A method for tower type matching and waypoint automatic recommendation of corner towers in power transmission lines

Through the matching of point cloud towers and automatic waypoint recommendation methods of corner towers on transmission lines, the problem of infrared thermal imaging relying on manual experience is solved, and more efficient drone inspection is achieved, ensuring high-quality shooting of insulator infrared images and Simplified operation.

CN118506036BActive Publication Date: 2024-11-12STATE GRID JIANGXI ELECTRIC POWER CO LTD RES INST +2
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410956156.0
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-07-17
Publication Date
2024-11-12
Estimated Expiration
2044-07-17

AI Technical Summary

Technical Problem

At the power transmission line maintenance site, the waypoint position of infrared thermal imaging method used to detect defective insulators depends on artificial subjective experience, resulting in poor imaging quality, cumbersome operation and error-prone, and drone inspection requires manual setting of a large number of waypoints. The workload is huge.

Method used

The method of matching the point cloud tower type of the transmission line angle tower and automatic waypoint recommendation is adopted. By setting the key points of the straight-line tower point cloud model, coarse matching and fine matching registration are performed, and the target detection intelligent algorithm and Hough transformation are used to calculate the conductor and tower. The angle of the point cloud model automatically recommends the waypoint position to ensure the better imaging quality of the insulator infrared images captured by the drone.

Benefits of technology

Improve the quality and efficiency of drone inspection work. By accurately matching the distance between the tops of the tower and automatically calculating the rotation angle, ensure that the waypoints fall on the perpendicular line of the insulator space, improve the infrared image imaging effect and reduce artificial Operation errors and workload.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118506036B_ABST
    Figure CN118506036B_ABST
Patent Text Reader

Abstract

The invention discloses a method for tower type matching and automatic waypoint recommendation of a transmission line corner tower point cloud, comprising the following steps: setting key points of a single straight tower point cloud model; cutting out a transmission line tower point cloud model from the top of the tower of the transmission line point cloud model; performing coarse matching and improved fine matching registration on the point cloud models of the straight tower and the transmission line tower to obtain a recommended waypoint position; calculating the angle between the conductor and the transmission line tower point cloud model, rotating the waypoint and the corresponding auxiliary window point position according to the angle between the conductor and the tower; determining the yaw angle and the pitch angle according to the rotated waypoint and the corresponding auxiliary window point position; the invention considers the distance between the tower tops during matching, so as to make the result more accurate, automatically calculates the degree of tower rotation through an intelligent plus classical algorithm, and is suitable for tower type matching of corner towers, and the waypoint falls on the mid-vertical line of the insulator space. Compared with the existing tower type matching, the imaging effect of the infrared image of the insulator taken by the waypoint is better.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of power transmission lines, and in particular to a method for tower type matching and automatic waypoint recommendation of a corner tower of a power transmission line. Background Art

[0002] Insulators are a common key component of power transmission lines and are widely used in power transmission, transformation and distribution systems. Insulator strings are affected by a variety of complex factors such as mechanical stress, electric fields, lightning strikes, hail, etc., and are prone to defects. Defective insulators cannot withstand normal electromechanical loads, resulting in local heating, cracking and other phenomena, endangering the safe operation of transmission lines.

[0003] At present, there are many types of insulator defect detection methods, mainly including voltage distribution method, insulation resistance method, leakage current method, infrared thermal imaging method, etc.; among them, the most commonly used non-contact live detection method is infrared thermal imaging method using handheld or airborne infrared thermal imagers for patrol inspections, which has the advantages of high efficiency, low safety risks and low costs.

