Cable quality detection and supervision system

Abstract

The invention relates to the technical field of cable quality detection, and discloses a cable quality detection supervision system. The invention aims to solve the problem of large eccentricity calculation deviation caused by incapability of accurately extracting the center of a conductor with an irregular cross section formed by twisting a plurality of wires in the prior art. According to the system, a three-dimensional coordinate system is established through a laser mark positioning module, a high-precision cross-section image is obtained through an ultrasonic discharge imaging module, an intelligent image processing module extracts multiple feature points to calculate an initial center, and an iterative registration module constructs an axial drift trajectory. The center decision module determines a real center and visual eccentricity distribution of the quality supervision platform in combination with a probability model and discharge verification, the center position of the irregular section conductor is accurately positioned through multi-module cooperation, the eccentricity detection precision is effectively improved, and cable quality full-process supervision is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of cable quality detection, and in particular to a cable quality detection and supervision system. Background Art

[0002] In cable construction, conductors formed by arranging or twisting multiple wires are a common design. Their cross-sections present irregular shapes due to the gaps between the wires and their arrangement. The eccentricity of such conductors refers to the degree of spatial deviation between the center of the conductor and the center of the outer insulation layer, significantly impacting electrical performance. Eccentricity can lead to uneven thickness distribution of the insulation layer in different directions, resulting in abnormally high electric field strength at the thinnest point, which can easily cause partial discharge and insulation material aging. Over long-term operation, this can reduce the cable's voltage withstand capability and increase the risk of insulation breakdown and circuit failure.

[0003] Existing cable eccentricity detection technology has the following technical flaws: First, traditional methods are mostly designed for single circular conductors and rely on fitting the geometric center of the circular contour. This method cannot adapt to irregular cross-section conductors formed by arranging or twisting multiple wires. Second, conventional detection methods lack a verification mechanism for the relationship between the conductor's discharge position and the geometric center, resulting in insufficient reliability of the measurement results. Because such irregular cross-section conductors are widely used in medium and low voltage cables, existing technologies are unable to accurately extract the center of the overall conductor formed by multiple wires, resulting in large deviations in eccentricity calculations and making it difficult to meet the quality control needs in actual production. In response to the above problems, existing technologies are in urgent need of improvement. Summary of the Invention

[0004] The technical problem to be solved by the present invention is: irregular cross-section conductors are widely used in medium and low voltage cables in the existing technology. The existing technology cannot accurately extract the center of the overall conductor formed by multiple wires, resulting in a large deviation in eccentricity calculation. For this reason, we propose a cable quality detection and supervision system.

[0005] In order to achieve the above-mentioned purpose, the present application adopts the following technical solutions: a cable quality inspection and supervision system, comprising: a laser marking positioning module: used to establish a spatial coordinate system on the cable surface, generate a spatial positioning mark on the surface of the insulation layer through a laser engraving device, and use a machine vision system to identify the mark position to establish a three-dimensional coordinate system; an ultrasonic discharge imaging module: used to obtain a cross-sectional image at the conductor discharge position, stimulate ultrasonic induced cavitation discharge through a multi-channel piezoelectric array, and capture the conductor cross-sectional image in conjunction with an optical coherence tomography; an intelligent image processing module: used to identify the structural features of the conductor, use an edge detection algorithm to extract the conductor contour and the insulation layer contour, delineate the candidate area based on multiple feature points on the contour, and calculate the initial center position; an iterative alignment module: used to establish a multi-section spatial correlation model, realize the alignment of different cross-sectional images through a feature point matching algorithm, and construct an axial drift trajectory of the center; a center decision module: used to determine the true conductor center position, determine the true conductor center based on the probability distribution model of the multi-section center coordinates, and correct the coordinates in combination with the discharge position verification results; a quality supervision platform: used to visualize the quality assessment results, generate an eccentricity distribution map through a three-dimensional modeling engine, and store the detection data.

