Power transmission line icing thickness measurement method based on YOLOv8 segmentation algorithm

Abstract

The invention relates to a power transmission line icing thickness measurement method based on a YOLOv8 segmentation algorithm. The method comprises the following steps: S1, obtaining a group of icing pictures with positioning information; s2, acquiring the total number P2 of icing pixel points and the total number of picture lead pixel points; s3, calculating a plane projection distance d between the unmanned aerial vehicle and the iron tower; s4, counting the total number P1 of manually labeled lead mask pixels; s5, acquiring related parameters according to a camera imaging principle; s6, according to the triangle similarity principle and the related parameters, calculating the proportion of the image wire diameter X to the actual wire diameter Wr; s7, calculating the diameter D of the actual wire; s8, calculating the ice coating area S of the actual wire; and S8, calculating the actual conductor icing average thickness b. According to the method for measuring the icing thickness of the power transmission line, the danger of manual observation is avoided, the operation risk of personnel in a severe environment is reduced, and compared with a traditional method, the cost is low, and application and popularization are easy.

Description

Technical Field

[0001] The present invention belongs to the field of electric power technology, and particularly relates to a method for measuring the icing thickness of transmission lines based on the YOLOv8 segmentation algorithm. Background Art

[0002] The power system is constantly progressing towards ultra-high voltage, large-capacity, and long-distance power transmission. As the artery of the power system, transmission lines undertake the important task of safely and efficiently transmitting electric energy from the power plant to the user side. However, transmission lines face many challenges during operation, among which the icing phenomenon is particularly prominent. Icing on transmission lines not only increases the weight of the line and intensifies the conductor galloping, but may also cause serious accidents such as wire breakage and tower collapse.

[0003] Existing methods for measuring the icing thickness of transmission lines mainly adopt traditional manual observation and mechanical measurement methods. Manual observation has low accuracy, high danger, and low efficiency. Mechanical measurement methods require the installation of special equipment on transmission lines, with high costs and difficulties in installation and maintenance. Some icing measurement methods based on technologies such as lidar and radar are also under research, but the equipment is expensive and the operation is complex, making it difficult to be applied on a large scale.

[0004] Therefore, the present invention provides a method for measuring the icing thickness of transmission lines based on the YOLOv8 segmentation algorithm to solve the above technical problems. Summary of the Invention

[0005] Aiming at the above problems, the purpose of the present invention is to provide a method for measuring the icing thickness of transmission lines based on the YOLOv8 segmentation algorithm, which avoids the danger of manual observation, reduces the operation risk of personnel in harsh environments, and has lower costs compared with traditional methods, being easy to promote and apply.

[0006] The present invention provides a method for measuring the icing thickness of transmission lines based on the YOLOv8 segmentation algorithm, which includes:

[0007] S1. Using a drone equipped with a positioning system, set the coordinate system as CGCS2000, import a preset fixed inspection route, and perform image acquisition on the icing transmission line to obtain a set of icing pictures with positioning information;

[0008] S2. Adopt the YOLOv8n-seg algorithm to perform instance segmentation on the icing and conductors in the icing pictures, and accurately obtain the total number of icing pixel points P2 and the total number of picture conductor pixel points;

[0009] S3. Calculate the planar projection distance d between the drone and the iron tower based on the icing pictures, the coordinates of the icing pictures, the iron tower coordinates, the camera parameters, and the conductor diameter;

[0010] S4. For the first use or under specific circumstances, perform a manual annotation operation on the icing image to determine the wire area in the image, and statistically obtain the total number of wire mask pixels P1 manually annotated.

[0011] S5. According to the camera imaging principle, obtain relevant parameters; the relevant parameters include the lens focal length f, the width w of the lens target surface, the height h of the lens target surface, the width W of the icing image, the height H of the icing image, the diameter X of the image wire in the icing image, the direct distance Z between the lens and the actual wire, the actual wire diameter X, and the width W of the image formed on the target surface. L Among them, the lens focal length f is obtained through the specification information corresponding to the camera model, the width w and the height h of the lens target surface are determined based on the physical properties of the camera itself, the width W and the height H of the icing image are obtained through the image resolution information, and the direct distance Z between the lens and the actual wire is measured and obtained through the positioning system carried on the drone.

[0012] S6. According to the principle of similar triangles and the relevant parameters, calculate the ratio of the image wire diameter X to the actual wire diameter W. r The ratio.

[0013] S7. According to the ratio of the image wire diameter X to the actual wire diameter W. r And the total number of wire mask pixels P1, calculate the diameter D of the actual wire.

