Method for detecting outer diameter of high-voltage cable

Abstract

The invention provides a high-voltage cable outer diameter detection method, which comprises the following steps of S1, acquiring point cloud data of the surface of a cable joint, and preprocessing the point cloud data; s2, performing bidirectional slicing on the preprocessed point cloud data along the radial direction and the axial direction of the axis of the cable joint to form a plurality of local surface slices; each local surface slice comprises a plurality of point clouds; s3, based on the curvature weight of each point cloud on each local surface slice, constructing a key point of each local surface slice; and S4, key points of the local surface slices are matched according to the angles and the distances of the slices to form key point pairs, the distance between two key points in the key point pairs is calculated, and the measured value of the outer diameter of the high-voltage cable is obtained by calculating the average value of the distances of the multiple key point pairs. According to the method, the key points of the local surface slices are constructed, subjective errors and uncertainty in traditional manual evaluation are reduced, and the measurement precision of the outer diameter of the high-voltage cable is improved.

Description

Technical Field

[0001] The invention belongs to the technical field of cable detection, and in particular relates to a method for detecting the outer diameter of a high-voltage cable. Background Art

[0002] Power cables are the foundation of national power transmission. The quality of power cable construction directly determines the transmission capacity of power cables and the stability of power transmission. Power cable construction is long and involves a large amount of work, which poses a lot of safety hazards and makes fault detection difficult. Therefore, it is extremely important to control the quality of power cable construction. Only by ensuring the quality of power cable construction can the stable transmission of power by power cables be fundamentally guaranteed.

[0003] With the continuous growth of the demand for electricity in urban life and industrial production, the scale of laying power grids is getting larger and larger, and the scope of application is getting wider and wider. Since power cable lines are less affected by natural weather, save air resources and space, and have little impact on the surrounding environment, they have gradually become the first choice for urban power transmission. The high-voltage cables currently used in my country are cross-linked polyethylene (XLPE) cables. XLPE cables have become the most widely used cable type in urban power transmission lines due to their better heat resistance, insulation, mechanical strength, and safe laying. They are widely used in high-voltage and ultra-high-voltage long-distance and ultra-long-distance transmission lines.

[0004] However, due to the strong concealment of cable lines, it is difficult for technicians to diagnose cable faults, and it is not easy to discover and make accurate judgments in time. In order to ensure that cable lines are safer and more reliable during operation, it is necessary to improve the construction quality of cable construction and provide effective support for cable operation and maintenance. When constructing cables, power supply companies should actively take quality control measures and do a good job in the management and control of engineering construction.

[0005] Before high-voltage cables are buried underground, construction workers usually remove the protective cover and grind the semi-conductive material to make intermediate joints on site to achieve the connection between high-voltage cables. Affected by factors such as processing site conditions, workers' grinding level, and installation operations, the surface of the intermediate joints of high-voltage cables is prone to defects such as insufficient grinding, excessive grinding, and scratches. Defects on the surface of the intermediate joints will cause gaps between the cable and the insulating sheath, causing the high-voltage cable to be broken down and catch fire under the action of long-term high voltage, thereby causing serious safety accidents. In the domestic power system safety accident records, the failure rate caused by surface defects of high-voltage cable intermediate joints is as high as 70%, so the quality inspection of high-voltage cable grinding is an indispensable and important part of power system maintenance. Summary of the invention

[0006] In view of the fact that the quality of high-voltage cables currently still relies on manual inspection before laying, which has the problems of strong subjectivity and poor precision, the present invention provides a method for detecting the outer diameter of a high-voltage cable to avoid the subjective measurement deviation caused by manual measurement, while reducing the influence of surface defects and shape deformation of the high-voltage cable on the point cloud of the cable joint when measuring the outer diameter of the high-voltage cable, thereby improving the detection accuracy of the outer diameter of the high-voltage cable and the construction quality of the cable joint.

[0007] To achieve the above-mentioned purpose, the present invention provides a method for detecting the outer diameter of a high-voltage cable, comprising the following steps: step S1, obtaining point cloud data of the surface of a cable joint, and preprocessing the point cloud data; step S2, bidirectionally slicing the preprocessed point cloud data along the radial and axial directions of the cable joint axis to form a plurality of local surface slices; each of the local surface slices contains a plurality of point clouds; step S3, constructing key points of each local surface slice based on the curvature weights of each point cloud on each local surface slice; step S4, matching the key points of each local surface slice according to the angle and distance of the slice to form a key point pair, calculating the distance between two key points in the key point pair, and obtaining the outer diameter measurement value of the high-voltage cable by calculating the average distance of a plurality of key point pairs.

