A method for detecting local deformation of a tower based on different-phase laser point clouds

Abstract

A kind of tower local deformation detection method based on different period laser point cloud, comprising: using three-dimensional laser scanner to scan detection overhead transmission line, according to time sequence, the tower point cloud collected in different period is divided into reference point cloud and comparison point cloud sequence 1,2,…n;Tower point cloud is sliced and layered, and the first transverse surface of tower is automatically identified according to tower point cloud structure determination criterion;With the first transverse surface of reference point cloud as reference object, using Super-4PCS coarse registration and local ICP fine registration algorithm, point cloud sequence 1,2,…n is registered to reference point cloud;The displacement of key node of point cloud sequence 1,2,…n relative to reference point cloud is extracted, compared with key node structure abnormality determination threshold, to assess whether tower structure is abnormal.This application effectively detects tower tilt, main material bending, foundation settlement and cross arm deformation and other structural abnormalities caused by icing galloping, extreme gale and extreme adverse conditions of geological disasters.

Description

Technical Field

[0001] This invention relates to the field of abnormal detection of transmission line tower structures, specifically a method for detecting local deformation of towers based on laser point clouds from different periods. Background Technology

[0002] Because power transmission lines stretch for hundreds to thousands of kilometers, covering a wide geographical area, and are exposed to harsh external environments, they are highly susceptible to extreme natural disasters. Under severe conditions such as icing, galloping, extreme winds, and geological disasters, transmission line towers can experience stress yielding due to factors such as load imbalance, exceeding design loads, and foundation displacement and settlement. This can lead to tower tilting and crossarm skewing, threatening the safe operation of the line. Since different parts of the tower experience different stresses and have varying strengths, the degree of deformation varies from part to part. In particular, crossarms and ground wire supports are prone to localized bending under heavy loads such as icing, galloping, and ice-shedding jumps. If the load continues to increase, it may cause damage to the tower head or tower collapse.

[0003] Chinese patent application CN115810012B discloses a method, device, equipment, and storage medium for detecting the tilt of transmission towers. The method includes: acquiring multiple initial point cloud data sets for the space where the transmission tower to be detected is located; solving for the geometric features of each initial point cloud data set to obtain geometric feature information corresponding to each initial point cloud data set; obtaining target point cloud data corresponding to each initial point cloud data set based on the geometric feature information; extracting the transmission tower point cloud data corresponding to the transmission tower and the ground point cloud data corresponding to the ground where the transmission tower is located from the target point cloud data; and obtaining the tilt detection result corresponding to the transmission tower based on the transmission tower point cloud data and the ground point cloud data. This method is suitable for extracting tower point clouds from transmission line point clouds and then calculating the tower tilt. However, for this non-uniform deformation, the single state quantity of tower tilt cannot effectively reflect local yield deformation, and there is a possibility of missed detection.

[0004] Chinese patent application CN115880276A discloses a method for evaluating the operational status of towers based on multi-period point cloud comparison. A three-dimensional cone is constructed based on the tower point cloud to obtain the tilt angle. The operational status of the tower is determined based on the tilt angle and multiple tower tilt degrees. The method further determines the tower tilt degree by extracting information from the front and side using two-dimensional convolution kernels and from the top using a three-dimensional convolution kernel. This allows for accurate determination of tower shape information not only from the front and side views but also by overcoming the limitations of angle-based assessment from the top. While this method can detect and evaluate the overall tilt of the tower, it struggles to reflect local displacement deformations such as those at crossarm attachment points and grounding wire supports.

[0005] In summary, existing methods primarily use 3D laser point cloud analysis to identify the overall tilt of the tower and assess its operational status. However, since the tower load is mainly concentrated at the conductor / ground wire anchorage and foundation, these critical nodes are most prone to local deformation under overload conditions. Single state parameters such as overall tower tilt and crossarm skew cannot effectively reflect this local deformation. To reflect the local deformation of the tower, this invention analyzes multi-period point cloud data to extract the displacement of key tower nodes and then measures the local deformation. Furthermore, because weak displacements and deformations such as foundation displacement and main material deformation are amplified through the tower structure, even small displacements can cause significant displacements at key nodes such as conductor / ground wire anchorages, making them easier to detect. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a method for detecting local deformation of transmission towers based on laser point clouds from different periods. This method can be used to detect structural anomalies in transmission towers caused by external loads, such as foundation displacement and settlement, bending deformation of main materials, and stress yielding of crossarms and ground wire supports.

