A method for measuring sag of power transmission line in flat area based on airborne laser radar

Abstract

The present application relates to a kind of flat area transmission line sag measurement method based on airborne laser radar, comprising: obtaining transmission line point cloud data, and pre-processing;Point cloud segmentation is carried out using the point cloud semantic segmentation model based on RandLA-Net constructed in advance, to obtain electric tower point cloud and conductor point cloud;The conductor point cloud closest to electric tower point cloud is clustered, to determine the conductor suspension point and unit span transmission line point cloud;Based on unit span transmission line point cloud, projection and straight line fitting are carried out, and multiple three-dimensional conductor point clouds are identified;Fitting curve point cloud, and the quality evaluation and determination lowest point missing conductor point cloud of three-dimensional conductor point cloud;Incomplete point cloud is reconstructed;Based on the point cloud after reconstruction, according to the situation that two conductor suspension points are the same height and two conductor suspension points are different height, the unit span transmission line sag numerical value is calculated.Compared with prior art, the present application has the advantages of improving measurement efficiency and precision.

Description

Technical Field

[0001] This invention relates to the field of power transmission technology, and in particular to a method for measuring the sag of power transmission lines in flat areas based on airborne lidar. Background Technology

[0002] In power transmission systems, the safety and stability of transmission lines are crucial for power operation. Sag is one of the important indicators for assessing the condition of transmission lines. Sag refers to the bending or drooping state of a transmission line due to its own weight and external forces. In transmission lines, due to the weight of the line itself and the influence of external factors such as wind and temperature, the line will exhibit a certain degree of curvature; this curvature is called sag. Excessive or insufficient sag can cause serious damage to transmission lines and the power system. When sag is too large, the risk of contact between the transmission line and the ground or surrounding objects increases, potentially leading to electric shock accidents or fires. Furthermore, large sag increases line vibration, affecting the stability and reliability of the power system. When sag is too small, it may cause excessive stress on the transmission line, leading to line breakage or tower tilting. Therefore, during the operation of transmission lines, sag needs to be measured regularly to ensure that it remains within a safe range.

[0003] The main methods for measuring the sag of transmission lines are: (1) Traditional manual measurement method, which uses measuring tools such as measuring ropes or rangefinders for measurement, but this method is limited by factors such as terrain and weather, and the labor cost is high; (2) Stress measurement method, which uses devices such as tension sensors or strain gauges to directly measure the tension or stress in the transmission line, and then deduce the sag of the line, but this method requires the installation of sensors on the transmission line, resulting in high engineering costs and may affect the normal operation of the transmission line; (3) Image measurement method, which uses cameras to acquire images of the line and performs measurements through image processing and computer vision technology, avoiding the safety risks of contact with the line. However, due to environmental conditions, the image quality is unstable and requires complex image processing algorithms. Its applicability is limited by line obstruction and complex terrain. (4) Laser radar measurement method: Laser radar is used to obtain three-dimensional point cloud data of transmission lines and their sag points, and the sag is measured by analyzing the point cloud data. Based on different scanning methods, laser radar measurement method is divided into ground-based laser radar measurement method and airborne laser radar measurement method. Ground-based laser radar measurement method is limited by limited coverage and high time cost, while airborne laser radar measurement method has the advantages of quickly covering a large area and obtaining elevation information.

[0004] When using lidar to measure the sag of power transmission lines, manually selecting suspension points or segmenting the transmission lines and towers using machine learning methods before selecting suspension points is often very complex. These methods require significant effort and constant adjustment of threshold parameters to adapt to different point cloud data. In contrast, the RandLA-Net-based semantic segmentation network for power transmission line point clouds can more efficiently and intelligently complete the tasks of power transmission line segmentation and suspension point determination. Furthermore, when airborne lidar scans power transmission lines, differences in reflection intensity and limitations in lidar resolution and accuracy can lead to discontinuities in the scanned conductors. If the lowest point of the conductor is missing, it will significantly affect the accuracy of the sag measurement. Summary of the Invention

[0005] The purpose of this invention is to provide a method for measuring the sag of power transmission lines in flat areas based on airborne lidar, which improves the accuracy of sag measurement.