[0008] Optionally, the step S1 specifically includes: S11, collecting point cloud data along the circumference of the cable connector surface through a three-dimensional imaging device; S12, deleting noise areas with low point cloud density from the point cloud data collected in step S11 through a statistical filtering algorithm; S13, performing cylinder fitting on the point cloud data filtered in step S12 to form a fitted cable connector, and using the cylindrical axis of the fitted cylinder as the axis of the cable connector; S14, rotating and translating the fitted cable connector so that the axis of the fitted cable connector coincides with the Z axis of the coordinate system in which it is located, performing rotational and translational coordinate transformation on the point cloud data, and sorting the point cloud data in ascending order according to the Z coordinate value in the point cloud data to obtain preprocessed point cloud data.

[0009] Optionally, the bidirectional slicing in step S2 includes radial equiangular slicing and axial equidistant sliding slicing, which specifically includes the following steps: S21, determine the measurement area of ​​the cable joint, take the XOZ plane of the cable joint axis as the starting position, and set the rotation angle increment , the pre-processed point cloud data is divided into multiple arc intervals by rotating counterclockwise around the axis, and each arc interval is numbered; S22, traversing the point clouds within each numbered arc interval, and extracting the point clouds of the arc interval between the segmentation points according to the decision conditions; S23, using a plane perpendicular to the axis of the cable connector to set the sliding interval , perform equidistant sliding slicing on the point cloud within the arc interval along the axis direction, and obtain a plurality of local surface slices according to the judgment conditions.

[0010] Optionally, the decision conditions in step S22 include equation (3) and equation (4): (3) (4) in, is the rotation angle increment, N is the axial slice number corresponding to the measurement position, represents the original coordinates of the point cloud, represents the coordinates in the XOZ plane, and The lines between them form the point axis, The point axis and X The angle of the axis, n Indicates the arc interval number, Represents the radial slice point cloud of the arc interval that meets the decision condition; The expression (5) of the judgment condition in step S23 is as follows: (5) in, and Indicates the Z-axis coordinates corresponding to the start and end positions of the sliding window. Indicates the Z-axis coordinate of the point cloud during axial slicing. Represents the point cloud coordinates after bidirectional slicing, Represents a local surface slice formed after bidirectional slicing.

[0011] Optionally, the step S3 specifically includes: S31, calculating a first normal vector of a first point cloud arbitrarily selected on a local surface slice, and performing curvature estimation according to a second normal vector of a second point cloud adjacent to the first point cloud; S32, repeating step S31 to traverse all point clouds on the local surface slice to obtain curvature estimates of all point clouds; S33, performing normalized average on the curvature estimates of all point clouds, and calculating the curvature weight corresponding to each point cloud; S34, obtaining the original coordinate information of each point cloud of the local surface slice, calculating the weighted coordinate average of each point cloud on the local surface slice according to the curvature weight of each point cloud, and completing the construction of the key points of each local surface slice by calculating the ratio of the weighted coordinate average of each point cloud to the original coordinate average.

[0012] Optionally, the step S31 specifically includes: S311, selecting a first point cloud on the local surface slice , set the search radius, use the kd tree for nearest neighbor search, and use the least squares method to perform local plane fitting on all second point clouds obtained by the kd tree nearest neighbor search. The resulting fitting plane expression (6) is as follows: (6) in, To fit the plane P The normal vector of d To fit the plane P The distance to the origin of the coordinate system; S312, solving the fitting plane expression in step S311 to obtain the first point cloud The first normal vector ; S313, from the first point cloud of m Select a second point cloud from the neighboring points Calculate the normal curvature by taking the first point cloud is the origin and the first point cloud The normal vector As the axis, create the first point cloud The local orthogonal coordinate system of the second point cloud is determined The second normal vector ; S314, passing through the first point cloud The curvature circle is estimated from the first point cloud The normal curvature , where the curvature radius of the curvature circle represents the distance from the center O to the first point cloud The distance and the normal vector of the center O; establish the first point cloud , Second point cloud The trigonometric relationship between the circle center O is used to solve the normal curvature. ; S315, estimate the first point cloud The principal curvature of The normal vector of the plane F will be the second point cloud Projecting onto plane F to get the third point cloud , using the Euler formula, by iteratively performing least squares fitting on all adjacent points, constructing a least squares fitting expression, and constructing a loss function based on the sum of squared errors, to obtain an estimate of the optimal parameter δ relative to the principal curvature; the estimate of the optimal parameter δ is formed into a matrix, and the matrix is ​​decomposed to obtain the eigenvalue of the matrix, which is the first point cloud The principal curvature of .

[0013] Optionally, the step S312 of solving the fitting plane expression in step S311 is specifically: transforming the fitting plane expression (6) into an eigenvalue decomposition of a covariance matrix M composed of a set of local neighborhood points, as shown in formula (7): (7) in is the centroid of the neighborhood point set, represents the eigenvalue of the matrix M, express The eigenvector of > > , The first point cloud The first normal vector Estimates.