[0007] A method for detecting local deformation of a tower based on laser point clouds from different periods includes the following steps:

[0008] Step 1: Use a 3D laser scanner to scan and inspect the overhead transmission line. According to the time sequence, the tower point clouds collected at different times are divided into reference point clouds and comparison point cloud sequences 1, 2, ... n.

[0009] Step 2: Slice and layer the tower point cloud, and automatically identify the first transverse diaphragm of the tower according to the tower point cloud structure determination criteria;

[0010] Step 3: Using the first transverse plane of the reference point cloud as the reference object, the point cloud sequence 1, 2, ... n is registered to the reference point cloud using the Super-4PCS coarse registration and local ICP fine registration algorithms;

[0011] Step 4: Extract the displacement of key nodes of point cloud sequence 1, 2, ... n relative to the reference point cloud, and compare it with the key node structural anomaly judgment threshold to evaluate whether the tower structure has an anomaly.

[0012] Furthermore, step 1 includes:

[0013] A 3D lidar was used to scan the transmission line, followed by calculation, filtering, and thinning. Point cloud classification technology was used to extract the tower point cloud, removing conductors, ground, and tree point clouds. Based on the time sequence of point cloud acquisition, the first phase of tower point cloud was set as the baseline point cloud, denoted as S. 0 The subsequent point clouds are sequentially set as alignment sequence point clouds, i.e., sequence 1 point cloud, sequence 2 point cloud, ..., sequence n point cloud, denoted as S.1 S 2 …S n .

[0014] Furthermore, step 2 includes:

[0015] Step 2.1: Slice the tower point cloud: Intersect the point cloud with a set of planes of thickness δ, and slice the point cloud from top to bottom. The slice thickness δ of the ground-scanned point cloud is 50cm, and the slice thickness δ of the machine-scanned point cloud is 80cm. Cut the tower point cloud sequentially from top to bottom along the vertical z-axis to obtain the horizontal point cloud slice S of the tower. i :

[0016]

[0017] The point cloud slices from each layer are projected onto a horizontal plane, as shown below:

[0018]

[0019] Step 2.2, Extract the circumscribed polygon boundaries: The circumscribed polygon boundaries of each slice are extracted using the "bidirectional nearest point search method" from the point cloud slice projection set. In S′, arbitrarily choose a point Ps as the starting point of the boundary line, find the point Pr that is closest to Ps, and calculate the distance ds from Pr to Ps; i In the remaining point cloud, find the nearest point Pe to Ps and calculate the distance de from Ps to Pe; if ds ≤ de, insert Pr before Ps as the new starting point, and construct a line connecting Pr, Ps, and Pe pairwise, replacing Pr with Ps as the starting point; if ds ≥ de, then grow towards the other end Pe; then determine the set S′ i Check if there are any remaining points. If so, repeat the above steps; finally, traverse the set S′. i After all points are connected, until PrPe overlaps, connect PsPe to make the polygon closed loop, and obtain the outer contour polygon boundary of the point cloud.

[0020] Step 2.3, Calculate the vertices of the circumscribed polygon: Using a linear regression equation, fit the outermost boundary point cloud into four boundary lines, denoted clockwise as AB, BC, CD, and DA; the intersection of these four lines represents the four endpoints of the tower's transverse diaphragm, with the endpoint coordinates denoted as A(x...). A ,y A ,z A ), B(x) B ,y B ,z B ), C(x) C ,y C ,z C ), D(x D ,y D,z D );

[0021] Step 2.4, calculate the aspect ratio, area ratio parameters, and number of points of the circumscribed polygon of the slice: calculate the side length l of each layer of point cloud slice. ix ,l iy Area s i Number of points n i Then calculate the area of ​​the circumscribed rectangle. and aspect ratio In the formula l ix ,l iy Projecting S onto the i-th layer slice i The length and width of the circumscribed rectangle;

[0022] Step 2.5: Based on the number of points in the slice and the aspect ratio and area ratio of its outer polygon, identify the first transverse diaphragm of the tower. The identification criterion is: when slicing from top to bottom, the slice layer where the last peak of the number of slice points is located, and the number of slice points n i The number should be greater than 500, and the height of the transverse diaphragm should be less than the tower height. The height above the ground is generally no more than 15m; the aspect ratio of the circumscribed rectangle of the cross section satisfies: slice area ratio