[0006] The objective of this invention can be achieved through the following technical solutions:

[0007] A method for measuring the sag of transmission lines in flat areas based on airborne lidar includes the following steps:

[0008] The transmission line point cloud data is acquired and preprocessed, wherein the transmission line point cloud data is obtained by scanning a flat area using an airborne lidar.

[0009] Based on the preprocessed transmission line point cloud data, a pre-built point cloud semantic segmentation model based on RandLA-Net is used to segment the point cloud, and the semantic segmentation results are obtained, including tower point clouds and conductor point clouds respectively.

[0010] Cluster the conductor point cloud closest to the power tower point cloud to determine the conductor suspension point, and determine the unit span transmission line point cloud based on the conductor suspension point;

[0011] Based on the point cloud of the unit span transmission line, projection and line fitting are performed to identify multiple three-dimensional conductor point clouds in the point cloud of the unit span transmission line.

[0012] Based on the straight line fitting direction and the identified 3D traverse point cloud, a curve point cloud is fitted, and the quality of the 3D traverse point cloud is evaluated and the lowest point missing traverse point cloud is determined based on the curve point cloud.

[0013] Based on the quality assessment results and the point cloud of the missing conductor at the lowest point, the residual point cloud is reconstructed;

[0014] Based on the reconstructed point cloud, the sag value of the transmission line per unit span is calculated, depending on whether the two conductor suspension points are at the same height or at different heights.

[0015] Furthermore, the preprocessing operation includes filtering and denoising, and the filtering and denoising steps include:

[0016] Take any point in the point cloud data of the transmission line as a sampling point, and take the k nearest neighbor points of the sampling point;

[0017] Calculate the Euclidean distance d from the sampling point to its k nearest neighbors. i The mean m and standard deviation σ of the Euclidean distance are calculated using the following formulas:

[0018]

[0019] In the formula, d i For sampling point P j (x j ,y j z j ) is the Euclidean distance to the k nearest neighbors, where m is the mean and σ is the standard deviation;

[0020] Euclidean distance d i Points with a value greater than m±σ are considered outliers and removed from the transmission line point cloud data.

[0021] Furthermore, the step of determining the conductor suspension point includes:

[0022] Calculate the centroids of the highest 'a' points in each cluster as the suspension points of the conductor. The formula for calculating the centroid coordinates for a given cluster is:

[0023]

[0024] In the formula, P b The coordinates of the centroid, i.e., the position of the suspension point of the conductor, are (x... i y i , z i ( ) is the coordinate information of the point cloud in each cluster.

[0025] Furthermore, the step of determining the point cloud of a unit span transmission line includes:

[0026] Using two adjacent power towers, calculate the Euclidean distance between the conductor suspension points. The formula for calculating the Euclidean distance is as follows:

[0027]

[0028] In the formula, d1 is the suspension point P of the conductor. b0 (x0, y0, z0) and the suspension point P of the conductor b1 Euclidean distance between (x1, y1, z1);

[0029] Those with a distance less than the set value are classified as suspension points in the same group;

[0030] The point cloud of multiple conductors between two nearest suspension point groups is defined as the point cloud of a unit span transmission line.

[0031] Furthermore, the step of identifying multiple three-dimensional conductor point clouds in the point cloud of a unit span transmission line includes:

[0032] Projecting the point cloud of the unit span transmission line from the XOY direction yields a two-dimensional projected point cloud.

[0033] The RANSAC algorithm is used to fit a straight line to each wire in the two-dimensional projected point cloud, resulting in multiple fitted straight lines.

[0034] Calculate the Euclidean distance between the two-dimensional projected point cloud and the multiple fitted straight lines to achieve the identification of multiple three-dimensional conductor point clouds in the point cloud of a unit span transmission line.