[0014] Optionally, the curvature estimation of all point clouds described in step S33 is normalized and averaged, and the curvature weight corresponding to each point cloud is calculated by the following formula (13): (13) Where C represents the average curvature of each cloud point, represents the maximum mean curvature value among all cloud points on the local surface slice, represents the minimum mean curvature value among all cloud points on the local surface slice, Represents the curvature weight of each cloud point.

[0015] Optionally, the calculation formula of the local surface slice key points in step S34 is as shown in the following formula (14): (14).

[0016] Optionally, the step S4 specifically includes: S41, determining the measurement position of the cable connector where the outer diameter needs to be measured, extracting all radial slice point clouds corresponding to the radial direction of the measurement position, and a total of A radial slice point cloud; all local surface slices in the axial direction are calculated by equation (15); (15) in, N is the axial slice number corresponding to the measurement position, location_j Indicates the Z-axis coordinate value of the measurement position, end is the minimum point in the Z-axis direction of the point cloud model, represents the axial slice distance; S42, extract the key point set K of the measurement position, and perform key point matching according to the angle constraint; specifically: select a local surface slice as the initial line segment, and its key point coordinates are , rotate 180° counterclockwise along the radial direction to find the local surface slice relative to the initial line segment, and its key point coordinates are And save it to the key point pair set H; wherein the expression of the angle constraint is shown in the following formula (16): (16); S43, calculate the Euclidean distance between the key points of two local surface slices, and traverse the remaining local surface slices, a total of The distance value of the key point pair is calculated, and the average distance value of the key point pair is calculated to represent the outer diameter of the measurement position, and the outer diameter measurement value of the high-voltage cable is obtained, which is specifically shown in the following formula (17): (17).

[0017] Compared with the prior art, the high-voltage cable outer diameter detection method of the present invention realizes accurate scanning and quality assessment of cable joints by collecting three-dimensional point cloud data on the surface of cable joints, thereby reducing subjective errors and uncertainties in traditional manual assessments, greatly reducing the demand for human resources, and improving work efficiency; further, by bidirectionally slicing the point cloud data to form local surface slices, the uniformity of the segmented point cloud is improved; at the same time, by constructing key points of local surface slices, the influence of surface defects and shape deformation on the point cloud measurement of cable joints during outer diameter measurement is reduced, thereby improving the measurement accuracy of the outer diameter of the high-voltage cable. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a flow chart of the high-voltage cable outer diameter detection method of the present invention; Figure 2 This is a schematic diagram of point cloud data of a high-voltage cable connector collected in an embodiment of the present invention; Figure 3 It is a schematic diagram of point cloud data splicing of a high-voltage cable joint in an embodiment of the present invention; Figure 4 is a schematic diagram of point cloud data of a high-voltage cable joint before statistical filtering in an embodiment of the present invention; Figure 5 is a schematic diagram of point cloud data of a high-voltage cable joint after statistical filtering in an embodiment of the present invention; Figure 6 It is a schematic diagram of cylindrical fitting of point cloud data of a high-voltage cable joint in an embodiment of the present invention; Figure 7A schematic diagram of rotation, translation and sorting of point cloud data of a high-voltage cable connector in an embodiment of the present invention; Figure 8 is a bidirectional slice diagram of a high voltage cable connector in an embodiment of the present invention; Fig. 9 A visualization schematic diagram of key points of a local surface slice of a high-voltage cable joint in an embodiment of the present invention; Fig.10 It is a schematic diagram of measuring the outer diameter parameters of the high-voltage cable connector in the embodiment of the present invention. DETAILED DESCRIPTION

[0019] The following will be combined with the attached embodiment of the present invention Figure 1 ~Attached Fig.10 , the technical solutions, structural features, objectives achieved and effects in the embodiments of the present invention are described in detail.

[0020] It should be noted that the drawings are in a very simplified form and use non-precise proportions. They are only used to conveniently and clearly assist in explaining the embodiments of the present invention, and are not used to limit the conditions for the implementation of the present invention. Therefore, they have no substantive technical significance. Any structural modification, change in proportional relationship or adjustment of size, without affecting the effects and purposes that can be achieved by the present invention, should still fall within the scope of the technical content disclosed by the present invention.