[0004] At the transmission line equipment maintenance site, the usual infrared detection method is to use drones to carry out flight missions, arrive at the designated location to take infrared images of the target, extract the temperature in the image, and determine the defective insulators through the temperature matrix diagnosis criteria; the flight mission waypoints rely on people's subjective experience, and the actual deviation of the waypoint position will lead to poor imaging quality of the captured image. In addition, a single transmission line usually contains hundreds or thousands of iron towers, and each iron tower requires manual setting of multiple waypoints, which is cumbersome and repetitive, with a huge workload and prone to errors. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention provides a method for tower type matching and automatic waypoint recommendation of corner towers of transmission lines, which aims to improve the quality and efficiency of drone inspection work.

[0006] To achieve the above object, the present invention provides the following technical solution: a method for tower type matching and waypoint automatic recommendation of corner towers of power transmission lines, comprising the following steps:

[0007] Step S1: setting key points of a single straight tower point cloud model;

[0008] Step S2: Click the top of the tower in the transmission line point cloud model to cut out the transmission line tower point cloud model;

[0009] Step S3: performing rough matching and improved fine matching registration on the straight tower point cloud model and the transmission line tower point cloud model, transforming the straight tower point cloud model coordinates to the line tower point cloud model coordinates, and obtaining the recommended waypoint position;

[0010] The specific process of step S3 is: downsampling the straight tower point cloud model and the transmission line tower point cloud model to reduce the number of points and the amount of calculation, using coarse matching to randomly select 3 points on the straight tower point cloud model and the transmission line tower point cloud model for feature calculation, and then perform matrix transformation, calculate the overlap, set the maximum number of iterations, repeat the iteration until the transformation with the highest overlap, transform the straight tower point cloud model coordinates to the line tower point cloud model coordinates, and correspondingly, transform the key point coordinates on the straight tower point cloud model to the relative position on the line tower, and use the relative position on the line tower point cloud model as the recommended waypoint position; similarly, improve the fine matching to use the least squares method to find the optimal matrix in the feature calculation, and at the same time judge the distance between the top plane of the straight tower point cloud model and the transmission line tower point cloud model;

[0011] Step S4: using the target detection intelligent algorithm and Hough transform to calculate the angle between the conductor and the point cloud model of the transmission line tower, and rotating the waypoint and the corresponding auxiliary window point according to the angle between the conductor and the point cloud model of the transmission line tower;

[0012] Step S5: Determine the yaw angle and pitch angle through the rotated waypoint and the corresponding auxiliary window point.

[0013] Furthermore, the specific process of step S1 is: constructing a single straight tower point cloud model, cropping the single straight tower point cloud model to obtain a template, setting four key points for the template, namely, a hanging point, two endpoints of the insulator and a waypoint; the position of the hanging point is the connection between the insulator and the straight tower point cloud model, that is, the horizontal rotation center of the insulator string, the position of the hanging point will not change, the two endpoints of the insulator are the insulator targets visualized in the auxiliary window, which are used to subsequently determine the orientation of the UAV fuselage, and the waypoint is the UAV flight hovering position.

[0014] Furthermore, the specific process of step S2 is: construct a point cloud model of the transmission line, click on the top of the tower in the transmission line point cloud model, obtain the current height of the top of the tower, use the top of the tower in the transmission line point cloud model as the center of the plane circle to crop the point cloud model within the cylindrical range downward, and remove the ground in the point cloud model within the cylindrical range. After completion, the point cloud model of the transmission line tower is obtained.

[0015] Furthermore, the point cloud model of the straight tower and the point cloud model of the transmission line tower are downsampled to reduce the number of points and the amount of calculation. Three points are randomly selected using coarse matching for feature calculation, and then matrix transformation is performed to calculate the overlap. The maximum number of iterations is set, and the iteration is repeated until the transformation with the highest overlap is achieved. The coordinates of the straight tower point cloud model are transformed to the coordinates of the line tower point cloud model. Correspondingly, the coordinates of the four key points of the hanging point, the two endpoints of the insulator and the waypoint on the straight tower point cloud model are transformed to the relative positions on the line tower point cloud model. The relative positions on the line tower point cloud model are used as the recommended waypoint positions, which are expressed as:

[0016] (1);

[0017] (2);

[0018] (3);

[0019] (4);

[0020] (5);

[0021] In the formula, , are the centroids of the point cloud model of the straight tower and the point cloud model of the transmission line tower, , They are the points of the straight line tower point cloud model and the transmission line tower point cloud model, , are the distances from the points to the centroid on the straight tower point cloud model and the transmission line tower point cloud model, respectively. J is a 3×3 matrix. is the total number of points on the straight tower point cloud model, T is the transposed matrix, M and N are orthogonal matrices, For the optimal rotation, is the optimal translation;

[0022] Improved fine matching uses the least squares method to find the optimal matrix in feature calculation, and at the same time judges the top plane distance between the straight tower point cloud model and the transmission line tower point cloud model, which is expressed as:

[0023] (6);

[0024] In the formula, min is the minimum approximate value calculated by the least squares method, n is the number of points, and R and t are , Adjoint matrix.

[0025] Furthermore, the specific process of step S4 is: intercepting the point cloud model of the top of the transmission line tower point cloud model with the top endpoint of the transmission line tower point cloud model downward by 0.5 units, transforming the 3D coordinates of the intercepted point cloud model of the top of the transmission line tower point cloud model into 2D coordinates, and converting it into the top image of the transmission line tower point cloud model, using the target detection algorithm to identify the wire target in the top image of the transmission line tower point cloud model, and performing Hough transform on the identified wire, that is, the wire in the detection frame area to obtain the straight line equation; similarly, using Hough transform to obtain the straight line equation of the longest edge of the transmission line tower point cloud model in the top image of the transmission line tower point cloud model, obtaining the rotation angle by calculating the angle between the two straight lines, and horizontally rotating the two end points and the waypoints of the insulator of the straight line tower point cloud model with the hanging point as the center.

[0026] Furthermore, the specific process of step S5 is: the endpoint center is obtained according to the two end points of the insulator, the endpoint center is projected to the horizontal plane of the waypoint as the center projection point, the angle between the straight line from the center projection point to the waypoint and the initial orientation of the UAV fuselage is the yaw angle, and the angle between the straight line from the endpoint center to the waypoint and the straight line from the center projection point to the waypoint is the pitch angle. Similarly, the yaw angle and pitch angle are obtained by the straight line angle formula.

[0027] Furthermore, the calculation of the angle between the two straight lines is expressed as:

[0028] (7);

[0029] Where θ is the degree of rotation, k1 and k2 are the slopes of the straight line of the conductor and the longest edge of the point cloud model of the transmission line tower obtained by Hough transform.

[0030] Furthermore, the horizontal rotation of the two end points and the waypoints of the insulator of the straight tower point cloud model with the hanging point as the center is expressed as:

[0031] A = [ cos i − sin i sin i cos i ] (8);

[0032] Where A is the coordinate transformation rotation matrix.

[0033] Compared with the existing technology, the present invention has the following beneficial effects: the present invention considers the distance between the tower tops during matching, which can make the result more accurate, and automatically calculates the angle of the corner tower through the intelligent plus classical algorithm, which is suitable for the tower type matching of the corner tower, and the waypoint falls on the vertical line in the insulator space. Compared with the existing tower type matching, the imaging effect of the infrared image of the insulator taken by this waypoint is better; the present invention creatively proposes an automatic waypoint recommendation method for corner tower type insulators, which has high technical advancement and wide applicability; by adding plane to plane matching on the basis of point cloud point to point matching, the matching result is more accurate; by converting three-dimensional data into two-dimensional images, target detection and Hough transform are used to detect the straight line to calculate the angle between the tower and the conductor. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 The figure is a flow chart of the method of the present invention.

[0035] Figure 2 It is the tower type matching principle diagram of the present invention.

[0036] Figure 3 It is a principle diagram of the target detection result of the present invention.