[0014] S8. Calculate the average value of the ratio of the icing image pixel scale to the corresponding width and height of the sensor physical scale through the camera sensor size and resolution, and calculate the icing area S of the actual wire according to the total number of icing pixels P2, the average value of the ratio of the icing image pixel scale to the corresponding width and height of the sensor physical scale, and the diameter D of the actual wire.

[0015] S9. Take the width W of the icing image as the length L of the actual wire, and calculate the average icing thickness b of the actual wire according to the length L of the actual wire and the icing area S of the actual wire.

[0016] Preferably, the planar projection distance between the drone and the iron tower is calculated in the following way: Among them, d represents the planar projection distance between the drone and the iron tower, x1 and y1 are the east coordinate and north coordinate of the iron tower, and x2 and y2 are the east coordinate and north coordinate of the drone.

[0017] Preferably, the steps of calculating the ratio of the image wire diameter X to the actual wire diameter W. r Specifically include:

[0018] S51. According to the principle of similar triangles, the focal length f of the lens, the direct distance Z between the lens and the actual wire, and the width W of the image formed on the target surface L and the diameter W of the actual wire r have the first proportional relationship as follows:

[0019] S52. According to the principle of similar triangles, the width W of the ice-covered image, the width W of the lens target surface B , the diameter X of the wire in the ice-covered image, and the width W of the image formed on the target surface L have the second proportional relationship as follows:

[0020]

[0021] S53. Calculate the ratio of the diameter X of the image wire to the diameter W of the actual wire according to the first proportional relationship and the second proportional relationship r as follows:

[0022] Preferably, the ice-covered area S of the actual wire is obtained through the following calculation formula:

[0023]

[0024] Preferably, the average ice-covered thickness b of the actual wire is calculated through the following formula:

[0025] Compared with the related technology, a method for measuring the ice-covered thickness of a transmission line based on the YOLOv8 segmentation algorithm provided by the present invention can quickly cover a large area of the transmission line through image acquisition by an unmanned aerial vehicle (UAV), improving the efficiency of ice-covered measurement; the YOLOv8n-seg algorithm can accurately identify and segment the ice-covered area in the images taken by the UAV, and achieve high-precision ice-covered measurement by calculating parameters such as the area and thickness of the ice-covered area. At the same time, it avoids the danger of manual observation, reduces the operation risk of personnel in harsh environments, and has a low cost, being easy to promote and apply; through the positioning system carried on the UAV, the position of the UAV can be determined more accurately, providing more accurate spatial information for ice-covered measurement, and further improving the performance and reliability of the measurement system. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 is a schematic flow chart of a method for measuring the ice-covered thickness of a transmission line based on the YOLOv8 segmentation algorithm of the present invention;

[0027] Figure 2 is a schematic flow chart of calculating the ratio of the diameter X of the image wire to the diameter W of the actual wire of the present invention r ;

[0028] Figure 3 Schematic diagram for calculating the planar projection distance d between the drone and the iron tower in the present invention;

[0029] Figure 4 Schematic diagram for manually annotating the icing picture in S4 of the present invention;

[0030] Figure 5 Schematic diagram of the principle of camera imaging in S5 of the present invention. Detailed implementation manners

[0031] The present invention provides a method for measuring the icing thickness of a transmission line based on the YOLOv8 segmentation algorithm, aiming to solve the problems of low manual observation accuracy, high danger, low efficiency and difficulty in large-scale application of the existing methods for measuring the icing thickness of a transmission line.

[0032] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0033] Please refer to the attached Figure 1 As shown, the present invention provides a method for measuring the icing thickness of a transmission line based on the YOLOv8 segmentation algorithm, which includes:

[0034] S1, using a drone equipped with a positioning system to set the coordinate system as CGCS2000, importing a preset fixed inspection route, and collecting images of the icing transmission line to obtain a set of icing pictures with positioning information;

[0035] In this embodiment, the positioning system carried on the drone is an RTK positioning system. Through the RTK positioning system, the CGCS2000 coordinate system can more accurately determine the position of the drone, provide more accurate spatial information for icing measurement, and further improve the performance and reliability of the measurement.

[0036] S2, using the YOLOv8n-seg algorithm to perform instance segmentation on the icing and conductors in the icing pictures, and accurately obtaining the total number of icing pixel points P2 and the total number of picture conductor pixel points.