[0021] It should be noted that, in the present invention, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only the elements explicitly listed, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0022] In order to avoid the subjective measurement deviation in manual measurement of the cable joint diameter and reduce the influence of surface defects and shape deformation of the high-voltage cable on the cable joint point cloud when measuring the outer diameter of the high-voltage cable, an embodiment of the present invention provides a high-voltage cable outer diameter detection method, such as Figure 1 As shown, the detection method includes the following steps S1 to S4: Step S1, obtaining point cloud data of the cable connector surface, and preprocessing the point cloud data; in specific implementation, the step S1 specifically includes: S11, collecting point cloud data along the circumference of the cable connector surface by a three-dimensional imaging device (a structured light camera is used in this embodiment); in this embodiment, when collecting point cloud data, the point cloud data is collected in front of the cable connector ( Figure 2 Figure a), left front ( Figure 2 b) and right front ( Figure 2 Point cloud data is collected at three different positions of the c figure (i.e., each position differs by 120°). The collected point cloud data is as follows Figure 2 As shown, the three sets of point cloud data are then spliced ​​using CloudCompare software, as shown in Figure 3 As shown, Figure 3 Figure a is the front view of the cable connector surface after the point cloud data is spliced. Figure 3 Figure b is a top view of the point cloud data of the cable connector surface after splicing. In other embodiments, a scanner can be used to scan around the cable connector surface for a circle, so that complete point cloud acquisition data can be obtained without splicing. The point cloud data includes the coordinate information of the X-axis, Y-axis, and Z-axis of the point cloud and RGB information.

[0023] S12, the point cloud data collected in step S11 is subjected to a statistical filtering algorithm to delete the noise area with low point cloud density; this is because filtering can reduce the influence of noise on the measurement accuracy. The point cloud data of the cable connector surface before filtering is as follows: Figure 4 As shown in the figure, the point cloud data of the cable joint surface after filtering is as follows: Figure 5 As shown, the specific process of the filtering is as follows: the statistical filter first statistically analyzes the distance from the point cloud in the neighborhood k of each point cloud to the point cloud, and calculates the mean value according to the distribution and standard deviation , and then set the standard deviation threshold To determine whether the mean value is within the distance between the point cloud and all point clouds in the neighborhood k If it is beyond the range, it is defined as an outlier point and deleted. Finally, the point filtering of all point clouds is completed iteratively.

[0024] S13, performing cylinder fitting on the point cloud data filtered in step S12 to form a fitted cable joint, and using the cylindrical axis of the fitted cylinder as the axis of the cable joint (eg Figure 6 shown); S14, rotating and translating the fitted cable joint so that the axis of the fitted cable joint coincides with the Z axis of the coordinate system in which it is located, performing a rotational and translational coordinate transformation on the point cloud data, and sorting the point cloud data in ascending order according to the Z coordinate values ​​in the point cloud data to obtain preprocessed point cloud data (such as Figure 7 as shown).

[0025] The point cloud rotation operation is achieved by calculating the rotation vector, and the rotation matrix is ​​shown in the following formula (1): (1) in, is perpendicular to both the cable connector axis and the spatial coordinate system Z The normal vector of the axis, represents the original point cloud coordinate vector, represents the matrix generated by the antisymmetric normal vector, R represents the rotation matrix, I represents the unit matrix, r represents the direction cosines of each axis of the rotated coordinate system along each dimension in the original coordinate system, Indicates the rotation angle; Further, the original point cloud coordinate vector Perform translation operation to complete the coordinate transformation, as shown in the following formula (2): (2) in, Indicates the translation distance. Finally, the point cloud data is sorted in ascending order according to the Z-axis coordinate value of the point cloud data, and the highest point on the cable connector surface is determined.

[0026] For the measurement of the actual cable joint point cloud, the original features must be retained to obtain an outer diameter parameter that is closer to the actual one. The surface of the cable joint polished by technicians inevitably contains defects and dimensional deviations. When the cable joint is sliced ​​horizontally along the axial direction of the cable joint axis, the cross section forms an irregular circle. At this time, directly fitting geometric models such as circles or ellipses to measure the outer diameter of the cable joint and the XY axis outer diameter deviation will lead to large measurement errors. Therefore, this detection method divides the surface of the cable joint into several small faces to construct local feature points. Based on this, the outer diameter parameters of the cable joint are measured.

[0027] Furthermore, since the surface formed by the point cloud on the surface of the cable joint is a curved surface and the normal vector generally points to the axis of the cable joint, on the one hand, the use of meshing and unidirectional slicing will lead to uneven distribution of the point cloud; on the other hand, feature-based segmentation faces the challenge of threshold determination, and its versatility and robustness are weak, making it difficult to directly apply to the point cloud segmentation of the cable joint surface.

[0028] In response to the above problems, an embodiment of the present invention provides a point cloud bidirectional slicing step S2 based on the cable joint axis: bidirectional slicing of the preprocessed point cloud data along the radial and axial directions of the cable joint axis to form a number of local surface slices; each of the local surface slices contains multiple point clouds.

