A method, apparatus, terminal equipment, and storage medium for assessing the aging risk of composite insulator skirts.

Abstract

This invention discloses a method, apparatus, terminal equipment, and storage medium for assessing the aging risk of composite insulator skirts, belonging to the technical field of power equipment risk assessment. The method includes: acquiring visible light and multispectral images of the composite insulator; generating a binary mask image using an image segmentation model and extracting water-related indicators to assess the hydrophobicity level; segmenting the skirt edge contour from the visible light image and fitting a curve to calculate the curvature integral value and the fractal dimension of the crack region; calculating the brightness value and damage index based on the reflectance of the multispectral image; and determining the aging level of the composite insulator based on the hydrophobicity level, curvature integral value, fractal dimension, and damage index. Implementing this invention can improve the efficiency and accuracy of assessing the aging risk of composite insulator skirts.

Description

Technical Field

[0001] This invention relates to the field of power equipment risk assessment technology, and in particular to a method, apparatus, terminal equipment and storage medium for assessing the aging risk of composite insulator skirts. Background Technology

[0002] Composite insulators are important electrical equipment, serving the dual functions of electrical insulation and mechanical support in power transmission and distribution lines and substations. However, after natural disasters such as wildfires, the high temperatures, thermal radiation, and ash deposits caused by the fires can lead to multiple aging damages on the sheds of composite insulators. These damages manifest as decreased or lost hydrophobicity, shed melting and deformation, surface crack propagation, and deterioration of the material's chemical composition. These damages significantly reduce the electrical insulation performance and mechanical strength of the insulators, and in severe cases, can cause flashover, breakage, and other faults, threatening the safe operation of the power grid.

[0003] In existing technologies, inspections mainly rely on patrol personnel to conduct ground inspections or climb towers for close-up inspections using binoculars. This method is slow and cannot meet the need for rapid inspection of a large area and multiple towers after a wildfire. Furthermore, the reliance on patrol personnel for inspections is highly subjective and not accurate enough. Summary of the Invention

[0004] This invention provides a method, apparatus, terminal equipment, and storage medium for assessing the aging risk of composite insulator skirts, in order to solve the problems of slow speed and low accuracy in the prior art for assessing the aging risk of composite insulator skirts.

[0005] To address the aforementioned technical problems, embodiments of the present invention provide a method for assessing the aging risk of composite insulator skirts, comprising: Acquire visible light and multispectral images of the composite insulator to be evaluated; The visible light image is input into an image segmentation model to generate a binary mask image representing the pixel type of each pixel in the visible light image; wherein, the pixel type includes water droplets and background; Identify independent connected regions in a binary mask image, and extract the percentage of water-covered area, the number of water droplets, and the average roundness index based on the connected regions; evaluate the hydrophobicity level based on the percentage of water-covered area, the number of water droplets, and the average roundness index. The edge contour and crack location of the composite insulator skirt are segmented and extracted from the visible light image; A curve is fitted based on the set of points of the edge contour; the absolute value of the principal curvature of the points on the curve is integrated to obtain the curvature integral value; The fractal dimension of the image of the region where the crack is located is calculated using the box counting method; The reflectance of the composite insulator skirt surface in multiple bands is identified based on multispectral images, and the current brightness value is calculated based on the reflectance. The damage index of the composite insulator skirt surface is calculated based on each reflectance, the current brightness value, and the insulator reference brightness value. The aging level of the composite insulator is determined based on the hydrophobicity level, the curvature integral value, the fractal dimension, and the damage index.

[0006] Understandably, compared with existing technologies, this invention analyzes the relevant parameters of water adhesion on the surface of composite insulators and matches the hydrophobicity level. It combines the curvature integral value, crack fractal dimension, and surface damage index to assess the aging risk of composite insulators, achieving full coverage of multi-dimensional indicators and improving the accuracy of aging assessment. Furthermore, this invention achieves automated batch testing, which can effectively reduce the intensity of manual inspection, significantly improve inspection efficiency, avoid the problem of strong subjectivity caused by relying on the judgment of inspection personnel, and improve the detection accuracy.

[0007] As a preferred embodiment, before inputting the visible light image into the image segmentation model, the method further includes: Adjust the visible light image of the composite insulator to a size that matches the image segmentation model; The pixel values ​​of the resized visible light image are normalized. The normalized image is converted to the HSV color space to obtain the preprocessed visible light image.

[0008] Understandably, preprocessing the visible light image before inputting it into the image segmentation model can uniformly adjust the image to an input format suitable for the model. Size adjustment ensures the image meets the model's input requirements, and pixel value normalization improves the model's numerical stability and convergence. Furthermore, converting the normalized image to the HSV color space highlights the color and brightness differences between water droplets and the background, reducing the impact of lighting, environment, and other interference factors on the segmentation results. The preprocessed image facilitates accurate identification and differentiation of water droplet pixels from background pixels by the image segmentation model, effectively improving the generation accuracy of the binary mask image. This provides a reliable data foundation for the accurate extraction of subsequent hydrophobicity-related indicators, thereby enhancing the stability and accuracy of the entire aging risk assessment method.