[0037] Figure 4 It is the principle diagram of the yaw angle and pitch angle of the present invention. DETAILED DESCRIPTION

[0038] like Figure 1 As shown, the present invention provides a technical solution: a method for matching tower type and automatic waypoint recommendation of a transmission line corner tower point cloud, comprising the following steps:

[0039] Step S1: setting key points of a single straight tower point cloud model;

[0040] Step S2: Click the top of the tower in the transmission line point cloud model to cut out the transmission line tower point cloud model;

[0041] Step S3: performing rough matching and improved fine matching registration on the straight tower point cloud model and the transmission line tower point cloud model, transforming the straight tower point cloud model coordinates to the line tower point cloud model coordinates, and obtaining the recommended waypoint position;

[0042] Step S4: using the target detection intelligent algorithm and Hough transform to calculate the angle between the conductor and the point cloud model of the transmission line tower, and rotating the waypoint and the corresponding auxiliary window point according to the angle between the conductor and the point cloud model of the transmission line tower;

[0043] Step S5: Determine the yaw angle and pitch angle through the rotated waypoint and the corresponding auxiliary window point.

[0044] Among them, the specific process of step S1 is: construct a single straight tower point cloud model, crop the single straight tower point cloud model to obtain a template, set four key points for the template, namely, the hanging point, the two endpoints of the insulator and the waypoint; the position of the hanging point is the connection between the insulator and the straight tower point cloud model, that is, the horizontal rotation center of the insulator string, the hanging point position will not change, the two endpoints of the insulator are the insulator targets visualized in the auxiliary window, which are used to subsequently determine the orientation of the drone fuselage, and the waypoint is the drone's flying hovering position.

[0045] Among them, the specific process of step S2 is: construct a point cloud model of the transmission line, click on the top of the tower in the transmission line point cloud model, obtain the current height of the top of the tower, use the top of the tower in the transmission line point cloud model as the center of the plane circle to cut the point cloud model within the cylindrical range downward, and remove the ground in the point cloud model within the cylindrical range. After completion, the point cloud model of the transmission line tower is obtained.

[0046] like Figure 2 As shown, the specific process of step S3 is: downsampling the straight tower point cloud model and the transmission line tower point cloud model to reduce the number of points and the amount of calculation, using coarse matching to randomly select 3 points on the straight tower point cloud model and the transmission line tower point cloud model for feature calculation, and then perform matrix transformation, calculate the overlap, set the maximum number of iterations, repeat the iteration until the transformation with the highest overlap, transform the straight tower point cloud model coordinates to the line tower point cloud model coordinates, and correspondingly, transform the four key point coordinates of the hanging point, the two end points of the insulator and the waypoint on the straight tower point cloud model to the relative position on the line tower point cloud model, and use the relative position on the line tower point cloud model as the recommended waypoint position. Similarly, the improved fine matching uses the least squares method to find the optimal matrix in the feature calculation, and at the same time, the distance between the straight tower point cloud model and the top plane of the transmission line tower point cloud model is judged;

[0047] (1);

[0048] (2);

[0049] (3);

[0050] (4);

[0051] (5);

[0052] (6);

[0053] In the formula, , are the centroids of the point cloud model of the straight tower and the point cloud model of the transmission line tower, , They are the points of the straight tower point cloud model and the transmission line tower point cloud model, , are the distances from the points to the centroid on the straight tower point cloud model and the transmission line tower point cloud model, respectively. J is a 3×3 matrix. is the total number of points on the straight tower point cloud model, T is the transposed matrix, M and N are orthogonal matrices, For the optimal rotation, is the optimal translation, min is the minimum approximation calculated by the least squares method, n is the number of points, R and t are , Adjoint matrix.