[0006] Preferably, the laser wavelength of the laser engraving device is 1064nm±50nm, the marking point diameter is ≤0.1mm, and the distance between adjacent marks is 50cm±10cm; the machine vision system uses a CMOS sensor with a resolution of ≥20 million pixels, and the mark recognition error is ≤0.05mm.

[0007] Preferably, the multi-channel piezoelectric array is a 40-channel PZT-8 piezoelectric ceramic array, and the ultrasonic excitation frequency range is 0.8-1.2 MHz; the optical coherence tomography device is polarization sensitive, with an axial resolution of ≤5 μm and a lateral scanning range of ≥10 mm.

[0008] Preferably, the steps of calculating the initial center coordinates by the intelligent image processing module include: S1. Feature point extraction: screening multiple key feature points from the conductor contour, including but not limited to: curvature extreme points, the vertex of the conductor, the maximum or minimum curvature point of the arc contour; discharge association points, the contour points closest to the discharge position detected by the ultrasonic discharge imaging module; symmetrical feature points, the contour has a local symmetrical structure, and the intersection of the symmetry axis is extracted; the number of feature points is not less than 8, and they are evenly distributed around the contour, with a circumferential spacing of ≤45°; S2. Candidate area delineation: based on the coordinate distribution of the feature points, the candidate area of ​​the initial center is delineated using the minimum circumscribed circle method, and the radius of the area is the maximum value of the distance from the feature point to the distribution center × 1.2, ensuring that all feature points are within the area; S3. Initial center calculation: within the candidate area, the weighted centroid method is used to calculate the initial center coordinates, and the formula is: ;in, is the initial center coordinate, is the coordinate of the i-th feature point extracted from the cross-sectional image, is the geometric weight, the greater the curvature, the higher the weight. is the discharge weight, the closer the distance to the discharge point, the higher the weight, and n is the total number of feature points.

[0009] Preferably, the feature point matching technology of the iterative registration module includes: extracting feature points of each cross-sectional image using the ORB algorithm, with the number being ≥50; eliminating mismatched points using the PROSAC algorithm, with a retention rate of ≥90%; and achieving image alignment based on the affine transformation formula: Where P is the coordinate of the feature point in the original cross-section image, M is the rotation matrix, which describes the angular deviation between the two cross-sections and is calculated by ORB feature point matching, and T is the translation vector, which represents the positional deviation between the two cross-sections. is the coordinate of the feature point after alignment.

[0010] Preferably, the probability distribution model is a Gaussian mixture model, which is expressed as: ;in is the probability density function of the center coordinates, indicating the possibility of the center appearing. is the weight of the kth Gaussian component , and are the mean and variance of the kth component respectively. The parameters are solved by the maximum expectation algorithm, and the number of components m is determined by the Bayesian information criterion.

[0011] Preferably, the discharge position verification includes: S71. measuring the distance from the discharge point to the current center ; S72. Calculate theoretical distance Where k is the insulation material constant. For example, the cross-linked polyethylene (XLPE) is 0.2 kV mm. To calculate the maximum electric field intensity at the discharge point through electric field simulation, is the laboratory calibration correction item to compensate for the measurement error; S73. , judged as an outlier and removed, is the measurement standard deviation.

[0012] Preferably, a vibration compensation unit is also included, which collects six-axis vibration data through a MEMS accelerometer and calculates the value of the vibration compensation unit based on the formula Calculate the compensation displacement, is the real-time acceleration, is the material elastic coefficient, which is used to correct the position offset during image acquisition.

[0013] Preferably, in the quality control platform, the three-dimensional modeling engine uses WebGL technology to generate an eccentricity heat map with a display accuracy of ≤0.01mm.