[0029] In specific implementation, the bidirectional slicing described in step S2 includes radial equiangular slicing and axial equidistant sliding slicing. For the radial slicing of the point cloud on the cable joint surface, only the x and y coordinates of the point cloud are considered, such as Figure 8 As shown, the specific steps include: S21, determine the measurement area of ​​the cable joint, take the XOZ plane of the cable joint axis as the starting position, and set the rotation angle increment , the pre-processed point cloud data is divided into multiple arc intervals by rotating counterclockwise around the axis, and each arc interval is numbered; S22, traversing the point clouds within each numbered arc interval, and extracting the point clouds of the arc interval between the segmentation points according to the decision conditions; wherein the decision conditions include equations (3) and (4): (3) (4) in, is the rotation angle increment, N is the axial slice number corresponding to the measurement position, represents the original coordinates of the point cloud, represents the coordinates in the XOZ plane, and The lines between them form the point axis, The point axis and X The angle of the axis, n Indicates the arc interval number, Represents the radial slice point cloud of the arc interval that meets the decision condition; S23, using a plane perpendicular to the axis of the cable connector to set the sliding interval , perform equidistant sliding slicing on the point cloud in the arc interval along the axis direction, and obtain a number of local surface slices according to the judgment condition; wherein the expression (5) of the judgment condition is as follows: (5) in, and Indicates the Z-axis coordinates corresponding to the start and end positions of the sliding window. Indicates the Z-axis coordinate of the point cloud during axial slicing. Represents the point cloud coordinates after bidirectional slicing, represents the local surface slice formed after bidirectional slicing, such as Figure 8 It should be noted that the original point cloud data is retained on the local surface slices formed after bidirectional slicing.

[0030] Furthermore, in the process of detecting the outer diameter of the cable, it is necessary to calculate the distance between each local surface slice. Since each local surface slice contains many point clouds, and there is no direct correspondence between the point clouds on different local surface slices, it is not feasible to directly measure the outer diameter of the cable. Therefore, it is necessary to construct key points to simplify the surface slices for subsequent analysis. In the process of cable joint detection, due to the deformation of the cable joint surface and the influence of unfiltered noise, it is necessary to construct appropriate key points to accurately measure the outer diameter parameters of the cable joint. Commonly used key points include center points, centroid points, and extreme points. Among them, center points and centroid points are less sensitive to noise or have less changes, while extreme points are easy to extract the features of the surface edge, resulting in feature point offset, affecting the subsequent measurement accuracy.

[0031] Therefore, in order to reduce the impact of surface defects and shape deformation on the measurement of cable joint point cloud during outer diameter measurement, an embodiment of the present invention provides a step S3 of constructing surface slice key points based on the coordinate weighted average of the cable joint surface curvature: based on the curvature weight of each point cloud on each local surface slice, construct the key points of each local surface slice.

[0032] In specific implementation, this embodiment uses principal component analysis (PCA) to calculate the normal vector of the space plane and regards it as the normal vector of the sampling point. The step S3 specifically includes: S31, calculating a first normal vector of a first point cloud arbitrarily selected on a local surface slice, and performing curvature estimation according to a second normal vector of a second point cloud adjacent to the first point cloud; S32, repeating step S31 to traverse all point clouds on the local surface slice to obtain curvature estimates of all point clouds; S33, performing normalized average on the curvature estimates of all point clouds, and calculating the curvature weight corresponding to each point cloud; S34, obtaining the original coordinate information of each point cloud of the local surface slice, calculating the weighted coordinate average of each point cloud on the local surface slice according to the curvature weight of each point cloud, and completing the construction of the key points of each local surface slice by calculating the ratio of the weighted coordinate average of each point cloud to the original coordinate average.

[0033] Further, in this embodiment, the step S31 specifically includes: S311, selecting the first point cloud on the local surface slice , set the search radius, use the kd tree for nearest neighbor search, and use the least squares method to perform local plane fitting on all second point clouds obtained by the kd tree nearest neighbor search. The resulting fitting plane expression (6) is as follows: (6) in, To fit the plane P The normal vector of d To fit the plane P The distance to the coordinate origin.

[0034] S312, solving the fitting plane expression in step S311 to obtain the first point cloud The first normal vector In this embodiment, the fitting plane expression (6) is transformed into the eigenvalue decomposition of the covariance matrix M composed of a set of local neighborhood points, as shown in equation (7): (7) in is the centroid of the neighborhood point set, represents the eigenvalue of the matrix M, express The eigenvector of > > , The first point cloud The first normal vector Estimates.