[0023] Furthermore, the Super-4PCS coarse registration algorithm in step 3 includes:

[0024] Step 3.1.1, on the reference point cloud S 0 Four coplanar but non-collinear points are selected, which are common points of the multi-phase point clouds, and the distance between them should be as large as possible. This is denoted as B = {P}. a P b P c P d} Calculate the distance and intersection ratio of lines AB and CD using the following formula:

[0025] d1=||P a -P b ||,d2=||P c -P d || (3)

[0026]

[0027] Step 3.1.2, according to the principle of affine invariance, in the point cloud S of sequence 1... 1 Find the set of points in the array that have the same distance and intersection ratio with lines AB and CD, as follows:

[0028] In point cloud S of sequence 1 1 Determine the set of point pairs S1 and S2, with each point qi With q1 as the center, draw spheres with radii R = d1 ± ε and R = d2 ± ε respectively. Set S1 consists of q1 and points distributed in the range [d1-ε, d1+ε], and set S2 consists of q1 and points distributed in the range [d2-ε, d2+ε]. At the same time, the point cloud surface is rasterized to establish a three-dimensional mesh G with a unit size of ε.

[0029] Step 3.1.3, in the point cloud S of sequence 1 1 Extracting from the baseline point cloud S 0 The corresponding four-point set: Traverse all candidate point pairs in point pair sets S1 and S2, calculate all intersection nodes e according to the cross ratio consistency, and store them in the grid G, based on the reference point cloud S. 0 The angle between two corresponding point pairs is θ. In the grid G, the nodes are extracted to be approximately equal, and the angle between the lines connecting the two point pairs is approximately equal to θ. The four-point basis is matched within the error range of ζ.

[0030] Step 3.1.4, in the point cloud S of sequence 1 1 All points in the reference point cloud S that meet the conditions 0 The corresponding four-point set U = {U1, U2, ..., U...} n}, calculate the baseline point cloud S 0 With each U i The transformation matrices between the common point set (LCP) are compared, and the transformation matrix with the highest registration accuracy is selected for global transformation.

[0031] Furthermore, the local ICP fine registration algorithm in step 3 includes: for point cloud S 0 and S 1 From the baseline point cloud S 0 Select reference point p i From point cloud S 1 Find the distance point p in the middle i The point q with the shortest Euclidean distance i , with p i and q i The transformation matrix is ​​calculated for the corresponding point set. This process is iterated continuously, using Equation 5 as the objective error function. The iteration stops when the value of f is less than a threshold, thus obtaining the optimal transformation matrix.

[0032]

[0033] Where R represents the rotation transformation matrix, T represents the translation transformation matrix, and k represents the number of corresponding points.

[0034] Furthermore, step 4 includes:

[0035] Step 4.1, Point Cloud Segmentation: Segment the point cloud S separately. 0 and S1 Using principal component analysis, the principal direction feature vector of the two-dimensional projected point cloud is obtained. Based on this, the principal direction rotation angle is calculated, and the tower point cloud is rotated to obtain the redirected point cloud data. The direction perpendicular to the rotated principal direction on the horizontal plane is selected as the horizontal direction, and the point cloud is cut along the horizontal direction. The cut point cloud is divided into left and right parts.

[0036] Step 4.2, Scattered Point Cloud Indexing: Index the baseline point cloud S 0 And Sequence 1 point cloud S 1 Use the Octree method to create a scattered point cloud index;

[0037] Step 4.3, Euclidean Clustering Segmentation: This involves segmenting the baseline point cloud S based on the established Octree index. 0 And Sequence 1 point cloud S 1 Euclidean clustering is performed on the insulators to distinguish the phases of the insulators. Each tower point cloud is set as P, which consists of n points. The Euclidean distance between the i-th sample and the j-th sample is calculated. The points with the smallest distance are grouped into one class. Then, the Euclidean distance between the new classes is calculated and iterated until a specified threshold is less than the Euclidean distance between any two classes or the number of classes is less than a specified number. The Euclidean clustering is then completed.

