CN-122016880-A - Nondestructive testing method and system for power transmission line

CN122016880ACN 122016880 ACN122016880 ACN 122016880ACN-122016880-A

Abstract

The invention relates to the technical field of transmission line detection, in particular to a nondestructive detection method and system for a transmission line. According to the invention, the unmanned aerial vehicle is used for throwing the airborne nondestructive testing robot to the power transmission line through the special mounting device, and various line structures and the number of sub-wires are adapted. And the ground remote workstation drives the robot to move through 5G communication and performs X-ray detection to acquire high-resolution images. Based on the normalized matching degree of the X-ray image of the crimping fitting to be detected and the corresponding structural feature template, the adaptive denoising threshold value of wavelet transformation is determined, the X-ray image is denoised and divided into a steel core area and an aluminum pipe area through gray level histogram analysis, the local self-adaptation enhancement is carried out by adopting a multi-scale Retinex algorithm, and a nondestructive detection result of the power transmission line is obtained through a multi-task defect identification network fused with a geometric attention mechanism through gesture correction. The invention effectively improves the power transmission line nondestructive testing accuracy.

Inventors

ZHANG YUJIN
QU HANWEI
DAI XIN
Jiang Zaidong
WANG ZIYANG
GUO NAN
GAO XIAN
LI DANGDANG
GONG TIANTIAN
HU TAO

Assignees

国网江苏省电力有限公司扬中市供电分公司
国网江苏省电力有限公司镇江供电分公司

Dates

Publication Date: 20260512
Application Date: 20260119

Claims (10)

1. The nondestructive testing method for the power transmission line is characterized by comprising the following steps of: Extracting frequency, direction and gray scale characteristics of periodic winding textures of defect-free steel-cored aluminum strand X-ray images corresponding to different types of crimping fittings, and generating steel-cored aluminum strand texture maps of all types of crimping fittings; Based on the spatial position and the dimensional parameter of each part of each model of crimping hardware, performing feature point matching and spatial registration on the corresponding X-ray image of the nondefective steel-cored aluminum strand, and generating hardware structure position masks of each model of crimping hardware; Carrying out channel fusion on the texture atlas of the steel-cored aluminum strand of each type of crimping hardware fitting and the structural position mask of the hardware fitting to obtain a structural feature template of each type of crimping hardware fitting; acquiring an X-ray image of the crimping fitting to be detected, matching the X-ray image with a corresponding structural feature template, acquiring normalized matching degrees of the X-ray image and the structural feature template, and determining a wavelet transformation self-adaptive denoising threshold value based on the normalized matching degrees; and obtaining a nondestructive testing result of the crimping fitting to be tested based on the target X-ray image.
2. The nondestructive testing method for power transmission lines according to claim 1, wherein the process of obtaining the normalized matching degree between the X-ray image of the crimp fitting to be tested and the corresponding structural feature template comprises the following steps: Dividing an X-ray image to be detected into a plurality of sub-windows, extracting wavelet coefficient distribution in each sub-window, calculating normalized cross-correlation coefficients of the sub-windows in the same coordinate range in a matched structural feature template, and taking the normalized cross-correlation coefficients as normalized matching degrees of the sub-windows.
3. The method for non-destructive testing of a power transmission line according to claim 2, wherein the formula for determining the adaptive denoising threshold value of the wavelet transform of each sub-window based on the normalized matching degree of each sub-window is: , Wherein, the The first X-ray image of the crimping fitting to be detected Line 1 Adaptive denoising thresholds for sub-windows of columns, As a global base threshold value, In order to adjust the factor(s), The first X-ray image of the crimping fitting to be detected Line 1 Normalized cross-correlation coefficients of the sub-windows of the columns and the sub-windows of the same coordinate range in the matched structural feature templates.
4. The nondestructive testing method for power transmission lines according to claim 1, wherein the method for acquiring the X-ray image of the crimp fitting to be tested comprises the steps of: accurately throwing the airborne nondestructive testing robot to the position of the crimping fitting to be tested of the power transmission line by using the unmanned aerial vehicle through a special mounting device; And driving the airborne nondestructive testing robot to move along the power transmission line by using a ground remote workstation, and acquiring an X-ray image of the crimping fitting to be tested by adopting X-ray equipment.
5. The nondestructive testing method for power transmission lines according to claim 4, wherein the method for obtaining the nondestructive testing result of the crimp fitting to be tested based on the target X-ray image comprises the following steps: acquiring a rotation angle of a lead in an image plane based on a main edge characteristic of the lead in a current target X-ray image coordinate system in the target X-ray image; Affine transformation is carried out on the target X-ray image based on the rotation angle of the lead in the image plane and the inclination angle parameter of the robot body, so as to obtain the geometrically corrected target X-ray image; And obtaining a nondestructive testing result of the crimping fitting to be tested by passing the geometrically corrected target X-ray image through a nondestructive testing model.
6. The nondestructive testing method for the power transmission line according to claim 5, wherein the nondestructive testing model is a deep convolutional neural network, and the deep convolutional neural network comprises a main network, a geometric attention mechanism module and a classifier.
7. The nondestructive testing method for power transmission lines according to claim 6, wherein the method for obtaining the nondestructive testing result of the crimping fitting to be tested by passing the geometrically corrected target X-ray image through the nondestructive testing model comprises the following steps: extracting a deep feature map from the geometrically corrected target X-ray image through a backbone network; A structural feature template corresponding to an X-ray image of the crimping fitting to be detected passes through a geometric attention mechanism module to generate a geometric attention mask; fusing the deep feature map with the geometric attention mask to obtain a target feature map; and the target feature image is passed through a classifier to obtain a nondestructive testing result of the crimping fitting to be tested.
8. The nondestructive testing method for power transmission lines according to claim 1, wherein the method for obtaining the nondestructive testing result of the crimp fitting to be tested based on the target X-ray image further comprises: Carrying out gray level histogram analysis on the target X-ray image, and dividing the target X-ray image into a steel core area and an aluminum pipe area; aiming at different areas obtained by segmentation, carrying out local self-adaptive enhancement by adopting a multi-scale Retinex algorithm to obtain an enhanced target X-ray image, wherein the Gaussian surrounding scale parameter and the gain coefficient of the steel core area are larger than those of the aluminum pipe area; and acquiring a nondestructive testing result of the crimping fitting to be tested based on the enhanced target X-ray image.
9. The nondestructive testing method for the power transmission line according to claim 1, wherein the nondestructive testing result of the crimping fitting to be tested comprises defect type, defect position coordinates, defect size and confidence report.
10. A transmission line nondestructive testing system, comprising: The texture map construction module is used for extracting frequency, direction and gray scale characteristics of periodic winding textures of the defect-free steel-cored aluminum strand X-ray images corresponding to the crimping fittings of different types to generate texture maps of the steel-cored aluminum strands of the crimping fittings of all types; The position mask generating module is used for carrying out characteristic point matching and space registration on the corresponding X-ray images of the nondefective steel-cored aluminum strand based on the space position and the size parameter of each part of each model of crimping hardware, and generating hardware structure position masks of each model of crimping hardware; The structural feature template construction module is used for carrying out channel fusion on the texture atlas of the steel-cored aluminum strand of each type of crimping hardware fitting and the structural position mask of the hardware fitting to obtain the structural feature template of each type of crimping hardware fitting; The denoising module is used for acquiring the X-ray image of the crimping fitting to be detected and matching the X-ray image with the corresponding structural feature template, acquiring the normalized matching degree of the X-ray image and the structural feature template, and determining the adaptive denoising threshold value of wavelet transformation based on the normalized matching degree; And the detection module is used for acquiring a nondestructive detection result of the crimping fitting to be detected based on the target X-ray image.