[0054] like Figure 3 As shown, the specific process of step S4 is: intercepting the point cloud model of the top of the transmission line tower point cloud model with the top endpoint of the transmission line tower point cloud model downward by 0.5 units, transforming the 3D coordinates of the intercepted point cloud model of the top of the transmission line tower point cloud model into 2D coordinates, and converting it into the top image of the transmission line tower point cloud model, using the target detection algorithm to identify the wire target in the top image of the transmission line tower point cloud model, and performing Hough transform on the identified wire, that is, the wire in the detection frame area to obtain the straight line equation; similarly, using Hough transform to obtain the straight line equation of the longest edge of the transmission line tower point cloud model in the top image of the transmission line tower point cloud model, and obtaining the rotation angle by calculating the angle between the two straight lines, and horizontally rotating the two end points and the waypoints of the insulator of the straight line tower point cloud model with the hanging point as the center;

[0055] (7);

[0056] A = [ cos i − sin i sin i cos i ] (8);

[0057] Where θ is the degree of rotation, k1 and k2 are the slopes of the straight line of the conductor and the longest straight line slope of the edge of the transmission line tower point cloud model obtained by Hough transform, and A is the coordinate transformation rotation matrix.

[0058] like Figure 4 As shown, the specific process of step S5 is: the endpoint center is obtained according to the two end points of the insulator, the endpoint center is projected to the horizontal plane of the waypoint as the center projection point, the angle between the straight line from the center projection point to the waypoint and the initial orientation of the UAV fuselage is the yaw angle, and the angle between the straight line from the endpoint center to the waypoint and the straight line from the center projection point to the waypoint is the pitch angle. Similarly, the yaw angle and pitch angle are obtained by the straight line angle formula.

[0059] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for tower type matching and waypoint automatic recommendation of transmission line corner tower point cloud, It is characterized in that The steps include: Step S1: setting key points of a single straight tower point cloud model; Step S2: Click the top of the tower in the transmission line point cloud model to cut out the transmission line tower point cloud model; Step S3: performing rough matching and improved fine matching registration on the straight tower point cloud model and the transmission line tower point cloud model, transforming the straight tower point cloud model coordinates to the line tower point cloud model coordinates, and obtaining the recommended waypoint position; The specific process of step S3 is: downsampling the straight tower point cloud model and the transmission line tower point cloud model to reduce the number of points and the amount of calculation, using coarse matching to randomly select 3 points on the straight tower point cloud model and the transmission line tower point cloud model for feature calculation, and then performing matrix transformation, calculating the overlap, setting the maximum number of iterations, repeating the iteration until the transformation with the highest overlap, transforming the straight tower point cloud model coordinates to the line tower point cloud model coordinates, and correspondingly, transforming the key point coordinates on the straight tower point cloud model to the relative position on the line tower, and taking the relative position on the line tower point cloud model as the recommended waypoint position; Similarly, the improved fine matching uses the least square method to find the optimal matrix in feature calculation, and at the same time, the distance between the top plane of the straight tower point cloud model and the transmission line tower point cloud model is judged; Step S4: using the target detection intelligent algorithm and Hough transform to calculate the angle between the conductor and the point cloud model of the transmission line tower, and rotating the waypoint and the corresponding auxiliary window point according to the angle between the conductor and the point cloud model of the transmission line tower; Step S5: Determine the yaw angle and pitch angle through the rotated waypoint and the corresponding auxiliary window point; The specific process of step S2 is: constructing a point cloud model of the transmission line, clicking the top of the tower in the point cloud model of the transmission line, obtaining the current height of the top of the tower, cutting the point cloud model within the cylindrical range downward with the top of the tower in the point cloud model of the transmission line as the center of the plane circle, and removing the ground in the point cloud model within the cylindrical range, and obtaining the point cloud model of the transmission line tower after completion; Downsample the point cloud model of the straight tower and the point cloud model of the transmission line tower to reduce the number of points and the amount of calculation. Use coarse matching to randomly select 3 points for feature calculation, then perform matrix transformation, calculate the overlap, set the maximum number of iterations, and repeat the transformation to the highest overlap. Transform the coordinates of the straight tower point cloud model to the line tower point cloud model coordinates. Correspondingly, the coordinates of the four key points of the hanging point, the two endpoints of the insulator and the waypoint on the straight tower point cloud model are transformed to the relative positions on the line tower point cloud model. The relative positions on the line tower point cloud model are used as the recommended waypoint positions, which are expressed as: (1); (2); (3); (4); (5); In the formula, , are the centroids of the point cloud model of the straight tower and the point cloud model of the transmission line tower, , They are the points of the straight line tower point cloud model and the transmission line tower point cloud model, , are the distances from the points to the centroid on the straight tower point cloud model and the transmission line tower point cloud model, respectively. J is a 3×3 matrix. is the total number of points on the straight tower point cloud model, T is the transposed matrix, M and N are orthogonal matrices, For the optimal rotation, is the optimal translation; Improved fine matching uses the least squares method to find the optimal matrix in feature calculation, and at the same time judges the top plane distance between the straight tower point cloud model and the transmission line tower point cloud model, which is expressed as: (6); In the formula, min is the minimum approximate value calculated by the least squares method, n is the number of points, and R and t are , adjoint matrix; The specific process of step S4 is: intercepting the point cloud model of the top of the transmission line tower point cloud model with the top endpoint of the transmission line tower point cloud model downward by 0.5 units, transforming the 3D coordinates of the intercepted point cloud model of the top of the transmission line tower point cloud model into 2D coordinates, and converting it into a top image of the transmission line tower point cloud model, using the target detection algorithm to identify the wire target in the top image of the transmission line tower point cloud model, and performing Hough transform on the identified wire, that is, the wire in the detection frame area to obtain the straight line equation; Similarly, the Hough transform is used to obtain the straight line equation of the longest edge of the transmission line tower point cloud model in the top image of the transmission line tower point cloud model. The rotation angle is obtained by calculating the angle between the two straight lines. The two end points and waypoints of the insulator of the straight line tower point cloud model are horizontally rotated with the hanging point as the center.