[0038] Step 4.4, Key Node Location Extraction: Extract the point cloud S for each period from the clustered point cloud. 0 and S 1 Calculate the three-dimensional coordinates of the crossarm mounting points of each phase insulator and the mounting points of the left and right phase ground wires, and then calculate the point cloud S. 0 and S 1 The displacement of key nodes at the same location along the x, y, and z axes;

[0039] Step 4.5, Structural Anomaly Judgment: Compare the displacement of each key node with the key node structural anomaly judgment threshold to determine whether the key nodes of the multi-phase point cloud have displaced, and then determine whether the tower structure has deformed.

[0040] Furthermore, the specific implementation process of step 4.3 is as follows:

[0041] Step 4.3.1: Establish an Octree data structure for each period's tower point cloud P;

[0042] Step 4.3.2: Create an empty cluster set C and a queue;

[0043] Step 4.3.3, for each point P i ∈P, perform the following steps:

[0044] (1) Add the point to the current queue Q;

[0045] (2) For each point P i ∈Q, perform the following operations: ① Using the search radius r < d th , for P i Perform a k-nearest neighbor search, and the searched points are: ②For each calculate With P i The two closest Euclidean distances are grouped into the same class; ③ Check each P i Check if the point has performed the above operations. If not, add the point to queue Q.

[0046] (3) When all points in the list of queue Q have performed the above operations, add the points in Q to the list of set C and clear the list of Q;

[0047] (4) When each point P i ∈P performs the above operation, and P i If all elements are part of set C, the algorithm terminates.

[0048] Compared with the prior art, the beneficial effects of the present invention are:

[0049] (1) Compared with the prior art, the present invention solves the problem that the prior art is difficult to determine local structural anomalies of towers. The present invention extracts the relative displacement of the key nodes of the tower by comparing the three-dimensional coordinates of the key nodes of the tower at different times, and compares them with the threshold for determining the structural anomalies of the key nodes, so as to detect whether the tower has undergone local structural deformation in a timely manner.

[0050] (2) Compared with existing technologies, this invention can quickly detect local deformation of crossarms and ground wire supports caused by icing and extreme winds. Existing technologies mainly use tower tilt to characterize structural anomalies, which cannot reflect local deformation of crossarms and ground wire supports. This invention uses three-dimensional laser scanning technology to perform multiple scans of the tower before and after experiencing extreme conditions, effectively obtaining the three-dimensional coordinates of key nodes at different times, and determining whether local deformation has occurred in the crossarms and ground wire supports. Attached Figure Description

[0051] Figure 1 This is a flowchart of the Super-4PCS coarse registration technology used in this invention;

[0052] Figure 2 This is a flowchart of the local ICP fine registration technology used in this invention;

[0053] Figure 3 This is a flowchart of the technology for extracting the relative displacement of key nodes of towers used in this invention;

[0054] Figure 4 These are the two point clouds to be registered in this embodiment of the invention;

[0055] Figure 5 This is a schematic diagram of the first transverse diaphragm of the tower extracted in an embodiment of the present invention;

[0056] Figure 6 This is a schematic diagram of two-stage point cloud registration in an embodiment of the present invention;

[0057] Figure 7 This is a schematic diagram illustrating the orientation and rotation of the tower using principal component analysis in an embodiment of the present invention;

[0058] Figure 8 This is a schematic diagram of insulator clustering analysis in an embodiment of the present invention;

[0059] Figure 9 This is a schematic diagram of displacement extraction of key nodes of the tower in an embodiment of the present invention. Detailed Implementation

[0060] The technical solutions of the present invention will now be clearly and completely described with reference to the accompanying drawings.

[0061] This invention first employs a 3D laser scanner to perform two scans on a potentially hazardous tower, dividing the two point clouds into a baseline point cloud and a comparison point cloud. The comparison point cloud is then registered to the baseline point cloud using the registration algorithm proposed in this invention, and the relative displacement at the conductor / ground wire attachment point is extracted. The specific implementation steps are as follows.

[0062] Step 1: Two scans were performed on a certain hazardous pole to obtain two phases of point cloud data for the pole. See details below. Figure 4 .

[0063] Step 2: According to Figure 1 The process shown involves slicing the point cloud data of the tower. A set of planes with a thickness of δ = 80 cm are intersected with the point cloud to slice it, and the first transverse plane of the tower point cloud is extracted. See details in [link to documentation]. Figure 5 .

