain signal, y2 is the interactive frequency-domain signal, and W is the fusion function. t Let be the interaction kernel at time t, and Softmax be the normalized exponential function; The step of adjusting the original discharge spectrum based on the time-frequency fusion signal to obtain the joint discharge spectrum includes: The parameters of the first Gaussian mixture model and the parameters of the second Gaussian mixture model are substituted into the fusion function and fused to obtain the parameters of the third Gaussian mixture model; The time-frequency fusion signal is subjected to Gaussian transformation using the third Gaussian mixture model. If the Gaussian transform is completed, the time-frequency fused signal is then normalized. If normalization is completed, the original discharge spectrum is convolved using the time-frequency fusion signal as the convolution kernel to obtain the joint discharge spectrum.

2. The method according to claim 1, characterized in that, Extracting time-domain and frequency-domain feature information from the joint discharge spectrum includes: Perform a Fourier transform on the combined discharge spectrum to obtain the transformed discharge spectrum; The transformed discharge spectrum and the combined discharge spectrum are fused to obtain a fused discharge image; The fused discharge image is input into a preset feature extraction model to output feature information.

3. The method according to claim 2, characterized in that, The process of fusing the transformed discharge spectrum and the combined discharge spectrum to obtain a fused discharge image includes: The pixels in the transformed discharge spectrum are clustered to divide the transformed discharge spectrum into multiple segmented transformed blocks; each segmented transformed block has a first block centroid. The pixels in the joint discharge spectrum are clustered to divide the joint discharge spectrum into multiple segmented discharge blocks; each segmented discharge block has a second block centroid. The positional relationship between the segmented transformation block and the segmented discharge block is identified based on the centroid of the first block and the centroid of the second block; If the positional relationship is overlapping, then the overlapping area of ​​the segmented transformation block and the segmented discharge block is determined; In the region, if the distance between the pixel in the segmented transformation block and the pixel in the overlapping segmented discharge block is the smallest, then the pixel in the segmented transformation block and the pixel in the overlapping segmented discharge block are weighted and summed to obtain the fused discharge image; In the weighted summation, the weight of the pixel in the segmented transformation block is the variance of the pixel in the segmented transformation block, and the weight of the pixel in the overlapping segmented discharge block is the variance of the pixel in the segmented discharge block.

4. The method according to any one of claims 1-3, characterized in that, The calculation of the correlation between the feature information and the bispectral features includes: The bispectral features and the feature information are mapped to the same vector space; In the vector space, determine the first feature centroid of the bispectral feature and the second feature centroid of the feature information; The correlation between the feature information and the bispectral feature is calculated using the first feature centroid and the second feature centroid.

5. The method according to claim 4, characterized in that, The step of calculating the correlation between the feature information and the bispectral feature using the first feature centroid and the second feature centroid includes: Calculate the distance between the first feature centroid and the second feature centroid, and use it as the overall correlation quantity; Clustering the bispectral features yields multiple bispectral clusters; each bispectral cluster has a third feature centroid. The feature information is clustered to obtain multiple feature clusters; each feature cluster has a fourth feature centroid. The spatial relationship between the bispectral cluster and the feature cluster is determined based on the third feature centroid and the fourth feature centroid; If the spatial relationship is overlapping, then the distance between the third feature centroid and the fourth feature centroid is set as a local correlation quantity; Calculate the average value of all the local correlation values, and use it as the average correlation value; The correlation between the overall correlation value and the average correlation value is obtained by weighted summation.

6. The method according to any one of claims 1-3, characterized in that, The calculation of the correlation between the feature information and the bispectral features includes: Perform a convolution operation on the bispectral features to obtain the first convolutional feature; Perform a convolution operation on the feature information to obtain the second convolutional feature; Calculate the cosine value between the first convolutional feature and the second convolutional feature, and use it as the correlation between the feature information and the bispectral feature; The process of fusing the bispectral features with the feature information to form the fingerprint data of the cable includes: If the phase information of the bispectral feature is less than the cosine value, the feature information is multiplied by the cosine value to obtain the fingerprint data of the cable.

7. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the cable defect detection method as described in any one of claims 1-6.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the cable defect detection method as described in any one of claims 1-6.

Images