[0014] Preferably, the detection process includes: the laser marking positioning module initializes the spatial coordinate system; the ultrasonic discharge imaging module collects cross-sectional images at the marked point; the intelligent image processing module outputs the initial center coordinates of each section; the iterative alignment module constructs the center drift trajectory; the center decision module inputs the initial center coordinates of multiple sections and the discharge position to calculate the real conductor center coordinates; the quality control platform inputs the real conductor center and the insulation layer center, calculates the eccentricity, generates a detection report and uploads it.

[0015] The technical effects and advantages of the present invention are as follows: In the present invention, a three-dimensional coordinate system is established through a laser marking positioning module, a high-precision cross-sectional image is obtained through an ultrasonic discharge imaging module, and an intelligent image processing module is used to extract multiple feature points to calculate the initial center. The true center coordinates are determined by combining a probability model with a discharge verification mechanism, thereby effectively solving the problem of eccentricity detection of conductors with irregular cross-sections and improving the accuracy of detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The disclosure of the present invention is illustrated with reference to the accompanying drawings. It should be understood that the drawings are for illustrative purposes only and are not intended to limit the scope of protection of the present invention. In the accompanying drawings, the same reference numerals are used to refer to the same components: Figure 1 It is a module schematic diagram of the present invention. DETAILED DESCRIPTION

[0017] It is easy to understand that according to the technical solution of the present invention, without changing the essential spirit of the present invention, a person skilled in the art can propose a variety of interchangeable structural modes and implementation modes. Therefore, the following specific embodiments and drawings are only exemplary descriptions of the technical solution of the present invention and should not be regarded as the entire invention or as a limitation or restriction of the technical solution of the present invention.

[0018] Reference Figure 1As shown, the present invention provides a technical solution: a cable quality inspection and supervision system, characterized in that it includes: a laser marking positioning module: used to establish a spatial coordinate system on the cable surface, generate a spatial positioning mark on the surface of the insulation layer through a laser engraving device, and use a machine vision system to identify the mark position to establish a three-dimensional coordinate system; an ultrasonic discharge imaging module: used to obtain a cross-sectional image at the conductor discharge position, stimulate ultrasonic induced cavitation discharge through a multi-channel piezoelectric array, and capture the conductor cross-sectional image in conjunction with an optical coherence tomography; an intelligent image processing module: used to identify the structural features of the conductor, use an edge detection algorithm to extract the conductor contour and the insulation layer contour, delineate the candidate area based on multiple feature points on the contour, and calculate the initial center position; an iterative alignment module: used to establish a multi-section spatial correlation model, realize the alignment of different cross-sectional images through a feature point matching algorithm, and construct an axial drift trajectory of the center; a center decision module: used to determine the true conductor center position, determine the true conductor center based on a probability distribution model of the multi-section center coordinates, and correct the coordinates based on the discharge position verification result; a quality supervision platform: used to visualize the quality assessment results, generate an eccentricity distribution map through a three-dimensional modeling engine, and store the inspection data.

[0019] Specifically, after the system establishes a three-dimensional spatial reference through laser marking, it uses ultrasonic imaging to obtain the conductor cross-sectional structural data; the intelligent image processing module extracts the conductor contour feature points and calculates the initial center coordinates based on the geometric and discharge characteristics; the iterative alignment module spatially aligns the feature points of multiple sections to construct the axial drift trajectory of the conductor center; the center decision module integrates multi-section data to establish a probability model, and corrects the final center coordinates based on physical verification of the discharge position; the quality control platform converts the test results into a three-dimensional eccentricity distribution map, which intuitively displays the trend of changes in the insulation layer thickness.

[0020] The laser wavelength of the laser engraving device is 1064nm±50nm, the marking point diameter is ≤0.1mm, and the distance between adjacent marks is 50cm±10cm; the machine vision system uses a CMOS sensor with a resolution of ≥20 million pixels, and the mark recognition error is ≤0.05mm.

[0021] The multi-channel piezoelectric array is a 40-channel PZT-8 piezoelectric ceramic array with an ultrasonic excitation frequency range of 0.8-1.2 MHz; the optical coherence tomography equipment is polarization-sensitive with an axial resolution of ≤5 μm and a lateral scanning range of ≥10 mm.