[0064] Step 3: According to Figure 2 The process shown employs Super-4PCS coarse registration and local ICP algorithms to register the Sequence 1 point cloud of the tower to the reference point cloud. See details below. Figure 6 Specifically, the first transverse diaphragm of the tower is selected as the registration reference, and the Super-4PCS coarse registration technique is used. See details... Figure 1 The sequence of point cloud S of the tower 1 Registered to reference point cloud S 0 Then, a local ICP algorithm is used for fine registration, see details. Figure 2 The sequence of point cloud S of the tower 1 Registered to reference point cloud S 0 .

[0065] Step 4, according to Figure 3 The process shown extracts the relative displacements of key tower nodes (ground wire attachment point and insulator crossarm side attachment point). First, principal component analysis is used to orient and rotate the tower, such as... Figure 7 Then, cluster the insulators, such as... Figure 8 The three-dimensional coordinates of key nodes such as the crossarm mounting point of the tower conductor and ground wire are obtained by extracting the coordinates of the closest point between the insulator and the tower. See details in [link to documentation]. Figure 9 Calculate the point cloud S 0 and S 1 The displacements of key nodes at the same location along the x, y, and z axes are shown in Table 1.

[0066] Table 1. Displacement of key nodes of the tower

[0067] Key Nodes Δx(m) Δy(m) Δz(m) Δd(m) threshold Is it abnormal? Point #1 -41.2 12.1 -56.2 70.73 0.25 yes Point #2 2.5 3.6 0.4 4.40 0.23 no Point #3 -3.5 3.7 -5.6 7.57 0.23 no Point #4 2.4 2.6 -5.9 6.88 0.19 no Point #5 -1.3 3.2 2 3.99 0.19 no Point #6 0.2 -2.7 2.6 3.75 0.15 no Point #7 -1.1 2.9 -2.1 3.75 0.15 no Point #8 -6.4 -3.7 -6.8 10.04 0.25 no Point #9 9.8 5.4 8.9 14.30 0.23 no Point #10 -7.9 2.1 2.4 8.52 0.23 no Point #11 2.5 2.6 -6.8 7.70 0.19 no Point #12 -7.8 3.6 9.8 13.03 0.19 no Point #13 7.4 -2.9 -2.1 8.22 0.15 no Point #14 5.2 5.1 -4.7 8.67 0.15 no

[0068] Step 5: The threshold for determining structural anomalies at key tower nodes is set to (h × 5‰), where h is the height of each node relative to the tower base. The displacement of the key tower nodes is compared with the threshold for determining structural anomalies. It is found that the ground wire support on the left side of the tower (Point #1) exceeds the threshold (see Table 1), and is therefore determined to be structurally abnormal. On-site inspection revealed that the ground wire support on the left side of the tower was bent and deformed, verifying the effectiveness and feasibility of this method.

[0069] For multi-phase point clouds, repeat steps 2-4 respectively, and process the sequence point cloud S. 2 ,S 3 ,…S n The data is sequentially registered to the reference point cloud, and then the displacement of the key nodes of the tower is extracted and compared with the judgment threshold to determine whether the tower has experienced structural abnormalities.

[0070] Compared with the prior art, the beneficial effects of the present invention are:

[0071] (1) Compared with the prior art, the present invention solves the problem that the prior art is difficult to determine local structural anomalies of towers. The present invention extracts the displacement of the key nodes of the tower by comparing the three-dimensional coordinates of the key nodes of the tower at different times, and compares them with the threshold for determining the structural anomalies of the key nodes, so as to detect whether the tower has undergone local structural deformation in a timely manner.

[0072] (2) Compared with existing technologies, this invention can quickly detect local deformation of crossarms and ground wire supports caused by icing and extreme winds. Existing technologies mainly use tower tilt to characterize structural anomalies, which cannot reflect local deformation of crossarms and ground wire supports. This invention uses three-dimensional laser scanning technology to perform multiple scans of the tower before and after experiencing extreme conditions, effectively obtaining the three-dimensional coordinates of key nodes at different times, and determining whether local deformation has occurred in the crossarms and ground wire supports.

[0073] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied 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.

[0074] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. 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... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0075] 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.

[0076] 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.