[0035] Furthermore, the step of recognizing multiple three-dimensional conductor point clouds in a unit span transmission line point cloud includes:

[0036] In the two-dimensional projected point cloud, the Euclidean distance from each point to each fitted line is calculated using the following expression:

[0037]

[0038] In the formula, d2 is the Euclidean distance from each point in the two-dimensional projected point cloud to each fitted line, (x,y) represents a point in the two-dimensional projected point cloud, Ax+By+C=0 represents the fitted line, and A, B, and C are the line coefficients;

[0039] Based on the calculated Euclidean distance, the point closest to each fitted line is regarded as the two-dimensional point cloud corresponding to each fitted line, thereby realizing the identification of multiple two-dimensional traverse point clouds.

[0040] Based on the X and Y coordinate information and Z coordinate index of the two-dimensional traverse point cloud, the Z coordinate index information is restored to realize the identification of multiple three-dimensional traverse point clouds.

[0041] Furthermore, the step of fitting the point cloud curve includes:

[0042] The 3D guide point cloud is rotated along the Z-axis to the X-axis and parallel to the straight line fitting direction, wherein the rotation angle θ and rotation matrix R of the 3D guide point cloud are respectively:

[0043]

[0044] In the formula, A and B are linear coefficients;

[0045] Based on the rotation results, the rotated guideline point cloud is obtained, expressed as:

[0046] A2 = R * A1

[0047] In the formula, A2 is the rotated traverse point cloud, and A1 is the original traverse point cloud, i.e., the three-dimensional traverse point cloud before rotation.

[0048] The rotated conductor point cloud is projected onto the XOZ plane to obtain a two-dimensional point cloud B1. A quadratic term curve is then fitted onto the two-dimensional point cloud B1 using the least squares method.

[0049] Based on the fitted quadratic curve, generate curve point cloud B2 at equal intervals.

[0050] Furthermore, the steps of performing quality assessment of the 3D traverse point cloud and determining the traverse point cloud with missing lowest points include:

[0051] Calculate the distance from each point in the curve point cloud to the nearest multiple points on the rotated traverse point cloud A2, and calculate the average distance;

[0052] The relationship between the average distance and the set threshold is determined to obtain the quality assessment results of the three-dimensional guide point cloud, which include complete point cloud, slightly damaged point cloud, moderately damaged point cloud and severely damaged point cloud respectively.

[0053] Take the points with the lowest height in the two-dimensional point cloud and perform defect detection on these points. The three-dimensional traverse point cloud with detected defects is taken as the traverse point cloud with the lowest missing point.

[0054] Furthermore, the step of reconstructing the residual cloud includes:

[0055] For the point cloud with missing lowest point, the point cloud with moderately damaged points, and the point cloud with severely damaged points, the two-dimensional point cloud B1 generated by the projection of the curve point cloud B2 onto A2 in the XOZ plane direction is fused to obtain the fused point cloud A3.

[0056] In order to restore the Y coordinate information of the fused point cloud A3, the average Y coordinate of the A2 point cloud is used as the Y coordinate of the curve point cloud A3, and the fused point cloud A4 after restoring the Y coordinate information is obtained.

[0057] Rotate A4 to the coordinate system of the original traverse point cloud A1 to obtain point cloud A5. Point cloud A5 is the incomplete reconstructed point cloud, where A5 is:

[0058] A5 = R1 * A4

[0059]

[0060] In the formula, R1 is the rotation matrix and θ is the rotation angle.

[0061] Furthermore, the step of calculating the sag value of the transmission line per unit span includes:

[0062] When the two conductor suspension points are at the same height, in the incomplete reconstructed point cloud A5, the straight line equation l1 is determined based on the conductor suspension points A and B, and the distance f from the lowest point C to the straight line equation l1 is calculated, where the distance f is the sag.

