A detection method for strain clamps of transmission lines

Abstract

The present invention relates to a detection method for a strain clamp of a transmission line. The method includes the following steps: 1) Obtain the spatial point cloud data of the strain clamp, convert it into the form of a.txt file, and establish a point cloud model; 2) Use a statistical filtering algorithm to denoise the point cloud model, and use a grid-based simplification method based on curvature features to simplify the point cloud model; 3) Use the RANSAC algorithm to achieve clustering segmentation of the point cloud model; 4) Use the NURBS surface fitting method to fit the point cloud data to realize the three-dimensional model reconstruction of the strain clamp; 5) According to the characteristic surface equation of the strain clamp, realize the registration between the measured strain clamp and the standard strain clamp, calculate the characteristic distance between the point cloud model of the measured strain clamp and the point cloud model of the standard strain clamp, and thus complete the quantitative calculation and analysis of the model deviation value. Compared with the prior art, the present invention has the advantages of high precision, visualization, high efficiency, etc.

Description

Technical Field

[0001] The present invention relates to the field of appearance dimension measurement of strain clamps for transmission lines, and in particular to a detection method for strain clamps of transmission lines. Background Art

[0002] With the continuous development of UHV power transmission and transformation projects in China, the requirements for the safety, stability, and reliability of UHV transmission lines are also constantly increasing. As a key device in UHV (extra-high voltage) transmission lines, strain clamps are widely used in the connection of conductors and ground wires; many line operation accidents are caused by the crimping quality problems of strain clamps. Therefore, accurately and efficiently detecting the crimping quality of strain clamps and promptly remedying corresponding defects is conducive to the safe, stable, and reliable operation of transmission lines.

[0003] Currently, the quality detection methods for crimped type electrical fittings mainly include: manual appearance dimension detection, grip force test, and radiographic non-destructive detection. (1) Appearance dimension detection is to measure and record the appearance dimensions of the strain clamp before and after crimping using a vernier caliper, mainly relying on manual measurement and visual identification. (2) X-ray non-destructive detection utilizes the principle that different metal structures absorb different rays, and can conveniently, quickly, and accurately capture the image of the internal crimping state of electrical fittings without damaging the object to be detected. (3) The grip force test is a destructive sampling method carried out in the laboratory, which can intuitively detect the external and internal crimping states of electrical fittings.

[0004] In engineering practice, in order to detect the crimping quality of strain clamps, front-line workers will first conduct manual appearance dimension detection on the strain clamps, measure parameters such as the crimping length, opposite side distance, and bending degree of the strain clamps using a vernier caliper, and evaluate whether the appearance dimensions of the strain clamps are qualified based on this. If the appearance dimensions of the strain clamp do not meet the standards, immediate remedial measures such as re-crimping need to be taken; if the appearance dimensions of the strain clamp meet the standards, internal defect detection or grip force test is considered. Thus, it can be seen that the measurement of the appearance dimensions of strain clamps is particularly important, which is the first and most crucial step in verifying whether the crimping quality is qualified.

[0005] Currently, there are obvious deficiencies in manual appearance dimension detection. On the one hand, due to the influence of factors such as manual operation error and subjective judgment when measuring with a vernier caliper, multiple measurements are often required for confirmation. Even the results of multiple measurements by the same person will deviate, and it is difficult to guarantee the accuracy of the measurement results. On the other hand, repeated measurement confirmation is time-consuming and laborious, and the measurement data needs to be manually copied, resulting in problems such as cumbersome recording, difficult format unification, and difficult digital preservation of results. In addition, the appearance dimension parameters of the strain clamp can only be displayed in a one-dimensional or two-dimensional form, lacking an ideal visualization carrier and showing the tension difference.

[0006] Therefore, there is an urgent need to implement a convenient and fast detection method for strain clamps to achieve high-efficiency, high-precision, and digital detection of the appearance dimensions of strain clamps and visualize the detection results. Summary of the Invention

[0007] The purpose of the present invention is to provide a detection method for strain clamps of transmission lines to overcome the defects existing in the above-mentioned prior art.