[0022] Specifically, during the conductor cross-section detection process, the multi-channel piezoelectric array generates an ultrasonic field covering the conductor cross-section by time-sharingly exciting the piezoelectric ceramic units of different channels, inducing cavitation discharge at the interface between the conductor and the insulating layer; the polarization-sensitive optical coherence tomography equipment collects the polarization state change information of the reflected light by emitting a polarization-modulated light beam, and combines axial high resolution with a large lateral scanning range to clearly capture the arrangement of the wires inside the conductor and the microstructure of the discharge area; by synchronously controlling the excitation frequency and optical scanning timing of the piezoelectric array, multi-dimensional information fusion of the conductor cross-section image is achieved.

[0023] The present application further proposes that the steps for calculating the initial center coordinates by the intelligent image processing module include: S1. Feature point extraction: screening multiple key feature points from the conductor contour, including but not limited to: curvature extreme points, the vertex of the conductor, the maximum or minimum curvature point of the arc contour; discharge association points, the contour points closest to the discharge position detected by the ultrasonic discharge imaging module; symmetrical feature points, the contour has a local symmetrical structure, and the intersection of the symmetry axis is extracted; the number of feature points is not less than 8, and they are evenly distributed around the contour, with a circumferential spacing of ≤45°; S2. Candidate area delineation: based on the coordinate distribution of the feature points, the candidate area of ​​the initial center is delineated using the minimum circumscribed circle method, and the radius of the area is the maximum value of the distance from the feature point to the distribution center × 1.2, to ensure that all feature points are within the area; S3. Initial center calculation: within the candidate area, the weighted centroid method is used to calculate the initial center coordinates, and the formula is: ;in, is the initial center coordinate, is the coordinate of the i-th feature point extracted from the cross-sectional image, is the geometric weight, the greater the curvature, the higher the weight. It is the geometric weight, and the greater the curvature, the higher the weight. In the conductor contour analysis process, the curvature extreme points are first used to capture significant geometric features such as the vertices of the fan-shaped conductor. At the same time, the discharge correlation points are combined to reflect the actual influence of the discharge position of the conductor, and the symmetrical feature points are used to enhance the characterization ability of the regular structure. The circumferentially uniform distribution of feature points ensures that all orientation features of the conductor are effectively covered, avoiding deviations caused by excessive concentration of local features. In the candidate area delineation stage, the radius of the circumscribed circle is dynamically adjusted to form an effective search space covering all feature points. When calculating the initial center, the geometric weight is dynamically adjusted according to the curvature size, so that the contour mutation area obtains a higher weight, and the discharge weight is inversely distributed according to the distance to the discharge point, so that the feature points with strong physical correlation have greater influence, and finally the high-precision positioning of the conductor center is achieved through a composite weighting method.

[0024] This application further proposes a feature point matching technology for the iterative registration module, including: using the ORB algorithm to extract feature points of each cross-sectional image, with a number of ≥50; using the PROSAC algorithm to eliminate mismatched points, with a retention rate of ≥90%; and achieving image alignment based on the affine transformation formula: , where P is the coordinate of the feature point in the original cross-sectional image, M is the rotation matrix, which describes the angular deviation between the two cross-sections and is calculated by ORB feature point matching, and T is the translation vector, which represents the positional deviation between the two cross-sections. is the coordinate of the feature point after alignment.

[0025] Specifically, in the cable cross-section image registration process, the ORB algorithm is first used to extract no less than 50 feature points in each cross-section to form a basic matching data set. The PROSAC algorithm is then used to screen the initial matching points. By calculating the geometric consistency score of the matching point pairs, the matching points that meet the spatial constraints are retained first, and the rejection rate of mismatched points is controlled within 10%. Finally, based on the retained matching point pairs, the affine transformation matrix is ​​constructed and the translation vector is calculated. Through matrix operations, the different cross-section images are mapped to a unified coordinate system to form a continuous axial drift trajectory.