[0063] When the heights of the two conductor suspension points are not the same, in the point cloud A5 after incomplete reconstruction, the equation of the straight line l1 is determined based on the conductor suspension points A and B. The vertical distance f1 from the lowest point C1 to the equation of the straight line l1 is calculated. The vertical distance f1 is the sag of the lowest point. The vertical distance from each point in A5 to l1 is calculated and the maximum value f2 is taken. The maximum value f2 is the maximum sag.

[0064] Compared with the prior art, the present invention has the following beneficial effects:

[0065] (1) This invention reduces the difficulty of selecting conductor suspension points by extracting conductor suspension points based on point cloud segmentation. By identifying multiple three-dimensional conductor point clouds and fitted curve point clouds, the quality of conductor point clouds is evaluated and the lowest missing conductor point cloud is determined. The incomplete conductor point cloud is supplemented and restored by using reconstruction methods, which effectively solves the problem of conductor incompleteness in transmission line point cloud data, thereby improving the accuracy of sag measurement.

[0066] (2) The present invention uses a rotational projection method to reconstruct the incomplete traverse point cloud, which further improves the integrity of the traverse point cloud and the accuracy of sag measurement.

[0067] (3) This invention uses non-contact airborne 3D lidar for measurement, which avoids dangerous manual measurement by personnel and provides more accurate and faster measurement results, providing strong support and technical foundation for the intelligent development of power transmission systems. Attached Figure Description

[0068] Figure 1 This is a schematic diagram of the method flow of the present invention;

[0069] Figure 2 This is a schematic diagram of the linear fitting and rotation angle calculation of the projected point cloud according to the present invention;

[0070] Figure 3 This is a schematic diagram of the point cloud of the rotating conductor according to the present invention;

[0071] Figure 4 This is a schematic diagram of the projection of the point cloud of the rotating guide wire and the curve fitting of the projected point cloud according to the present invention.

[0072] Figure 5This is a schematic diagram of the Y-coordinate recovery of the fused point cloud according to the present invention;

[0073] Figure 6 This is a schematic diagram of the reconstructed guideline point cloud of the present invention;

[0074] Figure 7 This is a schematic diagram illustrating the calculation of sag at the same suspension point height according to the present invention;

[0075] Figure 8 This is a schematic diagram illustrating the calculation of sag at different suspension point heights according to the present invention;

[0076] Figure 9 This is a schematic diagram of the transmission line reconstruction and sag measurement process of the present invention;

[0077] Figure 10 This is a visualization comparison of the semantic segmentation of point clouds of transmission lines according to the present invention;

[0078] Figure 11 This is a flowchart of the wire reconstruction process of the present invention. Detailed Implementation

[0079] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. These embodiments are based on the technical solution of the present invention and provide detailed implementation methods and specific operating procedures. However, the scope of protection of the present invention is not limited to the following embodiments.

[0080] This embodiment provides a method for measuring the sag of power transmission lines in flat areas based on airborne lidar, such as... Figure 1 As shown, the method includes the following steps:

[0081] S1. Use airborne lidar to scan the power transmission lines to complete the point cloud data acquisition of the power transmission corridor, and filter the acquired power transmission line point cloud data.

[0082] Specifically, the filtering process described in step S1 includes:

[0083] Take any point in the point cloud data as a sampling point, and take the k nearest neighbor points of the sampling point;

[0084] Calculate the sampling point P j (x j ,y j z j Euclidean distance d to k nearest neighbors i The mean m and standard deviation σ are calculated using the following formulas:

[0085]

[0086] Its sampling point P j (x j ,yj z j Points whose Euclidean distance to their k nearest neighbors is greater than m ± σ are considered outliers and are removed.

[0087] S2. After filtering the transmission line point cloud, use Cloud Compare software to assign labels to the power towers, conductors, ground, vegetation, and buildings in the transmission line point cloud, and create a transmission line point cloud semantic segmentation dataset for training the RandLA-Net point cloud semantic segmentation network model.

[0088] The point cloud semantic segmentation dataset for power transmission lines is divided into a training set, a validation set, and a test set in a 6:2:2 ratio.