[0008] The purpose of the present invention can be achieved through the following technical solutions:

[0009] A detection method for strain clamps of transmission lines, the method comprising the following steps:

[0010] 1) Obtain the spatial point cloud data of the standard strain clamp and the strain clamp to be measured, convert it into the form of a.txt file, and establish the point cloud models of the standard strain clamp and the strain clamp to be measured;

[0011] 2) Use the statistical filtering algorithm to denoise the point cloud models of the standard strain clamp and the strain clamp to be measured, and use the grid refinement method based on curvature features to simplify the point cloud models of the standard strain clamp and the strain clamp to be measured, thereby completing the quality optimization of the point cloud models;

[0012] 3) Use the RANSAC algorithm to achieve clustering segmentation of the point cloud model of the standard strain clamp, thereby dividing the point cloud model into a crimping area model and a non-crimping area model;

[0013] 4) Use the NURBS surface fitting method to fit the point cloud data of the standard strain clamp to realize the three-dimensional model reconstruction of the standard strain clamp and obtain the digital model of the standard strain clamp with high precision;

[0014] 5) According to the characteristic surface equation of the strain clamp, realize the registration between the point cloud model of the measured strain clamp and the point cloud model of the standard strain clamp, calculate the characteristic distance between the point cloud model of the measured strain clamp and the point cloud model of the standard strain clamp, and complete the quantitative calculation and analysis of the model deviation value.

[0015] Furthermore, a handheld lidar scanner is used to obtain the spatial point cloud data of the standard strain clamp and the strain clamp to be measured.

[0016] Furthermore, the specific implementation of using the statistical filtering algorithm to denoise the point cloud model is as follows:

[0017] 201) For any target point cloud p i , determine the set P of point cloud data within its r neighborhood i, the surface equation of the r-neighborhood is calculated using the quadratic surface fitting method; according to the first fundamental form of the surface equation, the first fundamental quantities E, F, and G are calculated; according to the second fundamental form of the surface equation, the second fundamental quantities L, M, and N are calculated; using the first fundamental quantities and the second fundamental quantities, the average curvature H of the target point cloud p i is calculated i :

[0018]

[0019] (202) Perform statistical analysis on the average curvature H of the strain clamp point cloud i to determine the curvature judgment threshold H for point cloud data classification T :

[0020]

[0021] where represents the mean value of the point cloud average curvature; N represents the total number of data points in the point cloud model; σ H represents the standard deviation of the point cloud average curvature;

[0022] (203) Classify the point cloud data according to the curvature judgment threshold H T : For any point cloud p i , if its average curvature H i is less than the break threshold H T , it is defined as the point cloud of the flat area and retained; if its average curvature H i is greater than the break threshold H T , it is defined as the point cloud of the steep area and removed.

[0023] Furthermore, the grid-based reduction method based on curvature features is specifically as follows: For the point cloud P of the flat area A , set the grid side length to a; correspondingly, the three-dimensional space range where the point cloud P A is located is divided into m three-dimensional grids of a×a×a; for the point cloud P of the flat area B , set the grid side length to b; correspondingly, the three-dimensional space range where the point cloud P B is located is divided into n three-dimensional grids of b×b×b; where, the grid side length a > the grid side length b; a total of m + n three-dimensional grids Q i are obtained, and each three-dimensional grid Q i contains multiple point cloud data q ij =(x ij ,y ij ,z ij ); for the point cloud data in any three-dimensional grid, perform an averaging operation:

[0024]

[0025] Where k represents the three-dimensional grid Q i in the point cloud data volume.

[0026] Furthermore, the specific implementation of clustering segmentation using the RANSAC algorithm is as follows:

[0027] 301) Randomly select 6 data points from the original point cloud data P of the standard strain clamp;

[0028] 302) According to the three-dimensional coordinate values of the randomly selected 6 data points, use the least squares method to calculate the corresponding spatial cylindrical surface model equation;

[0029] 303) Traverse any point p in the original point cloud data P of the standard strain clamp i , calculate the distance from point p i to the spatial cylindrical surface: compare the distance d i with the distance judgment threshold d t , if d i < d t , define point p i as an inlier of the spatial cylindrical surface; if d i > d t , then define point p i as an outlier of the spatial cylindrical surface; count the number of inliers of the spatial cylindrical surface;

[0030] 304) Increment the iteration count by one and go to step 301), until the set maximum iteration count is reached, then go to step 305);

[0031] 305) Select the spatial cylindrical surface with the largest number of inliers as the fitting model for the non-crimping area of the standard strain clamp, and the point cloud data contained in this model is the point cloud data of the non-crimping area; the remaining point cloud data is the point cloud data of the crimping area.