[0037] In this embodiment, through the YOLOv8n-seg algorithm, the icing area in the images taken by the drone can be accurately identified and segmented. By calculating parameters such as the area and thickness of the icing area, high-precision icing measurement can be achieved. At the same time, the danger of manual observation is avoided, the operation risk of personnel in harsh environments is reduced, and the cost is relatively low, making it easy to promote and apply.

[0038] S3. Calculate the planar projection distance d between the UAV and the iron tower based on the icing picture, the coordinates of the icing picture, the coordinates of the iron tower, the camera parameters, and the wire diameter.

[0039] In this embodiment, the planar projection distance between the UAV and the iron tower is calculated as follows: Where d represents the planar projection distance between the UAV and the iron tower, x1 and y1 are the east coordinate and north coordinate of the iron tower, and x2 and y2 are the east coordinate and north coordinate of the UAV.

[0040] S4. For the first use or in specific situations, perform a manual annotation operation on the icing picture to determine the wire area in the picture, and count the total number of pixels P1 of the manually annotated wire mask.

[0041] It should be noted that due to the obstruction of the wire by ice, image segmentation cannot accurately identify the un-iced wire. To ensure the stability of the wire shape in the picture, for the first use, the trajectory of the wire in the picture needs to be manually marked to calculate the cross-sectional area of the wire when imported into the software later.

[0042] S5. According to the camera imaging principle, obtain relevant parameters; the relevant parameters include the lens focal length f, the width w of the lens target surface, the height h of the lens target surface, the width W of the icing picture, the height H of the icing picture, the diameter X of the wire in the icing picture, the direct distance Z between the lens and the actual wire, the actual wire diameter X, and the width W of the image on the target surface. L ; Among them, the lens focal length f is obtained through the specification information corresponding to the camera model, the width w and height h of the lens target surface are determined based on the physical properties of the camera itself, the width W and height H of the icing picture are obtained through the image resolution information, and the direct distance Z between the lens and the actual wire is measured and obtained through the positioning system carried on the UAV.

[0043] S6. According to the principle of similar triangles and the relevant parameters, calculate the ratio of the diameter X of the wire in the image to the actual wire diameter W. r Ratio.

[0044] Specifically, the steps of calculating the ratio of the diameter X of the wire in the image to the actual wire diameter W. r Specifically include: S51. According to the principle of similar triangles, obtain the first proportional relationship among the lens focal length f, the direct distance Z between the lens and the actual wire, the width W of the image on the target surface. L And the actual wire diameter W. r Is: S52. According to the principle of similar triangles, obtain the width W of the icing picture, the width W of the lens target surface.B The diameter X of the wire in the ice-covered picture and the width W imaged on the target surface L The second proportional relationship is: S52. Calculate the ratio of the image wire diameter X to the actual wire diameter W according to the first proportional relationship and the second proportional relationship r The ratio is:

[0045] S7. Calculate the diameter D of the actual wire according to the ratio of the image wire diameter X to the actual wire diameter W r and the total number of wire mask pixels P1

[0046] S8. Calculate the average value of the ratio of the pixel scale of the ice-covered picture to the corresponding width and height of the physical scale of the sensor through the camera sensor size and resolution, and calculate the ice-covered area S of the actual wire according to the total number of ice-covered pixels P2, the average value of the ratio of the pixel scale of the ice-covered picture to the corresponding width and height of the physical scale of the sensor, and the diameter D of the actual wire

[0047] In this embodiment, the ice-covered area S of the actual wire is obtained through the following calculation formula:

[0048]

[0049] S9. Use the width W of the ice-covered picture as the length L of the actual wire, and calculate the average ice-covered thickness b of the actual wire according to the length L of the actual wire and the ice-covered area S of the actual wire

[0050] In this embodiment, the average ice-covered thickness b of the actual wire is calculated through the following formula:

[0051]

[0052] Compared with the related technology, a method for measuring the ice-covered thickness of a transmission line based on the YOLOv8 segmentation algorithm provided by the present invention can quickly cover a large area of the transmission line through unmanned aerial vehicle (UAV) image acquisition, improving the ice-covered measurement efficiency; the YOLOv8n-seg algorithm can accurately identify and segment the ice-covered area in the images taken by the UAV, and achieve high-precision ice-covered measurement by calculating parameters such as the area and thickness of the ice-covered area. At the same time, it avoids the danger of manual observation, reduces the operation risk of personnel in harsh environments, and has a low cost, making it easy to promote and apply; through the positioning system carried on the UAV, the position of the UAV can be determined more accurately, providing more accurate spatial information for ice-covered measurement, and further improving the performance and reliability of the measurement system

[0053] The embodiments described above should be understood as illustrative and not limiting the scope of protection of the present invention. The scope of protection of the present invention shall be subject to the claims. For those skilled in the art, without departing from the essence and scope of the present invention, some non-essential improvements and adjustments made to the present invention still fall within the scope of protection of the present invention.