[0089] S3. Based on the power tower point cloud and conductor point cloud obtained from the semantic segmentation results, cluster the conductor point cloud that is closest to the power tower point cloud, and calculate the centroid of the top 50 points in each cluster as the conductor suspension point.

[0090] For a given cluster, the formula for calculating the centroid coordinates is:

[0091]

[0092] S4. Use two adjacent power towers and their suspension points to determine the point cloud of the transmission line per unit span;

[0093] Specifically, the step of determining the point cloud of a unit span transmission line in step S4 includes:

[0094] Calculate the distance between the suspension points of the conductor, and classify the Euclidean distance d1 between two suspension points that is less than 200 into the same group of suspension points.

[0095] Two suspension points P b0 (x0,y0,z0) and P b1 The Euclidean distance between (x1, y1, z1) is:

[0096]

[0097] The point cloud of multiple conductors between two nearest suspension point groups is defined as the point cloud of a unit span transmission line.

[0098] S5. In the point cloud coordinate system, XY represents the horizontal and vertical coordinates on the plane, while Z represents the height coordinate perpendicular to the plane. By projecting the point cloud of the unit span transmission line from the XOY direction, multiple straight lines are fitted to the two-dimensional transmission line point cloud.

[0099] Specifically, the guide point cloud projection and line fitting steps in step S5 include:

[0100] Using world coordinates as a reference, the point cloud of a unit span transmission line is projected from the XOY direction, and the RANSAC algorithm is used to fit a straight line to each conductor in the two-dimensional point cloud generated by the projection.

[0101] S6. Calculate the Euclidean distance between the projected point cloud and the fitted multiple straight lines to achieve the identification of multiple conductors in the point cloud of a unit span transmission line.

[0102] Specifically, the step of identifying multiple conductors in the point cloud of a unit span transmission line in step S6 includes:

[0103] In the projected point cloud, calculate the distance d2 from each point to each line;

[0104] The Euclidean distance formula from a point in a 2D point cloud to a line is:

[0105]

[0106] Then, the point closest to each straight line is regarded as the two-dimensional point cloud corresponding to each straight line, realizing the identification of multiple wires in the two-dimensional point cloud. Finally, based on the X and Y coordinate information and Z coordinate index of the two-dimensional point cloud, its Z information is restored, realizing the identification of multiple wires in the three-dimensional point cloud.

[0107] S7. Based on the direction of the fitted straight line, rotate the point cloud of the conductor along the Z-axis until the X-axis is parallel to the direction of the straight line. Project the rotated point cloud onto the XOZ plane and then fit the curve point cloud using the least squares method. Based on the curve point cloud, evaluate the quality of the single conductor point cloud and classify it into complete point cloud, slightly incomplete, moderately incomplete, and severely defective.

[0108] For details, please refer to Figure 2 , Figure 3 and Figure 4 The step S7, which involves rotating the conductor and fitting the projected curve point cloud, includes:

[0109] Based on the slope of the fitted straight line L1, the angle of rotation of the point cloud around the Z-axis and its rotation matrix R are calculated.

[0110] The formula for calculating the rotation angle around the Z-axis is:

[0111]

[0112] The rotation matrix for rotating about the Z-axis by θ degrees is:

[0113]

[0114] The original point cloud is A1, and the rotated point cloud is A2:

[0115] A2=R*A

[0116] Project the rotated point cloud A2 onto the XOZ plane to obtain a two-dimensional point cloud B1. Then, fit a quadratic curve onto the two-dimensional point cloud B1 using the least squares method.

[0117] Generate curve point cloud B2 at equal intervals based on the fitted quadratic polynomial, and calculate the average distance from each point on curve point cloud B2 to the nearest 16 points on point cloud A2.

[0118] Set a distance threshold and classify points with an average distance of less than 20 as complete point clouds, those with an average distance of more than 20 but less than 100 as slightly damaged point clouds, those with an average distance of more than 100 but less than 200 as moderately damaged point clouds, and those with an average distance of more than 200 as severely damaged point clouds.