[0032] Furthermore, the three-dimensional model reconstruction of the standard strain clamp is specifically as follows:

[0033] 401) For the point cloud model of the standard strain clamp, set an appropriate number of axial layers c, and cut the point cloud model into c point cloud layers {S1, S2,..., S c} along the central axis direction of the standard strain clamp;

[0034] 402) For any point cloud layer S i , select n + 1 data points q i (i = 0, 1, 2,..., n) as the value data points to be fitted; and analyze them with a cubic NURBS curve to determine the curve degree k = 3;

[0035] 403) Adjust the value data points to the parameter form required by the NURBS curve using the cumulative chord length parameterization method;

[0036] 404) According to the built-in knots u0, u1, …, u n and the curve degree k = 3, obtain the specific knot vector U;

[0037] 405) According to the knot vector U, use the deBoor-Cox recurrence equation to solve the B-spline basis function B i,k (u);

[0038] 406) For a cubic NURBS curve, there is the following relationship among the value data points, control vertices, and B-spline basis functions:

[0039]

[0040] where P(u) represents the cubic NURBS interpolation curve function, and D j (j = 0, 1, 2, …, n + 2) represents the control vertices;

[0041] 407) Set the weight factor w of the control vertices i = 1, substitute all the calculated parameter values into the following formula, and the cubic NURBS interpolation curve function P(u) passing through the value numerical points q i (i = 0, 1, 2, …, n) can be obtained;

[0042]

[0043] 408) Perform curve fitting on each layer of point clouds, perform surface interpolation based on all the obtained fitting curves, and finally realize the three-dimensional model reconstruction of the strain clamp.

[0044] Furthermore, the cumulative chord length parameterization method is specifically as follows: Define the total chord length d of the curve according to the value data points, and the total chord length d is equal to the sum of the chord lengths between consecutive points; Calculate the built-in knot corresponding to any value data point q i corresponding to

[0045] Furthermore, the number of control vertices is two more than the number of value data points. Solving the control vertices requires the following two additional conditions:

[0046] Condition 1 is that the first and last endpoints of the control vertices coincide with the first and last endpoints of the value data points;

[0047] Condition 2 is the consistency of the tangent vector condition, that is:

[0048]

[0049] Further, step 5) is specifically as follows:

[0050] 501) Use the Harris algorithm to extract the corner points p i (i = 1, 2, 3) of the point cloud model of the tension clamp to be measured as feature points; subsequently, use the curvature value as the similarity criterion to find the corresponding point q i matching p i (i = 1, 2, 3) in the three-dimensional reconstruction model of the standard tension clamp;

[0051] 502) Given the point cloud data W p of the tension clamp to be measured, respectively obtain the point set representation W q in the coordinate system of the three-dimensional reconstruction model through rotation and translation transformations;

[0052] 503) The point set representation W q in the coordinate system of the three-dimensional reconstruction model can be denoted as {(x i , y i , z i )|i = 1, 2,..., N}; for any data point in the point set representation W q , the minimum distance E i from this data point to the three-dimensional reconstruction model of the standard tension clamp can be calculated,

[0053]

[0054] where (x i , y i , z i ) represents the three-dimensional coordinate value of the point to be measured; (x i-min , y i-min , z i-min ) represents the three-dimensional coordinate value of the data point closest to the point to be measured;

[0055] 504) Define the minimum distance E i as the structural dimension deviation value of the tension clamp, and compare it with the structural deviation threshold E max to achieve the quantitative analysis of the structural dimensions.

[0056] Further, step 504) is specifically as follows: If E i ≤ E max , it is considered that the structural dimensions at this data point meet the detection requirements; if E i > E max , it is considered that the structural dimensions at this data point do not meet the detection requirements.

[0057] Compared with the prior art, the present invention has the following beneficial effects:

[0058] 1. The present invention uses the NURBS surface fitting method to fit the point cloud data of the strain clamp, solves the defects of the traditional surface fitting method in fitting accuracy and fitting efficiency, can realize the high-precision reconstruction of the three-dimensional model of the strain clamp, and helps to improve the accuracy of the measurement of the appearance size parameters and the calculation of the structural deviation; the present invention realizes the automatic measurement of the structural size parameters of the strain clamp, visualizes the measurement results, and can help the operators accurately obtain the appearance size parameters of the strain clamp in the first time, effectively improving the detection efficiency and reliability of the crimping quality of the strain clamp.