[0077] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for detecting local deformation of a tower based on laser point clouds from different periods, characterized in that, Includes the following steps: Step 1: Use a 3D laser scanner to scan and inspect the overhead transmission line. According to the time sequence, the tower point clouds collected at different times are divided into a reference point cloud and a comparison point cloud sequence 1, 2, ... m; Step 2: Slice and layer the tower point cloud, and automatically identify the first transverse diaphragm of the tower according to the tower point cloud structure determination criteria; Step 3: Using the first transverse plane of the reference point cloud as the reference object, the Super-4PCS coarse registration and local ICP fine registration algorithms are used to register the comparison point cloud sequence 1, 2, ... m to the reference point cloud; Step 4: Extract the displacement of key nodes of the comparison point cloud sequence 1, 2, ... m relative to the reference point cloud, and compare it with the key node structural anomaly judgment threshold to evaluate whether the tower structure has anomalies. Step 3, the Super-4PCS coarse registration algorithm, includes: Step 3.1.1, in the reference point cloud Four coplanar but non-collinear points are selected, which are common points of the multi-period point clouds, and the distance between them should be as large as possible. This is denoted as B = { , , , } Calculate the distance and intersection ratio of lines AB and CD using the following formula: ； Step 3.1.2, according to the principle of affine invariance, in the point cloud of sequence 1 Find the set of points in the array that have the same distance and intersection ratio with lines AB and CD, as follows: Point cloud in sequence 1 Set of Determined Point Pairs , , with each point Centered on the ball, and respectively with and Draw a sphere with radius, set for and distributed in Points within the range, and distributed in Points within the range are Simultaneously, the point cloud surface is rasterized to create a cell with a size of [missing information]. The three-dimensional mesh G; Step 3.1.3, in the point cloud of sequence 1 Extracting from the baseline point cloud The corresponding four-point set: a set of traversed point pairs and For all candidate point pairs, calculate all intersection nodes e based on cross-ratio consistency and store them in grid G, according to the baseline point cloud. The angle between two corresponding point pairs is θ. Nodes extracted from grid G ​​are approximately equal, and the angle between the lines connecting the two point pairs is also approximately equal to θ. The error range is within... Matching is performed using four basis points within the range; Step 3.1.4, in the point cloud of sequence 1 All cloud points that meet the conditions and benchmark points The corresponding set of four points U={ , , ..., }, calculate the baseline point cloud With each The transformation matrix between w is selected for global transformation by comparing the transformation matrix with the highest registration accuracy among the common point set LCP. Step 3, the local ICP fine registration algorithm, includes: for point clouds and From the benchmark point cloud Select reference point From point clouds Find the distance point The point with the shortest Euclidean distance ,by and The transformation matrix is ​​calculated for the corresponding point set, and the process is iterated continuously. Formula 5 is used as the target error function. The iteration stops when the value is less than the threshold, thus obtaining the optimal transformation matrix. ； Where R represents the rotation transformation matrix, T represents the translation transformation matrix, and k represents the number of corresponding points.

2. The method for detecting local deformation of a tower based on laser point clouds from different periods as described in claim 1, characterized in that, Step 1 includes: A 3D lidar was used to scan the transmission lines, followed by calculation, filtering, and thinning. Point cloud classification technology was used to extract the tower point cloud, removing conductors, ground, and tree point clouds. Based on the time sequence of point cloud acquisition, the first phase of tower point cloud was set as the baseline point cloud, denoted as [reference number missing]. The subsequent point clouds are sequentially set as alignment sequence point clouds, i.e., sequence 1 point cloud, sequence 2 point cloud, ..., sequence m point cloud, denoted as... .