[0119] S8. Based on the lowest part of the two-dimensional point cloud B1, detect whether this part of the points is missing. If it is missing, define the traverse point cloud as the traverse point cloud with missing lowest point.

[0120] Take the 150 points with the lowest height in the 2D point cloud B1. If any of these points are found to be missing, then the traverse point cloud is determined to be a traverse point cloud with missing lowest point.

[0121] S9. For point clouds with missing lowest points, moderately damaged point clouds, and severely damaged point clouds, the projected point cloud B1 is fused with the curved point cloud B2, and the average value of the Y coordinate of point cloud A2 is used as the Y coordinate of the fused point cloud. Finally, the fused point cloud is rotated back to the original coordinate system to complete the reconstruction of the damaged point cloud.

[0122] For details, please refer to Figure 5 and Figure 6 The residual cloud reconstruction step described in step 9 includes:

[0123] For point clouds with missing minimum points, moderate defects, and severe defects, the fitted curve point cloud B2 is fused with the point cloud B1 projected onto the XOZ plane by A2, and the fused point cloud is A3.

[0124] To restore the Y-coordinate information of the fused point cloud A3, the average Y-coordinate of the A2 point cloud is used as the Y-coordinate of point cloud A3, and the fused point cloud after restoring the Y-coordinate information is A4;

[0125] The point cloud after rotating point cloud A4 to the coordinate system of point cloud A1 is called point cloud A5. Point cloud A5 is the point cloud after incomplete reconstruction, and its rotation matrix is ​​R1.

[0126]

[0127] The process of transforming point cloud A4 to point cloud A1 coordinate system is as follows:

[0128] A5 = R1 * A4

[0129] S10. Calculate the sag value of the transmission line per unit span based on two cases: the two suspension points are at the same height and the two suspension points are at different heights.

[0130] For details, please refer to Figure 7 and Figure 8 The sag measurement steps described in step 9 for the two cases of the same suspension point height and different suspension point heights include:

[0131] When the two suspension points are at the same height, in the point cloud A5 after incomplete reconstruction, the equation of the straight line l1 is determined based on the suspension points A and B, and the distance f from the lowest point C to the equation of the straight line l1 is calculated, where f is the sag.

[0132] When the two suspension points are at different heights, in the reconstructed point cloud A5, the equation of the straight line l1 is determined based on suspension point A and suspension point B. The vertical distance f1 from C1 to l1 is calculated, where f1 is the sag of the lowest point. The vertical distance from each point in point cloud A5 to l1 is calculated and the maximum value is taken, which is the vertical distance f2 from C2 to l1, where f2 is the maximum sag.

[0133] This invention provides verification results for the above-mentioned method for measuring the sag of power transmission lines in flat areas based on airborne lidar. Specifically, a 26.32 km section of power transmission line point cloud data from a map is extracted as the semantic segmentation dataset for power transmission line point cloud. The power transmission line reconstruction and sag measurement process is as follows: Figure 9 As shown in Table 1, the accuracy comparison of the semantic segmentation results of the transmission line point cloud is as follows: the overall segmentation accuracy is 97.39%. The visualization comparison of the semantic segmentation results of the transmission line point cloud is shown in Table 1. Figure 10 As shown, the conductor reconstruction process is as follows: Figure 11 As shown in Table 2, the results of the sag measurement experiment are shown in Table 2.

[0134] Table 1: Semantic segmentation accuracy of point clouds for transmission lines

[0135]

[0136] Table 2: Experimental Results of Sag Measurement

[0137] Single line 1 9.444 9.316 0.128 98.64％ Single line 2 9.305 9.538 -0.233 97.55％ Single line 3 9.313 9.152 0.161 98.24％ Single line 4 9.291 9.189 0.102 98.89％ Single line 5 9.269 9.385 -0.116 98.76％ Single line 6 9.266 9.393 -0.127 98.65％

[0138] During the experiment, after using this method to complete the semantic segmentation of the transmission line, the reconstruction of the point cloud of the incomplete conductor, and the sag measurement, the measured sag value was compared with the true value. The overall sag measurement accuracy reached 98.455%, and the absolute error was less than 0.233m, which greatly improved the accuracy of the sag measurement of the transmission line.