[0059] 2. The present invention adopts a grid-based reduction method based on curvature features, which can adaptively adjust the point cloud reduction rate according to the curvature attributes of the point cloud, and while reducing the amount of point cloud data, retains the characteristic information of the point cloud data to the greatest extent; at the same time, it can also solve the problem of uneven point cloud distribution.

[0060] 3. The present invention takes the geometric shape of the aluminum tube of the strain clamp as prior knowledge, and uses the RANSAC algorithm to realize the clustering segmentation of the point cloud model of the strain clamp, effectively improving the segmentation efficiency of the point cloud model of the strain clamp on the premise of ensuring the segmentation accuracy of the point cloud model. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 is a schematic diagram of the method flow of the present invention;

[0062] Figure 2 is a flowchart of the grid-based reduction method based on curvature features of the present invention;

[0063] Figure 3 is a flowchart of the RANSAC algorithm for realizing the segmentation of the point cloud model of the strain clamp of the present invention;

[0064] Figure 4 is a flowchart of the NURBS surface fitting method of the present invention for realizing the three-dimensional model reconstruction of the strain clamp. DETAILED DESCRIPTION OF THE INVENTION

[0065] The present invention will be described in detail below with reference to the drawings and specific embodiments. This embodiment is implemented on the premise of the technical solution of the present invention, and gives the detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0066] As Figure 1 shown, a detection method for a strain clamp of a transmission line, the method includes the following steps:

[0067] 1) Use a high-precision handheld lidar scanner to obtain the spatial point cloud data of the standard strain clamp and the strain clamp to be measured, and convert it into the form of a.txt file to establish the point cloud models of the standard strain clamp and the strain clamp to be measured.

[0068] 2) Implement denoising of the point cloud models of the standard strain clamp and the to-be-tested strain clamp using a statistical filtering algorithm, and implement simplification of the point cloud models of the standard strain clamp and the to-be-tested strain clamp using a grid-based reduction method based on curvature features, so as to complete the quality optimization of the point cloud models.

[0069] Affected by subjective and objective factors such as human operation and external environment, the point cloud data of the strain clamp inevitably has problems such as noise point interference and uneven density distribution. In order to ensure the accuracy of model reconstruction, it is necessary to optimize the quality of the strain clamp point cloud. As Figure 2 shown, first, use a statistical filtering algorithm to remove the noise points in the point cloud model; second, use a grid-based reduction method based on curvature features to simplify the point cloud model, and the specific steps are as follows:

[0070] 201) For any target point cloud p i , determine the set P i of the point cloud data within its r-neighborhood, and calculate the surface equation of the r-neighborhood using the quadratic surface fitting method; according to the first fundamental form of the surface equation, calculate the first fundamental quantities E, F, G; according to the second fundamental form of the surface equation, calculate the second fundamental quantities L, M, N; using the first fundamental quantities and the second fundamental quantities, the mean curvature H i of the target point cloud p i can be calculated:

[0071]

[0072] 202) Conduct statistical analysis on the mean curvature H i of the strain clamp point cloud, and thereby determine the curvature judgment threshold H T for classifying the point cloud data:

[0073]

[0074] In the formula represents the mean value of the point cloud mean curvature; N represents the total number of data points in the point cloud model; σ H represents the standard deviation of the point cloud mean curvature;

[0075] 203) Classify the point cloud data according to the curvature judgment threshold H T : For any point cloud p i , if its mean curvature H i is less than the judgment threshold H T , then define it as the point cloud of the gentle area and retain it; otherwise, define it as the point cloud of the steep area and remove it;

[0076] (204) The curvature differential reduction of the point cloud data in the two regions is carried out by using the uniform grid method. For the point cloud P in the flat region A , the grid side length is set to a; correspondingly, the three-dimensional space range where the point cloud P A is located is divided into m three-dimensional grids of a×a×a. For the point cloud P in the flat region B , the grid side length is set to b; correspondingly, the three-dimensional space range where the point cloud P B is located is divided into n three-dimensional grids of b×b×b. Among them, the grid side length a > the grid side length b. Thus, a total of m + n three-dimensional grids Q i are obtained, and each three-dimensional grid Q i contains a number of point cloud data q ij =(x ij , y ij , z ij ). For the point cloud data in any three-dimensional grid, an averaging operation is performed.