3. The method for detecting local deformation of towers based on laser point clouds from different periods as described in claim 2, characterized in that, Step 2 includes: Step 2.1, slice the tower point cloud: use a set of slices with a thickness of Find the intersection of the plane with the point cloud, slice the point cloud from top to bottom, and measure the thickness of the ground-scanned point cloud slices. Take a 50cm section of machine-scanned point cloud with the following thickness. Take an 80cm section and cut the tower point cloud sequentially from top to bottom along the vertical z-axis to obtain horizontal point cloud slices S. i : ； The point cloud slices from each layer are projected onto a horizontal plane, as shown below: ； Step 2.2, Extract the circumscribed polygon boundaries: The "bidirectional nearest point search method" is used to extract the circumscribed polygon boundaries of each layer of slices, from the point cloud slice projection set. In the process, arbitrarily choose a point Ps as the starting point of the boundary line, find the point Pr that is closest to Ps, and calculate the distance ds from Pr to Ps; In the remaining point cloud, find the nearest point Pe to Ps and calculate the distance de from Ps to Pe; if ds ≤ de, insert Pr before Ps as the new starting point, and construct a line connecting Pr, Ps, and Pe pairwise, replacing Pr with Ps as the starting point; if ds ≥ de, then grow towards the other end Pe; then determine the set Check if there are any remaining points. If so, repeat the above steps; finally, traverse the set. After all points are connected, until PrPe overlaps, connect PsPe to make the polygon closed loop, and obtain the outer contour polygon boundary of the point cloud. Step 2.3, Calculate the vertices of the circumscribed polygon: Using a linear regression equation, fit the outermost boundary point cloud into four boundary lines, denoted clockwise as AB, BC, CD, and DA; the intersection of these four lines represents the four endpoints of the tower's transverse diaphragm, with the endpoint coordinates denoted as A(x...). A , y A , z A ), B(x) B , y B , z B ), C(x) C , y C , z C ), D(x D , y D , z D ); Step 2.4, calculate the aspect ratio, area ratio parameters, and number of points of the circumscribed polygon of the slice: calculate the side length of the point cloud slice for each layer. ,area Number of points Then calculate the area of ​​the circumscribed rectangle. and aspect ratio In the formula Projecting the i-th layer slice The length and width of the circumscribed rectangle; Step 2.5: Based on the number of points in the slice and the aspect ratio and area ratio of its outer polygon, identify the first transverse diaphragm of the tower. The identification criterion is: when slicing from top to bottom, the slice layer where the last peak of the number of slice points is located, and the number of slice points... The number should be greater than 500, and the height of the transverse diaphragm should be less than the tower height. The height above the ground is generally no more than 15m; the aspect ratio of the circumscribed rectangle of the cross-section satisfies: ; slice area ratio 4. The method for detecting local deformation of towers based on laser point clouds from different periods as described in claim 1, characterized in that, Step 4 includes: Step 4.1, Point Cloud Segmentation: Segment the point cloud separately. and Principal component analysis is used to obtain the principal direction eigenvector of the two-dimensional projected point cloud. Based on this, the principal direction rotation angle is calculated, and the tower point cloud is rotated to obtain the redirected point cloud data. The direction perpendicular to the rotated principal direction on the horizontal plane is selected as the horizontal direction, and the point cloud is cut along the horizontal direction. The cut point cloud is divided into left and right parts. Step 4.2, Scattered Point Cloud Indexing: Indexing the baseline point cloud And Sequence 1 point cloud Use the Octree method to create a scattered point cloud index; Step 4.3, Euclidean Clustering Segmentation: This involves segmenting the baseline point cloud based on the established Octree index. And Sequence 1 point cloud Euclidean clustering is performed on the insulators to distinguish the phases of the insulators. Each tower point cloud is set as P, which consists of v points. The Euclidean distance between the i-th sample and the j-th sample is calculated. The points with the smallest distance are grouped into the same class. Then, the Euclidean distance between the new classes is calculated and iterated until a specified threshold is less than the Euclidean distance between any two classes or the number of classes is less than a specified number. The Euclidean clustering is then completed. Step 4.4, Key Node Location Extraction: Extract the point cloud locations for each clustered period. and Calculate the three-dimensional coordinates of the crossarm attachment points of each phase insulator and the attachment points of the left and right phase ground wires, and then calculate the point cloud. and The displacement of key nodes at the same location along the x, y, and z axes; Step 4.5, Structural Anomaly Judgment: Compare the displacement of each key node with the key node structural anomaly judgment threshold to determine whether the key nodes of the multi-phase point cloud have displaced, and then determine whether the tower structure has deformed.

5. The method for detecting local deformation of towers based on laser point clouds from different periods as described in claim 4, characterized in that, The specific implementation process of step 4.3 is as follows: Step 4.3.1: Establish an Octree data structure for each period's tower point cloud P; Step 4.3.2: Create an empty cluster set C and a queue; Step 4.3.3, for each point ∈P, perform the following steps: (1) Add the point to the current queue Q; (2) For each point ∈Q, perform the following operations: ① Using the search radius r < ,right Perform a k-nearest neighbor search, and the searched points are: ; ② For each ∈ ,calculate and The two closest Euclidean distances are grouped into the same class; ③ Check each Check if the point has performed the above operations. If not, add the point to queue Q. (3) When all points in the list of queue Q have performed the above operations, add the points in Q to the list of set C and clear the list of Q; (4) When each point ∈P performs the above operations, and If all elements are part of set C, the algorithm terminates.

Images