[0139] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0140] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of the present invention can be implemented using various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0141] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0142] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0143] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0144] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0145] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for measuring the sag of transmission lines in flat areas based on airborne lidar, characterized in that, Includes the following steps: The transmission line point cloud data is acquired and preprocessed, wherein the transmission line point cloud data is obtained by scanning a flat area using an airborne lidar. Based on the preprocessed transmission line point cloud data, a pre-built point cloud semantic segmentation model based on RandLA-Net is used to segment the point cloud, and the semantic segmentation results are obtained, including tower point clouds and conductor point clouds respectively. Cluster the conductor point cloud closest to the power tower point cloud to determine the conductor suspension point, and determine the unit span transmission line point cloud based on the conductor suspension point; Based on the point cloud of the unit span transmission line, projection and line fitting are performed to identify multiple three-dimensional conductor point clouds in the point cloud of the unit span transmission line. Based on the straight line fitting direction and the identified 3D traverse point cloud, a curve point cloud is fitted, and the quality of the 3D traverse point cloud is evaluated and the lowest point missing traverse point cloud is determined based on the curve point cloud. The steps of fitting the curve point cloud include: The three-dimensional guide point cloud is rotated along the Z-axis to the X-axis and parallel to the straight line fitting direction, wherein the rotation angle of the three-dimensional guide point cloud is... and rotation matrix They are respectively: In the formula, A and B are linear coefficients; Based on the rotation results, the rotated guideline point cloud is obtained, expressed as: In the formula, For the rotated conductor point cloud, The original traverse point cloud, i.e., the 3D traverse point cloud before rotation; Projecting the rotated conductor point cloud onto the XOZ plane yields a two-dimensional point cloud. Using the least squares method in two-dimensional point clouds Fit a quadratic term curve to the upper bound; Generate curve point clouds at equal intervals based on the fitted quadratic curve. ; The steps for quality assessment of 3D conductor point clouds and determination of conductor point clouds with missing lowest points include: Calculate the distance from each point in the curve point cloud to the rotated guide point cloud. Find the distances between the nearest multiple points and calculate the average distance. The relationship between the average distance and the set threshold is determined to obtain the quality assessment results of the three-dimensional guide point cloud, which include complete point cloud, slightly damaged point cloud, moderately damaged point cloud and severely damaged point cloud respectively. Take the points with the lowest height of the curve point cloud, and perform defect detection on the points. The three-dimensional guide point cloud with detected defects is taken as the guide point cloud with the lowest missing point. Based on the quality assessment results and the point cloud of the missing conductor at the lowest point, the residual point cloud is reconstructed; Based on the reconstructed point cloud, the sag value of the transmission line per unit span is calculated, depending on whether the two conductor suspension points are at the same height or at different heights.

2. The method for measuring the sag of transmission lines in flat areas based on airborne lidar according to claim 1, characterized in that, The preprocessing operation includes filtering and denoising, and the steps of the filtering and denoising process include: Take any point in the point cloud data of the transmission line as a sampling point, and take the sampling point. k One nearest neighbor point; Calculate the sampling point to k Euclidean distance between nearest neighbors Mean of Euclidean distance and standard deviation The calculation formulas are as follows: In the formula, Sampling points arrive k Euclidean distance between nearest neighbors The mean, Standard deviation; European distance Greater than Points that are considered outliers are removed from the power transmission line point cloud data.