[0077]

[0078] where k represents the amount of point cloud data in the three-dimensional grid Q i .

[0079] (3) The RANSAC algorithm is used to realize the clustering segmentation of the point cloud model of the standard strain clamp, so as to divide the point cloud model into a crimping area model and a non-crimping area model.

[0080] Different regions of the strain clamp correspond to different geometric appearances. In order to ensure the accuracy of the reconstruction of the standard strain clamp model, it is necessary to perform differential model fitting processing on the point cloud data in different regions of the standard strain clamp. As Figure 3 shown, considering that the aluminum tube part of the standard strain clamp is a regular cylinder, the present invention uses the RANSAC algorithm to realize the segmentation of the point cloud model of the standard strain clamp. The specific steps are as follows:

[0081] (301) Randomly select 6 data points from the original point cloud data P of the standard strain clamp;

[0082] (302) According to the three-dimensional coordinate values of the randomly selected 6 data points, use the least square method to calculate the corresponding spatial cylindrical surface model equation:

[0083]

[0084] where r represents the radius of the cylindrical surface, and the parameters l, m, n correspond to the central axis of the cylindrical surface;

[0085]

[0086] (303) Traverse any point p in the original point cloud data P of the standard strain clamp, i calculate the distance from point p i to the spatial cylindrical surface:

[0087]

[0088] Compare the distance d i with the distance judgment threshold d t ; if d i < d t , define point p i as an interior point of the spatial cylindrical surface; otherwise, define point p i as an exterior point of the spatial cylindrical surface, and count the number N p of interior points of the spatial cylindrical surface;

[0089] (304) Increment the iteration count by one and repeat the above steps until the set maximum iteration count N max is reached;

[0090] (305) Select the spatial cylindrical surface with the largest number N p of interior points as the fitting model of the aluminum tube in the non-crimping area of the standard strain clamp. The point cloud data included in this model is the point cloud data of the aluminum tube in the non-crimping area; the remaining point cloud data is the point cloud data of the crimping area.

[0091] (4) Use the NURBS surface fitting method to fit the point cloud data of the standard strain clamp to realize the three-dimensional model reconstruction of the standard strain clamp and obtain a high-precision digital model of the standard strain clamp.

[0092] The NURBS surface can accurately represent quadratic regular curve surfaces, and can further use a unified mathematical model to describe regular curve surfaces and free curve surfaces, with the advantages of good intuitiveness, simple calculation, and high precision. As Figure 4 shown, the present invention uses the NURBS surface fitting method to realize the three-dimensional model reconstruction of the standard strain clamp. The specific steps are as follows:

[0093] (401) For the point cloud model of the standard strain clamp, set an appropriate number c of axial layers, and divide the point cloud model into c point cloud layers {S1, S2,..., S c} along the central axis direction of the standard strain clamp; the more the number c of layers, the more complete the surface information; of course, the number c of layers cannot be too large to prevent the number of data points in the point cloud layer from being too small and causing the fitting to fail;

[0094] (402) For any point cloud layer S i , select n + 1 data points q i(i = 0, 1, 2, …, n) are the data points of the form values to be fitted; and cubic NURBS curves are used for analysis to determine that the curve degree k = 3.

[0095] 403) Use the cumulative chord length parameterization method to adjust the form value data points to the parameter form required by the NURBS curve; first, define the total chord length d of the curve according to the form value data points,

[0096]

[0097] The total chord length d is equal to the sum of the chord lengths between consecutive points; then, calculate the built-in knot corresponding to any form value data point q i corresponding to

[0098]

[0099] 404) According to the built-in knot and the curve degree k = 3, the specific knot vector U can be obtained,

[0100]

[0101] where each knot value in the knot vector U can be determined by the following formula:

[0102]

[0103] 405) According to the knot vector U, use the deBoor-Cox recurrence equation to solve the B-spline basis function B i,k (u),

[0104]

[0105]