[0035] S313, from the first point cloud of m Select a second point cloud from the neighboring points Calculate the normal curvature by taking the first point cloud is the origin and the first point cloud The normal vector As the axis, create the first point cloud The local orthogonal coordinate system of the second point cloud is determined The second normal vector ; S314, passing through the first point cloud The curvature circle is estimated from the first point cloud The normal curvature , where the curvature radius of the curvature circle represents the distance from the center O to the first point cloud The distance and the normal vector of the center O; establish the first point cloud , Second point cloud The trigonometric relationship between the circle center O is used to solve the normal curvature. In this embodiment, the normal vector curvature is solved The specific formulas are shown in formula (8) and formula (9): (8) (9) in, For the first point cloud and the direction cosines between the second normal vector Q, is the first normal vector With the first point cloud The vector at The angle between Represents the first normal vector With the second point cloud The angle between the second normal vector Q at Represents the first point cloud The original coordinates of S315, estimate the first point cloud The principal curvature of The normal vector of the plane F will be the second point cloud Projecting onto plane F to get the third point cloud , using the Euler formula, by iteratively performing least squares fitting on all adjacent points, constructing a least squares fitting expression, and constructing a loss function based on the sum of squared errors, to obtain an estimate of the optimal parameter δ relative to the principal curvature; the estimate of the optimal parameter δ is formed into a matrix, and the matrix is ​​decomposed to obtain the eigenvalue of the matrix, which is the first point cloud The principal curvature of .

[0036] Specifically, in this embodiment, the least squares fitting expression is shown in formula (10): (10) in and The first point cloud The corresponding principal curvature is The direction vector of the principal curvature With vector The projected vector The angle between the plane F and is a vector and vector The angle between In this embodiment, the estimation of the optimal parameter δ is calculated by converting formula (10) into matrix A and matrix B, and constructing a loss function based on the sum of squared errors: , so as to minimize the sum of squared errors and obtain the relative principal curvature and The estimation of the optimal parameter δ is shown in formula (11): (11); Furthermore, in this embodiment, the matrix D is formed according to the estimation of the optimal parameter δ, and the matrix is ​​decomposed to obtain the eigenvalues ​​of the matrix D and , the eigenvalue and For the first point cloud The principal curvature and , the matrix D is shown in formula (12): (12).

[0037] In order to ensure that the principal curvature ratio of each point cloud is consistent, the principal curvature of each point cloud is normalized, and all point clouds on the local surface slice are traversed to complete the curvature estimation of all point clouds.

[0038] The curvature estimation of all point clouds is normalized and averaged in step S33, and the curvature weight corresponding to each point cloud is calculated as shown in the following formula (13): (13) Where C represents the average curvature of each cloud point, represents the maximum mean curvature value among all cloud points on the local surface slice, represents the minimum mean curvature value among all cloud points on the local surface slice, Represents the curvature weight of each cloud point.

[0039] Based on the calculation formula of curvature weight in formula (13), the calculation formula of the local surface slice key point in step S34 is as shown in formula (14): (14).

[0040] After completing the construction of the local surface slice key points, the visualized local surface slice key points are as follows Fig. 9 As shown, Fig. 9 The red points in the figure are the key points of local surface slices. It is further necessary to match these key points of local surface slices to realize the detection of the outer diameter parameters of the cable joint. Specifically, S4, according to the angle and distance of the slice, the key points of each local surface slice are matched to form a key point pair, the distance between two key points in the key point pair is calculated, and the outer diameter measurement value of the high-voltage cable is obtained by calculating the average distance of multiple key point pairs.

[0041] In the specific implementation, in this embodiment, the step S4 specifically includes: S41, determine the measurement position of the cable joint where the outer diameter needs to be measured, extract all radial slice point clouds corresponding to the radial direction of the measurement position, a total of A radial slice point cloud; all local surface slices in the axial direction are calculated by equation (15); (15) in,N is the axial slice number corresponding to the measurement position, location_j Indicates the Z coordinate value of the measurement position, end is the minimum point in the Z-axis direction of the point cloud model, represents the axial slice distance; S42, extract the key point set K of the measurement position, and perform key point matching according to the angle constraint; specifically: select a local surface slice as the initial line segment, and its key point coordinates are , rotate 180° counterclockwise along the radial direction to find the local surface slice relative to the initial line segment, and its key point coordinates are And save it to the key point pair set H; wherein the expression of the angle constraint is shown in the following formula (16): (16); S43, calculate the Euclidean distance between the key points of two local surface slices, and traverse the remaining local surface slices, a total of The distance value of the key point pair is calculated, and the average distance value of the key point pair is calculated to represent the outer diameter of the measurement position, and the outer diameter measurement value of the high-voltage cable is obtained, which is specifically shown in the following formula (17): (17) The outer diameter detection value of the high-voltage cable can be obtained from formula (17).

[0042] At the same time, the embodiment of the present invention can also measure the maximum deviation of the XY axis outer diameter according to the orthogonality between the key point pairs. The calculation formula of the XY axis outer diameter deviation at the current measurement position is as follows: (18) Where n represents the number of local surface slices, represents the key point coordinates, and Represents the distance between keypoint pairs.