[0026] This application further proposes that the probability distribution model is a Gaussian mixture model, which is expressed as: ;in is the probability density function of the center coordinates, indicating the possibility of the center appearing. is the weight of the kth Gaussian component , and are the mean and variance of the kth component respectively. The parameters are solved by the maximum expectation algorithm, and the number of components m is determined by the Bayesian information criterion.

[0027] During the calculation of the cable conductor center, the center coordinates of multiple cross-sections may present multi-peak distribution characteristics due to the irregularity of the conductor structure. The Gaussian mixture model can effectively fit such complex distributions by superimposing multiple Gaussian components. First, the model parameters are iteratively optimized based on the maximum expectation algorithm, and the weight, mean, and variance of each Gaussian component are updated by calculating the posterior probability until convergence. Then, the model's advantages and disadvantages under different numbers of components are evaluated according to the Bayesian information criterion, and the optimal number of components is selected. Finally, the true conductor center position is determined by the maximum value or weighted mean of the probability density function, thereby solving the problem of accurate calculation of the center of irregular conductors.

[0028] Discharge position verification includes: S71. Measuring the distance from the discharge point to the current center ; S72. Calculate theoretical distance , where k is the insulation material constant. For example, the cross-linked polyethylene (XLPE) is 0.2 kV·mm. To calculate the maximum electric field intensity at the discharge point through electric field simulation, is the laboratory calibration correction item to compensate for the measurement error; S73. When , it is judged as an abnormal point and removed. is the measurement standard deviation.

[0029] During the verification of the conductor center coordinates, the actual measured position of the discharge point in the current cross-sectional image is first obtained, and its Euclidean distance to the current center coordinate is calculated. Subsequently, the theoretical discharge distance for that position is generated based on the dielectric constant of the insulating material, the local maximum electric field strength data, and correction parameters pre-calibrated in the laboratory. By comparing the difference between the actual measured distance and the theoretical distance, center coordinate offsets caused by image registration errors or conductor structural anomalies can be effectively identified. When the difference between the two exceeds three standard deviations, the center coordinate corresponding to the discharge point is marked as unreliable data, thereby preventing the abnormal point from interfering with the final conductor center positioning.

[0030] The present application further proposes a vibration compensation unit, which collects six-axis vibration data through a MEMS accelerometer and is based on the formula Calculate the compensation displacement, Real-time acceleration, is the material elastic coefficient, which is used to correct the position offset during image acquisition.

[0031] During the cable quality inspection process, the vibration compensation unit continuously collects the vibration signal of the equipment and analyzes the spatial displacement deviation caused by the vibration through the time integration operation of the acceleration data. The compensated displacement is corrected by the elastic coefficient and converted into the coordinate offset correction parameter of the image acquisition module, thereby eliminating the image position error caused by equipment jitter or external impact. For example, during laser marking positioning or ultrasonic imaging, if the detection equipment is disturbed by lateral vibration, the compensation algorithm can dynamically adjust the spatial coordinates of the optical components to ensure the accurate correspondence between the cross-sectional image and the three-dimensional coordinate system.

[0032] This application further proposes that in the quality control platform, the three-dimensional modeling engine uses WebGL technology to generate eccentricity heat maps with a display accuracy of less than or equal to 0.01 mm; the blockchain storage data contains five tuples: cable ID, cross-sectional coordinates, center parameters, timestamp, and device code, and is encrypted using the SHA-256 algorithm.

[0033] This application further proposes that the detection process includes: the laser marking positioning module initializes the spatial coordinate system; the ultrasonic discharge imaging module collects cross-sectional images at the marking point; the intelligent image processing module outputs the initial center coordinates of each section; the iterative alignment module constructs the center drift trajectory; the center decision module inputs the initial center coordinates of multiple sections and the discharge position to calculate the real conductor center coordinates; the quality control platform inputs the real conductor center and the insulation layer center, calculates the eccentricity, generates a detection report and uploads it.