2. A method for matching tower type and automatically recommending waypoints for a transmission line corner tower point cloud according to claim 1, Features: The specific process of step S1 is: construct a single straight tower point cloud model, crop the single straight tower point cloud model to obtain a template, set four key points for the template, namely, the hanging point, the two endpoints of the insulator and the waypoint; the position of the hanging point is the connection between the insulator and the straight tower point cloud model, that is, the horizontal rotation center of the insulator string, the hanging point position will not change, the two endpoints of the insulator are the insulator targets visualized in the auxiliary window, which are used to subsequently determine the orientation of the UAV fuselage, and the waypoint is the UAV flight hovering position.

3. A method for matching tower type and automatically recommending waypoints for a transmission line corner tower point cloud according to claim 1, Features: The specific process of step S5 is: the endpoint center is obtained according to the two end points of the insulator, the endpoint center is projected to the horizontal plane of the waypoint as the center projection point, the angle between the straight line from the center projection point to the waypoint and the initial orientation of the UAV fuselage is the yaw angle, and the angle between the straight line from the endpoint center to the waypoint and the straight line from the center projection point to the waypoint is the pitch angle. Similarly, the yaw angle and pitch angle are obtained by the straight line angle formula.

4. A method for matching tower type and automatically recommending waypoints for a transmission line corner tower point cloud according to claim 3, Features: The calculation of the angle between two straight lines is expressed as: (7); Where θ is the rotation angle, k 1 , k 2 It is the straight line slope of the conductor obtained by Hough transform and the straight line slope of the longest edge of the point cloud model of the transmission line tower.

5. A method for matching tower type and automatically recommending waypoints for a transmission line corner tower point cloud according to claim 4, Features: The horizontal rotation of the two end points and waypoints of the insulator of the straight tower point cloud model with the hanging point as the center is expressed as: (8); Where A is the coordinate transformation rotation matrix.