[0043] Based on the high-voltage cable outer diameter detection method provided by the embodiment of the present invention, in a specific embodiment, Fig.10 As shown, the outer diameter of the cable connector calculated by steps S1 to S4 is 45.40 mm, which is only 0.2 mm different from the actual measured outer diameter of 45.20 mm, and the error is only 0.44%, which can meet the quality inspection requirements of high-voltage cables.

[0044] To sum up, compared with the prior art, the high-voltage cable outer diameter detection method of the present invention realizes accurate scanning and quality assessment of cable joints by collecting three-dimensional point cloud data on the surface of cable joints, thereby reducing subjective errors and uncertainties in traditional manual assessments, greatly reducing the demand for human resources, and improving work efficiency; further, by bidirectionally slicing the point cloud data to form local surface slices, the uniformity of the segmented point cloud is improved; at the same time, by constructing key points of local surface slices, the influence of surface defects and shape deformation on the point cloud measurement of cable joints during outer diameter measurement is reduced, thereby improving the measurement accuracy of the outer diameter of the high-voltage cable.

[0045] Although the content of the present invention has been described in detail through the above preferred embodiments, it should be appreciated that the above description should not be considered as a limitation of the present invention. After reading the above content, it will be apparent to those skilled in the art that various modifications and substitutions of the present invention will occur. Therefore, the protection scope of the present invention should be limited by the appended claims.

Claims

1. A method for detecting the outer diameter of a high-voltage cable, characterized in that: The steps include: Step S1, obtaining point cloud data of the cable connector surface, and preprocessing the point cloud data; Step S2, bidirectionally slicing the preprocessed point cloud data along the radial direction and the axial direction of the cable connector axis to form a plurality of local surface slices; each of the local surface slices contains a plurality of point clouds; Step S3, constructing key points of each local surface slice based on the curvature weight of each point cloud on each local surface slice; Step S4, matching the key points of each of the local surface slices according to the angle and distance of the slices to form a key point pair, calculating the distance between two key points in the key point pair, and obtaining the outer diameter measurement value of the high-voltage cable by calculating the average distance of multiple key point pairs.

2. The method for detecting the outer diameter of a high-voltage cable according to claim 1, characterized in that: The step S1 specifically includes: S11, collecting point cloud data along the circumference of the cable connector surface by a three-dimensional imaging device; S12, deleting the noise area with low point cloud density by using a statistical filtering algorithm on the point cloud data collected in step S11; S13, performing cylinder fitting on the point cloud data filtered in step S12 to form a fitted cable joint, and using the cylindrical axis of the fitted cylinder as the axis of the cable joint; S14, rotating and translating the fitted cable connector so that the axis of the fitted cable connector coincides with the Z axis of the coordinate system in which it is located, performing a rotational and translational coordinate transformation on the point cloud data, and sorting the point cloud data in ascending order according to the Z coordinate values ​​in the point cloud data to obtain preprocessed point cloud data.

3. The method for detecting the outer diameter of a high-voltage cable according to claim 2, characterized in that: The bidirectional slicing described in step S2 includes radial equiangular slicing and axial equidistant sliding slicing, and specifically includes the following steps: S21, determine the measurement area of ​​the cable joint, take the XOZ plane of the cable joint axis as the starting position, and set the rotation angle increment , the pre-processed point cloud data is divided into multiple arc intervals by rotating counterclockwise around the axis, and each arc interval is numbered; S22, traversing the point clouds within each numbered arc interval, and extracting the point clouds of the arc interval between the segmentation points according to the decision conditions; S23, using a plane perpendicular to the cable joint axis to set the sliding interval , perform equidistant sliding slicing on the point cloud within the arc interval along the axis direction, and obtain a plurality of local surface slices according to the judgment conditions.

4. The method for detecting the outer diameter of a high-voltage cable according to claim 3, characterized in that: The decision conditions described in step S22 include equation (3) and equation (4): （3） （4） in, is the rotation angle increment, N is the axial slice number corresponding to the measurement position, represents the original coordinates of the point cloud, represents the coordinates in the XOZ plane, and The lines between them form the point axis, The point axis and X The angle of the axis, n Indicates the arc interval number, Represents the radial slice point cloud of the arc interval that meets the decision condition; The expression (5) of the judgment condition in step S23 is as follows: （5） in, and Indicates the Z-axis coordinates corresponding to the start and end positions of the sliding window. Indicates the Z-axis coordinate of the point cloud during axial slicing. Represents the point cloud coordinates after bidirectional slicing, Represents a local surface slice formed after bidirectional slicing.