[0034] Among them, the laser marking positioning module initializes the spatial coordinate system and establishes a three-dimensional reference on the cable surface through a laser engraving device. It is specifically implemented by combining spatial positioning mark generation with machine vision recognition to provide a spatial reference for subsequent cross-sectional image acquisition; the ultrasonic discharge imaging module collects cross-sectional images at the marking point and obtains the internal structure data of the conductor through ultrasonic induced cavitation discharge in conjunction with optical imaging equipment. It is specifically implemented by a multi-channel piezoelectric array to excite ultrasonic waves and optical coherence tomography technology to work together to accurately capture the relative position relationship between the conductor and the insulation layer; the intelligent image processing module outputs the initial center coordinates of each section and generates the conductor center position based on feature point extraction and weighted calculation. It is specifically implemented by combining edge detection algorithm with geometric weight and discharge weight centroid calculation method. Used to solve the problem of center positioning of irregular conductor contours; the iterative alignment module constructs the center drift trajectory and establishes an axial position association model through multi-section image alignment. It is specifically implemented by combining the feature point matching algorithm with the affine transformation matrix calculation to eliminate the axial offset error in the detection process; the center decision module calculates the final eccentricity and determines the center deviation of the conductor and the insulation layer through the probability distribution model. It is specifically implemented by using a Gaussian mixture model to perform statistical analysis on the center coordinates of multiple sections to improve the reliability of eccentricity calculation; the quality supervision platform generates a test report and uploads it to the blockchain to encrypt, store and visualize the test data. It is specifically implemented by combining a three-dimensional modeling engine with blockchain encryption technology to achieve traceability of the detection process and data security.

[0035] The technical scope of the present invention is not limited to the contents of the above description. Those skilled in the art can make various deformations and modifications to the above embodiments without departing from the technical idea of ​​the present invention, and these deformations and modifications should all fall within the protection scope of the present invention.

Claims

1. A cable quality detection and supervision system, characterized in that: include: Laser marking positioning module: used to establish a spatial coordinate system on the cable surface, generate spatial positioning marks on the surface of the insulation layer through a laser engraving device, and use a machine vision system to identify the mark position to establish a three-dimensional coordinate system; Ultrasonic discharge imaging module: used to obtain cross-sectional images at the conductor discharge position, stimulate ultrasonic induced cavitation discharge through a multi-channel piezoelectric array, and capture the conductor cross-sectional image with an optical coherence tomography; Intelligent image processing module: used to identify the structural features of the conductor, use an edge detection algorithm to extract the conductor contour and the insulation layer contour, delineate the candidate area based on multiple feature points on the contour, and calculate the initial center position; Iterative registration module: used to establish a multi-section spatial correlation model, realize the alignment of different cross-sectional images through a feature point matching algorithm, and construct the axial drift trajectory of the center; Center decision module: used to determine the true conductor center position, determine the true conductor center based on the probability distribution model of the multi-section center coordinates, and correct the coordinates based on the discharge position verification results; Quality control platform: used to visualize quality assessment results, generate eccentricity distribution maps through a 3D modeling engine, and store inspection data.

2. A cable quality detection and supervision system according to claim 1, characterized in that: The laser wavelength of the laser engraving device is 1064nm±50nm, the marking point diameter is ≤0.1mm, and the distance between adjacent marks is 50cm±10cm; the machine vision system uses a CMOS sensor with a resolution of ≥20 million pixels, and the mark recognition error is ≤0.05mm.

3. A cable quality detection and supervision system according to claim 1, characterized in that: The multi-channel piezoelectric array is a 40-channel PZT-8 piezoelectric ceramic array, and the ultrasonic excitation frequency range is 0.8-1.2 MHz; the optical coherence tomography device is polarization-sensitive, with an axial resolution of ≤5 μm and a lateral scanning range of ≥10 mm.