Claims

1. A method for measuring the icing thickness of a transmission line based on the YOLOv8 segmentation algorithm, characterized in that, The method for measuring the icing thickness of a transmission line includes: S1. Use a drone equipped with a positioning system to set the coordinate system as CGCS2000, import a preset fixed inspection route, collect images of the iced transmission line, and obtain a set of iced pictures with positioning information; S2. Adopt the YOLOv8n-seg algorithm to perform instance segmentation on the ice and conductors in the iced pictures, and accurately obtain the total number of ice-covered pixel points P2 and the total number of conductor pixel points in the pictures; S3. Calculate the planar projection distance d between the drone and the iron tower based on the iced pictures, the coordinates of the iced pictures, the iron tower coordinates, the camera parameters, and the conductor diameter; S4. For the first use or in specific situations, perform a manual annotation operation on the iced pictures to determine the conductor area in the pictures, and count the total number of conductor mask pixels P1 obtained by manual annotation; S5. Obtain relevant parameters according to the camera imaging principle; the relevant parameters include the lens focal length f, the width w of the lens target surface, the height h of the lens target surface, the width W of the icing picture, the height H of the icing picture, the diameter X of the picture conductor in the icing picture, the direct distance Z between the lens and the actual conductor, the actual conductor diameter X, and the width W of the image formed on the target surface. L Among them, the lens focal length f is obtained through the specification information corresponding to the camera model, the width w of the lens target surface and the height h of the lens target surface are determined based on the physical properties of the camera itself, the width W of the icing picture and the height H of the icing picture are obtained through the image resolution information, and the direct distance Z between the lens and the actual conductor is measured and obtained through the positioning system carried on the unmanned aerial vehicle. S6. According to the principle of similar triangles and the relevant parameters, calculate the ratio of the diameter X of the image wire to the actual wire diameter W r ; S7. Calculate the diameter D of the actual wire based on the ratio of the image wire diameter X to the actual wire diameter W r and the total number of wire mask pixels P1; S8. Calculate the average value of the ratio of the pixel scale of the iced pictures to the corresponding width and height of the physical scale of the sensor through the camera sensor size and resolution, and calculate the actual ice-covered area S of the conductor according to the total number of ice-covered pixel points P2, the average value of the ratio of the pixel scale of the iced pictures to the corresponding width and height of the physical scale of the sensor, and the diameter D of the actual conductor; S9. Take the width W of the iced pictures as the length L of the actual conductor, and calculate the average icing thickness b of the actual conductor according to the length L of the actual conductor and the actual ice-covered area S of the conductor.

2. The method for measuring the icing thickness of a transmission line based on the YOLOv8 segmentation algorithm according to claim 1, characterized in that, The planar projection distance between the drone and the iron tower is calculated by the following method: Where d represents the planar projection distance between the drone and the iron tower, x1 and y1 are the east coordinate and north coordinate of the iron tower, and x2 and y2 are the east coordinate and north coordinate of the drone.

3. A method for measuring the icing thickness of a transmission line based on the YOLOv8 segmentation algorithm according to claim 2, characterized in that, The step of calculating the ratio of the diameter X of the image wire to the actual wire diameter W r specifically includes: S51, according to the principle of similar triangles, the first proportional relationship among the focal length f of the lens, the direct distance Z between the lens and the actual wire, the width W of the image formed on the target surface L and the diameter W of the actual wire r is: S52. According to the principle of similar triangles, the width W of the ice-covered picture, the width W of the lens target surface B , the diameter X of the picture wire in the ice-covered picture, and the width W of the image formed on the target surface L The second proportional relationship is: Calculate the ratio of the diameter X of the image wire to the actual wire diameter W according to the first proportional relationship and the second proportional relationship r as follows:

4. A method for measuring the icing thickness of a transmission line based on the YOLOv8 segmentation algorithm according to claim 3, characterized in that, The actual ice-covered area S of the conductor is obtained through the following calculation formula:

5. A method for measuring the icing thickness of a transmission line based on the YOLOv8 segmentation algorithm according to claim 3, characterized in that, The average actual ice thickness b of the conductor is calculated by the following formula:

Images