3. The method for measuring the sag of transmission lines in flat areas based on airborne lidar according to claim 1, characterized in that, The steps for determining the conductor suspension point include: Calculate the highest value in each cluster a The centroid of each point is taken as the suspension point of the conductor. For a cluster, the formula for calculating the centroid coordinates is: In the formula, These are the coordinates of the centroid, i.e., the position of the conductor suspension point. This provides the coordinate information of the point cloud in each cluster.

4. The method for measuring the sag of transmission lines in flat areas based on airborne lidar according to claim 1, characterized in that, The steps for determining the point cloud of a unit span transmission line include: Using two adjacent power towers, calculate the Euclidean distance between the conductor suspension points. The formula for calculating the Euclidean distance is as follows: In the formula, For conductor suspension point With the suspension point of the conductor The Euclidean distance between them; Those with a distance less than the set value are classified as suspension points in the same group; The point cloud of multiple conductors between two nearest suspension point groups is defined as the point cloud of a unit span transmission line.

5. The method for measuring the sag of transmission lines in flat areas based on airborne lidar according to claim 1, characterized in that, The steps for identifying multiple three-dimensional conductor point clouds in a point cloud of a unit span transmission line include: Projecting the point cloud of the unit span transmission line from the XOY direction yields a two-dimensional projected point cloud. The RANSAC algorithm is used to fit a straight line to each wire in the two-dimensional projected point cloud, resulting in multiple fitted straight lines. Calculate the Euclidean distance between the two-dimensional projected point cloud and the multiple fitted straight lines to achieve the identification of multiple three-dimensional conductor point clouds in the point cloud of a unit span transmission line.

6. The method for measuring the sag of transmission lines in flat areas based on airborne lidar according to claim 5, characterized in that, The steps for identifying multiple three-dimensional conductor point clouds in a point cloud of a unit span transmission line include: In the two-dimensional projected point cloud, the Euclidean distance from each point to each fitted line is calculated using the following expression: In the formula, Let Euclidean distance be the distance from each point in the 2D projected point cloud to each fitted line. x,y () represents a point in a two-dimensional projected point cloud. This represents the fitted straight line, where A, B, and C are the linear coefficients. Based on the calculated Euclidean distance, the point closest to each fitted line is regarded as the two-dimensional point cloud corresponding to each fitted line, thereby realizing the identification of multiple two-dimensional traverse point clouds. Based on the X and Y coordinate information and Z coordinate index of the two-dimensional traverse point cloud, the Z coordinate index information is restored to realize the identification of multiple three-dimensional traverse point clouds.

7. The method for measuring the sag of transmission lines in flat areas based on airborne lidar according to claim 1, characterized in that, The steps for reconstructing the incomplete cloud include: For the point clouds with missing lowest points, moderately incomplete point clouds, and severely incomplete point clouds, the curve point clouds are... and Two-dimensional point cloud generated by projection in the XOZ plane direction The points are then merged to obtain the merged point cloud. ; To restore the merged point cloud The Y-coordinate information will The average value of the Y-coordinate in the point cloud is used as the point cloud value. The Y-coordinate is used to obtain the fused point cloud after restoring the Y-coordinate information. ; Will Rotate to the original guide point cloud Using the coordinate system, we obtain the point cloud. Point cloud This is the point cloud after incomplete reconstruction, in which for: In the formula, For rotation matrix, The angle is the rotation angle.

8. A method for measuring the sag of transmission lines in flat areas based on airborne lidar according to claim 7, characterized in that, The steps for calculating the sag value of a transmission line per unit span include: When the two conductor suspension points are at the same height, in the incomplete reconstructed point cloud In the above, the equation of the straight line is determined based on the suspension points A and B of the conductor. Calculate the lowest point C To the equation of the line distance The distance It is a sag; When the two conductor suspension points are at different heights, in the incomplete reconstructed point cloud In the above, the equation of the straight line is determined based on the suspension points A and B of the conductor. Calculate the lowest point To the equation of the line vertical distance The vertical distance Calculate the sag at the lowest point. Each point in the middle Calculate the vertical distance and take the maximum value. maximum value This represents the maximum sag.

Images