[0106] For the k-th order B-spline basis function, it is linearly combined by two k-1 order B-spline basis functions. To prevent the denominator from being 0, it is agreed that 0 / 0 = 0;

[0107] 406) For the cubic NURBS curve, there is the following relationship between the form value data points, control vertices and B-spline basis functions,

[0108]

[0109] where P(u) represents the cubic NURBS interpolation curve function; D j(j = 0, 1, 2, …, n + 2) represents the control vertices. Given the known value data points and B-spline basis functions, the control vertices used to fit the NURBS curve can be obtained by inverse calculation; the number of control vertices is two more than the number of value data points, so additional relational equations are required to solve for the control vertices; the first condition is that the first and last endpoints of the control vertices coincide with the first and last endpoints of the value data points,

[0110] D0 = q0, D n+2 = q n

[0111] The second condition is the consistency of the tangent vector condition,

[0112]

[0113] From this, the matrix form of the control vertices can be determined, and all control nodes D of the cubic NURBS curve can be solved accordingly j (j = 0, 1, 2, …, n + 2).

[0114]

[0115] 407) Set the weight factor w of the control vertices i = 1, substitute all the calculated parameter values into the following formula, and the cubic NURBS interpolation curve function P(u) passing through the value numerical points q i (i = 0, 1, 2, …, n) can be obtained;

[0116]

[0117] 408) Perform curve fitting on each stratified point cloud, and perform surface interpolation based on all the obtained fitting curves to finally realize the three-dimensional model reconstruction of the standard strain clamp.

[0118] 5) According to the characteristic surface equation of the strain clamp, realize the registration between the point cloud model of the measured strain clamp and the point cloud model of the standard strain clamp, calculate the characteristic distance between the point cloud model of the measured strain clamp and the point cloud model of the standard strain clamp, and complete the quantitative calculation and analysis of the model deviation value.

[0119] In engineering practice, the crimping of the strain clamp is all realized manually, so the structural dimensions of the strain clamp will all be different. To quantify this structural dimension deviation, the point cloud model of the strain clamp to be measured can be compared with the three-dimensional reconstruction model of the standard strain clamp, so as to obtain the quantitative value of the strain clamp model deviation. The specific steps are as follows:

[0120] 501) Use the Harris algorithm to extract the corner points p of the point cloud model of the strain clamp to be measured i(i = 1, 2, 3) as feature points; subsequently, using the curvature value as the similarity criterion, find the corresponding point q i matching p i (i = 1, 2, 3) in the three-dimensional reconstruction model of the standard strain clamp; thus, three pairs of feature points (p i , q i ) can be determined, where i = 1, 2, 3. Centering process is performed on the above data points respectively,

[0121]

[0122] and define the fundamental matrix H based on this,

[0123]

[0124] Construct matrix HH using the fundamental matrix H T , and perform eigenvalue decomposition on this matrix to obtain the eigenvalues λ0, λ1, λ2 of matrix HH T and their corresponding three eigenvectors v0, v1, v2; thus, the initial rotation matrix R and translation matrix T can be calculated.

[0125]

[0126] 502) Given the point cloud data W p of the strain clamp to be measured, after rotation and translation transformations respectively, obtain the point set representation W q in the coordinate system of the three-dimensional reconstruction model.

[0127] W q = R·W p + T

[0128] 503) The point set representation W q in the coordinate system of the three-dimensional reconstruction model can be denoted as {(x i , y i , z i )|i = 1, 2, …, N}; for any data point in the point set representation W q , the minimum distance E i from this data point to the three-dimensional reconstruction model of the standard strain clamp can be calculated.

[0129] In the formula, (x i , y i , z i ) represents the three-dimensional coordinate values of the point to be measured; (x i-min , y i-min , z i-min ) represents the three-dimensional coordinate values of the data point closest to the point to be measured;

[0130] 504) Define the minimum distance E i as the structural dimension deviation value of the strain clamp, and compare it with the structural deviation threshold E max to achieve quantitative analysis of the structural dimensions. If E i ≤E max , it is considered that the structural dimensions at this data point meet the detection requirements; otherwise, it is considered that the structural dimensions at this data point do not meet the detection requirements.