5. The method for detecting the outer diameter of a high-voltage cable according to claim 1, characterized in that: The step S3 specifically includes: S31, calculating a first normal vector of a first point cloud arbitrarily selected on a local surface slice, and performing curvature estimation according to a second normal vector of a second point cloud adjacent to the first point cloud; S32, repeating step S31 to traverse all point clouds on the local surface slice to obtain curvature estimates of all point clouds; S33, performing normalized average on the curvature estimates of all point clouds, and calculating the curvature weight corresponding to each point cloud; S34, obtaining the original coordinate information of each point cloud of the local surface slice, calculating the weighted coordinate average of each point cloud on the local surface slice according to the curvature weight of each point cloud, and completing the construction of the key points of each local surface slice by calculating the ratio of the weighted coordinate average of each point cloud to the original coordinate average.

6. The method for detecting the outer diameter of a high-voltage cable according to claim 5, characterized in that: The step S31 specifically includes: S311, selecting the first point cloud on the local surface slice , set the search radius, use the kd tree for nearest neighbor search, and use the least squares method to perform local plane fitting on all second point clouds obtained by the kd tree nearest neighbor search. The resulting fitting plane expression (6) is as follows: （6） in, Fitting plane P The normal vector of d Fitting plane P The distance to the origin of the coordinate system; S312, solving the fitting plane expression in step S311 to obtain the first point cloud The first normal vector ; S313, from the first point cloud of m Select a second point cloud from the neighboring points Calculate the normal curvature by taking the first point cloud is the origin and the first point cloud The normal vector As the axis, create the first point cloud The local orthogonal coordinate system of the second point cloud is determined The second normal vector ; S314, passing through the first point cloud The curvature circle is estimated from the first point cloud The normal curvature , where the curvature radius of the curvature circle represents the distance from the center O to the first point cloud The distance and the normal vector of the center O; establish the first point cloud , Second point cloud The trigonometric relationship between the circle center O is used to solve the normal curvature. ; S315, estimate the first point cloud The principal curvature of The normal vector of the plane F will be the second point cloud Projecting onto plane F to get the third point cloud , using the Euler formula, by iteratively performing least squares fitting on all adjacent points, constructing a least squares fitting expression, and constructing a loss function based on the sum of squared errors, to obtain an estimate of the optimal parameter δ relative to the principal curvature; the estimate of the optimal parameter δ is formed into a matrix, and the matrix is ​​decomposed to obtain the eigenvalue of the matrix, which is the first point cloud The principal curvature of .

7. The method for detecting the outer diameter of a high-voltage cable according to claim 6, characterized in that: The step S312 of solving the fitting plane expression in step S311 is specifically: by transforming the fitting plane expression (6) into the eigenvalue decomposition of the covariance matrix M composed of the local neighborhood point set, as shown in formula (7): （7） in is the centroid of the neighborhood point set, represents the eigenvalue of the matrix M, express The eigenvector of > > , The first point cloud The first normal vector Estimates.

8. The method for detecting the outer diameter of a high-voltage cable according to claim 6, characterized in that: The curvature estimation of all point clouds described in step S33 is normalized and averaged, and the calculation formula for calculating the curvature weight corresponding to each point cloud is shown in the following formula (13): （13） Where C represents the average curvature of each cloud point, represents the maximum mean curvature value among all cloud points on the local surface slice, represents the minimum mean curvature value among all cloud points on the local surface slice, Represents the curvature weight of each cloud point.

9. The method for detecting the outer diameter of a high-voltage cable according to claim 8, characterized in that: The calculation formula of the local surface slice key points in step S34 is shown in the following formula (14): （14）。 10. The method for detecting the outer diameter of a high-voltage cable according to claim 9, characterized in that: The step S4 specifically includes: S41, determine the measurement position of the cable joint where the outer diameter needs to be measured, extract all radial slice point clouds corresponding to the radial direction of the measurement position, a total of A radial slice point cloud; all local surface slices in the axial direction are calculated by equation (15); （15） in, N is the axial slice number corresponding to the measurement position, location_j Indicates the Z-axis coordinate value of the measurement position, end is the minimum point in the Z-axis direction of the point cloud model, represents the axial slice distance; S42, extract the key point set K of the measurement position, and perform key point matching according to the angle constraint; specifically: select a local surface slice as the initial line segment, and its key point coordinates are , rotate 180° counterclockwise along the radial direction to find the local surface slice relative to the initial line segment, and its key point coordinates are And save it to the key point pair set H; wherein the expression of the angle constraint is shown in the following formula (16): （16）； S43, calculate the Euclidean distance between the key points of two local surface slices, and traverse the remaining local surface slices, a total of The distance value of the key point pair is calculated, and the average distance value of the key point pair is calculated to represent the outer diameter of the measurement position, and the outer diameter measurement value of the high-voltage cable is obtained, which is specifically shown in the following formula (17): （17）。

Images