4. A cable quality detection and supervision system according to claim 1, characterized in that: The steps of calculating the initial center coordinates by the intelligent image processing module include: S1. Feature point extraction: screening multiple key feature points from the conductor contour, including but not limited to: curvature extreme points, the vertex of the conductor, the maximum or minimum curvature point of the arc contour; discharge association points, the contour points closest to the discharge position detected by the ultrasonic discharge imaging module; symmetrical feature points, the contour has a local symmetrical structure, and the intersection of the symmetry axis is extracted; the number of feature points is not less than 8, and they are evenly distributed around the contour, with a circumferential spacing of ≤45°; S2. Candidate area delineation: based on the coordinate distribution of the feature points, the candidate area of ​​the initial center is delineated using the minimum circumscribed circle method, and the radius of the area is the maximum value of the distance from the feature point to the distribution center × 1.2, to ensure that all feature points are within the area; S3. Initial center calculation: within the candidate area, the weighted centroid method is used to calculate the initial center coordinates, and the formula is: ;in, is the initial center coordinate, is the coordinate of the i-th feature point extracted from the cross-sectional image, is the geometric weight, the greater the curvature, the higher the weight. is the discharge weight, the closer the distance to the discharge point, the higher the weight, and n is the total number of feature points.

5. A cable quality detection and supervision system according to claim 1, characterized in that: The feature point matching technology of the iterative registration module includes: using the ORB algorithm to extract feature points of each cross-sectional image, with the number of feature points ≥50; using the PROSAC algorithm to eliminate mismatched points, with a retention rate ≥90%; and achieving image alignment based on the affine transformation formula: , where P is the coordinate of the feature point in the original cross-sectional image, M is the rotation matrix, which describes the angular deviation between the two cross-sections and is calculated by ORB feature point matching, and T is the translation vector, which represents the positional deviation between the two cross-sections. are the coordinates of the feature points after alignment.

6. A cable quality detection and supervision system according to claim 1, characterized in that: The probability distribution model is a Gaussian mixture model, which is expressed as: ;in is the probability density function of the center coordinates, indicating the possibility of the center appearing. is the weight of the kth Gaussian component , and are the mean and variance of the kth component respectively. The parameters are solved by the maximum expectation algorithm, and the number of components m is determined by the Bayesian information criterion.

7. A cable quality detection and supervision system according to claim 1, characterized in that: The discharge position verification includes: S71. measuring the distance from the discharge point to the current center ; S72. Calculate theoretical distance , where k is the insulation material constant, To calculate the maximum electric field intensity at the discharge point through electric field simulation, is the laboratory calibration correction item to compensate for the measurement error; S73. When , it is judged as an abnormal point and removed. is the measurement standard deviation.

8. The cable quality detection and supervision system according to claim 1, characterized in that: It also includes a vibration compensation unit, which collects six-axis vibration data through a MEMS accelerometer and calculates the value of the six-axis vibration compensation unit based on the formula Calculate the compensation displacement, is the real-time acceleration, is the material elastic coefficient, which is used to correct the position offset during image acquisition.

9. A cable quality detection and supervision system according to claim 1, characterized in that: In the quality control platform, the three-dimensional modeling engine uses WebGL technology to generate eccentricity heat maps with a display accuracy of ≤0.01mm.

10. A cable quality detection method, implemented using the cable quality detection and supervision system according to claim 1, characterized in that: The detection process includes: the laser marking positioning module initializes the spatial coordinate system; the ultrasonic discharge imaging module collects cross-sectional images at the marked point; the intelligent image processing module outputs the initial center coordinates of each section; the iterative alignment module constructs the center drift trajectory; the center decision module inputs the initial center coordinates of multiple sections and the discharge position to calculate the true conductor center coordinates; the quality control platform inputs the true conductor center and the insulation layer center, calculates the eccentricity, generates a detection report, and uploads it.

Images