[0131] The preferred specific embodiments of the present invention have been described in detail above. It should be understood that those of ordinary skill in the art can make many modifications and variations based on the concept of the present invention without creative efforts. Therefore, all technical solutions that can be obtained by those skilled in the art in the technical field based on the concept of the present invention through logical analysis, reasoning, or limited experiments on the basis of the prior art should be within the protection scope determined by the claims.

Claims

1. A detection method for a strain clamp of a transmission line, characterized in that The method described above includes the following steps: 1) Obtain the spatial point cloud data of the standard strain clamp and the strain clamp to be measured, convert it into the form of a.txt file, and establish the point cloud models of the standard strain clamp and the strain clamp to be measured; 2) Use the statistical filtering algorithm to denoise the point cloud models of the standard strain clamp and the strain clamp to be measured, and use the grid simplification method based on curvature features to simplify the point cloud models of the standard strain clamp and the strain clamp to be measured, so as to complete the quality optimization of the point cloud models; 3) Use the RANSAC algorithm to achieve the clustering segmentation of the point cloud model of the standard strain clamp, so as to divide the point cloud model into a crimping area model and a non-crimping area model; 4) Use the NURBS surface fitting method to fit the point cloud data of the standard strain clamp, so as to realize the three-dimensional model reconstruction of the standard strain clamp and obtain the digital model of the standard strain clamp with high precision; 5) According to the characteristic surface equation of the strain clamp, realize the registration between the point cloud model of the measured strain clamp and the point cloud model of the standard strain clamp, calculate the characteristic distance between the point cloud model of the measured strain clamp and the point cloud model of the standard strain clamp, and complete the quantitative calculation and analysis of the model deviation value; The three-dimensional model reconstruction of the standard strain clamp described above is specifically as follows: 401) For the standard tension clamp point cloud model, set an appropriate number of axial layers c, and divide the point cloud model into c point cloud layers {S1, S2, …, S c} along the central axis direction of the standard tension clamp; 402) For any point cloud layer S i , select n + 1 data points q i as the value data points to be fitted, where i = 0, 1, 2, …, n; and analyze with a cubic NURBS curve to determine that the curve degree k = 3; 403) Use the cumulative chord length parameterization method to adjust the value data points into the parameter form required by the NURBS curve; 404) According to the built-in knots and the curve degree k = 3, the specific knot vector U is obtained; 405) Solve the B-spline basis function B i,k (u) according to the knot vector U using the deBoor-Cox recurrence equation;(u); 406) For the cubic NURBS curve, there is the following relationship between the value data points, control vertices and B-spline basis functions, where i = 3, 4, …, n + 3 where P(u) represents the cubic NURBS interpolation curve function, and D j represents the control vertex, where j = 0, 1, 2, …, n + 2; 407) Set the weight factor w of the control vertex i = 1, substitute all the calculated parameter values into the following formula, and the cubic NURBS interpolation curve function P(u) passing through the value point q i can be obtained, where i = 0, 1, 2, …, n; 408) Perform curve fitting on each stratified point cloud, perform surface interpolation on the basis of all the fitted curves, and finally realize the three-dimensional model reconstruction of the strain clamp.

2. The detection method of a strain clamp for a transmission line according to claim 1, characterized in that, Use a handheld lidar scanner to obtain the spatial point cloud data of the standard strain clamp and the strain clamp to be measured.

3. The detection method of a strain clamp for a transmission line according to claim 1, characterized in that, The specific method of using the statistical filtering algorithm to denoise the point cloud model is as follows: (201) For any target point cloud p i , determine the set of point cloud data P within its r-neighborhood i , and calculate the surface equation of the r-neighborhood using the quadratic surface fitting method; according to the first fundamental form of the surface equation, calculate the first fundamental quantities E, F, G; according to the second fundamental form of the surface equation, calculate the second fundamental quantities L, M, N; Using the first fundamental quantity and the second fundamental quantity, the average curvature H of the target point cloud p is calculated i is obtained i as follows (202) Mean curvature H of the strain clamp point cloud i Perform statistical analysis to determine the curvature judgment threshold H for point cloud data classification T : where represents the mean value of the average curvature of the point cloud; N represents the total number of data points in the point cloud model; σ H represents the standard deviation of the average curvature of the point cloud; Judge the threshold H according to the curvature T Classify the point cloud data: For any point cloud p i , if its average curvature H i is less than the break threshold H T , then define it as the point cloud of the gentle area and retain it; if its average curvature H i is greater than the break threshold H T , define it as the point cloud of the steep area and remove it.