Description

Nondestructive testing method and system for power transmission line Technical Field The invention relates to the technical field of transmission line detection, in particular to a nondestructive detection method and system for a transmission line. Background Crimping fittings such as strain clamps and splicing sleeves of overhead transmission lines are key components for ensuring the fixation of a conductive wire and the transmission of electric power, and the crimping quality of the crimping fittings directly influences the reliability and the safety of the transmission lines. Because the crimping quality problem is hidden in the hardware fitting, the traditional detection method mainly relies on X-ray flaw detection equipment operated by manual tower climbing, and the image defect is judged by naked eyes. The method is low in efficiency, the average time for detecting each hardware fitting is about 50 minutes, at least 5 persons are required to finish the detection, and safety risks such as high falling, ray radiation and electric shock exist. In addition, manual detection is limited by the skill level and operation standardization of constructors, detection consistency and accuracy are difficult to ensure, and especially in a complex line structure or a severe environment, detection difficulty is further increased, so that the detection omission rate of crimping defects is high, for example, the defect rate of a strain clamp in a three-span section is up to 29.09%. In recent years, some areas try to reduce manual overhead operation by directly mounting flaw detection equipment through an unmanned aerial vehicle based on an X-ray detection technology of the unmanned aerial vehicle, and research is also carried out by mounting a detection robot with a motion adjusting function through an unmanned aerial vehicle platform so as to realize accurate flaw detection operation on crimping fittings. When the method is operated, the unmanned aerial vehicle carries the robot to fly to the position of the target hardware fitting according to the flexible maneuvering characteristic of the unmanned aerial vehicle, the X-ray flaw detection module carried by the robot keeps the optimal detection distance and angle with the hardware fitting through the visual positioning and posture adjustment technology, and then the robot starts flaw detection equipment to collect the internal structural image of the hardware fitting, so that data support is provided for subsequent flaw identification. In the image preprocessing link, in order to reduce the interference of noise on defect judgment, a wavelet transformation technology is often adopted in the industry to carry out denoising treatment on an X-ray image, and the technology is used for solving an image signal into wavelet coefficients with different scales, inhibiting high-frequency small coefficients representing noise and reserving low-frequency large coefficients representing image details, so that the dual aims of noise filtering and image definition are realized, and a foundation is laid for defect identification. Although the detection method of the unmanned aerial vehicle-mounted robot has remarkable improvement in operation safety and efficiency, the X-ray image denoising technology based on wavelet transformation still has the defect to be solved. The key problem of the standard wavelet transformation technology widely applied in the current industry is that the complex distribution characteristics of noise and defect characteristics in X-ray images cannot be dynamically adapted by adopting a fixed wavelet basis function and a single decomposition scale for signal processing. The noise in the X-ray image is mostly presented as random high-frequency signals, and key defects such as micro cracks, undervoltage edges and the like in the compression fitting are presented, and the characteristic signals are also in a high-frequency band, and the signal strength is often similar to the noise. When the standard wavelet transformation is used for screening the high-frequency signals, an effective distinguishing mechanism is lacked, the high-frequency defect characteristics are easily misjudged as noise, and excessive smoothing processing is carried out, so that the defect characteristics are weakened or even completely eliminated. The technical defect directly causes missed detection of hidden danger in the crimping fitting, so that the detection method of the unmanned aerial vehicle carrying robot is greatly reduced in accuracy, the technical advantages of the unmanned aerial vehicle carrying robot are difficult to fully develop, and reliable guarantee cannot be provided for quality detection of the crimping fitting of the power transmission line. Disclosure of Invention Therefore, the technical problem to be solved by the invention is to overcome the defect that the nondestructive testing precision of the power transmission line is poor because the existing detection meth