4. The detection method of a strain clamp for a transmission line according to claim 3, characterized in that, The grid-based reduction method based on curvature features is specifically as follows: For the point cloud P in the flat area A , set the grid side length to a; correspondingly, the three-dimensional space range where the point cloud P A is located is divided into m three-dimensional grids of a×a×a; For the point cloud P in the flat area B , set the grid side length to b; correspondingly, the three-dimensional space range where the point cloud P B is located is divided into n three-dimensional grids of b×b×b; Among them, the grid side length a > the grid side length b; a total of m + n three-dimensional grids Q are divided i , and each three-dimensional grid Q i contains multiple point cloud data q ij =(x ij , y ij , z ij ); for the point cloud data in any three-dimensional grid, an averaging operation is performed: where k represents the amount of point cloud data in the three-dimensional grid Q i ​ 5. The detection method of a strain clamp for a transmission line according to claim 1, wherein The specific method of using the RANSAC algorithm to achieve clustering segmentation is as follows: 301) Randomly select 6 data points from the original point cloud data P of the standard strain clamp; 302) According to the three-dimensional coordinate values of the randomly selected 6 data points, use the least squares method to calculate the corresponding spatial cylindrical surface model equation; 303) Traverse any point p in the original point cloud data P of the standard strain clamp i , calculate the distance from point p i to the spatial cylindrical surface: Compare the distance d i with the distance judgment threshold d t , if d i < d t , define point p i as an inlier point of the spatial cylindrical surface; if d i > d t , then define point p i as an outlier point of the spatial cylindrical surface; count the number of inlier points of the spatial cylindrical surface; 304) Increment the iteration count and go to step 301), until the set maximum iteration count is reached, and then go to step 305); 305) Select the spatial cylindrical surface with the largest number of inliers as the fitting model of the non-crimping area of the standard strain clamp. The point cloud data contained in this model is the point cloud data of the non-crimping area; the remaining point cloud data is the point cloud data of the crimping area.

6. The detection method of a strain clamp for a transmission line according to claim 1, characterized in that The specific cumulative chord length parameterization method is as follows: Define the total chord length d of the curve according to the shape value data points, and the total chord length d is equal to the sum of the chord lengths between consecutive points; Calculate the built-in knot corresponding to any shape value data point q i corresponding to 7. The detection method of a strain clamp for a transmission line according to claim 1, characterized in that The number of control vertices is two more than the number of value data points. Two additional conditions are required to solve the control vertices: Condition 1 is that the start and end points of the control vertices coincide with the start and end points of the value data points; Condition 2 is the consistency of the tangent vector condition, that is:

8. A detection method for a strain clamp of a transmission line according to claim 1, characterized in that, The specific content of step 5) is as follows: 501) Extract the corner points p of the measured strain clamp point cloud model using the Harris algorithm, where i = 1, 2, 3; i as feature points; Subsequently, using the curvature value as the similarity criterion, the corresponding point q that matches p is found in the three-dimensional reconstruction model of the standard strain clamp, where i = 1, 2, 3; i i ​​ 502) Given the point cloud data W of the tension clamp to be measured p , and respectively obtain the point set representation W in the coordinate system of the 3D reconstruction model through rotation and translation transformations q ; 503) Point set representation W in the three-dimensional reconstruction model coordinate system q It can be denoted as {(x i , y i , z i ) | i = 1, 2, …, N}; For any data point in the point set representation W q , the minimum distance E between the data point and the three-dimensional reconstruction model of the standard strain clamp can be calculated i , where (x i , y i , z i ) represents the three-dimensional coordinate values of the point to be measured; (x i-min , y i-min , z i-min ) represents the three-dimensional coordinate values of the data point closest to the point to be measured; 504) Define the minimum distance E i as the structural dimension deviation value of the strain clamp, and compare it with the structural deviation threshold E max to achieve quantitative analysis of the structural dimensions.

9. The detection method of a strain clamp for a transmission line according to claim 8, characterized in that The specific step 504) is as follows: If E i ≤E max , it is considered that the structural dimension at this data point meets the detection requirements; if E i >E max , it is considered that the structural dimension at this data point does not meet the detection requirements.

Images