[0009] As a preferred approach, the reflectance of the composite insulator skirt surface in multiple bands is identified based on multispectral images, and the current brightness value is calculated based on the reflectance. Based on each reflectance, the current brightness value, and the insulator's reference brightness value, a damage index for the composite insulator skirt surface is calculated, including: Based on the reflectivity of the composite insulator skirt surface in a preset first band and the reflectivity in a preset second band, the normalized difference of the reflectivity of the skirt surface in the two bands is determined; wherein, the wavelength of the first band is less than the wavelength of the second band. The relative degree of color change on the surface of the composite insulator skirt is determined based on the current brightness value and the insulator reference brightness value; The damage index of the composite insulator skirt surface is calculated based on the normalized difference of reflectivity and the relative change in color of the composite insulator skirt surface.

[0010] It is understood that this embodiment obtains the reflectivity of the umbrella skirt surface in different bands through multispectral images and calculates the damage index by combining the brightness value. This can transform the complex state of the umbrella skirt surface, such as aging, pollution, and discoloration, into a unified and quantifiable numerical index, providing a reliable quantitative basis for subsequent aging risk level determination and improving the accuracy and repeatability of the assessment results.

[0011] As a preferred approach, before identifying independent connected regions in the binary mask image and extracting the percentage of water-covered area, the number of water droplets, and the average roundness index based on the connected regions, the following steps are also included: An opening operation is performed on the binary mask image to obtain a denoised image after removing tiny noise points and smoothing the edges of water droplets. The denoised image is subjected to a closing operation to fill holes in the denoised image with an area smaller than a preset area, resulting in a morphologically processed binary mask image.

[0012] Understandably, pre-processing the binary mask image using morphological methods can effectively remove minute noise points, smooth the edge contours of water droplet regions, and fill in tiny pores within the droplets, transforming scattered areas into continuous and complete water droplet regions. This processing step eliminates interference and optimizes image quality, providing a clear and reliable image foundation for subsequent accurate identification of independent connected regions, extraction of water-attached area ratio, water droplet quantity, and average roundness, thereby improving the accuracy and stability of hydrophobicity level assessment.

[0013] As a preferred approach, the fractal dimension of the image of the region where the crack is located is calculated using the box counting method, including: The image of the crack location area is covered by grids of the first preset size and the second preset size, respectively; The number of first grids required when covering the image of the crack location area with a grid of the first preset size, and the number of second grids required when covering the image of the crack location area with a grid of the second preset size; Based on the first grid number and the second grid number, determine the first change in the number of grids required to cover the crack when the grid size changes from the first preset size to the second preset size; Based on the first preset size and the second preset size, calculate the second change in the grid size at the logarithmic scale; The fractal dimension of the image of the region where the crack is located is determined based on the first change and the second change.

[0014] Understandably, calculating the fractal dimension of a crack region image using the box counting method can transform the complex morphology of the crack into a quantifiable numerical indicator. A higher fractal dimension indicates a more complex, denser, and more irregular crack network, while a lower dimension indicates a relatively simpler crack morphology. This quantitative indicator provides an objective and stable basis for subsequently assessing the aging degree and damage level of composite insulator skirts, avoiding the subjectivity of manual crack observation and improving the accuracy and repeatability of the entire aging risk assessment scheme.

[0015] As a preferred embodiment, the aging level of the composite insulator is determined based on the hydrophobicity level, the curvature integral value, the fractal dimension, and the damage index, including: Based on the aging index calculation formula, the aging index is calculated according to the hydrophobicity level, curvature integral value, fractal dimension and damage index. The formula for calculating the aging index is as follows: CAI=α⋅((HC-1) / 6)+β⋅(K_int / K_intmax )+γ⋅((D-1) / (D_max-1))+δ⋅FPI In the formula, CAI is the aging index, HC is the hydrophobicity level, α is the weighting coefficient corresponding to the hydrophobicity level, K_int is the curvature integral value, β is the weighting coefficient corresponding to the curvature integral value, D is the fractal dimension, γ is the weighting coefficient corresponding to the fractal dimension, FPI is the damage index, δ is the weighting coefficient corresponding to the damage index, K_intmax is the maximum value of the curvature integral value reached under extreme melting deformation of the composite insulator skirt, and D_max is the maximum value of the fractal dimension reached under severe cracking. The aging level of composite insulators is determined based on the aging index.

[0016] Understandably, this embodiment uses an aging index calculation formula to weight and fuse four key indicators—hydrophobicity level, curvature integral value, fractal dimension, and damage index—according to their respective weight coefficients, resulting in a unified aging index. The aging level is then determined based on this index. This process transforms the originally scattered multi-dimensional detection results into a quantifiable and comparable comprehensive value, objectively and comprehensively reflecting the overall aging status of composite insulators in terms of hydrophobicity, edge melting deformation, crack complexity, and surface damage. This avoids the one-sidedness of single-indicator evaluation, improves the accuracy and scientific nature of aging risk assessment, and provides a stable and reliable quantitative basis for subsequent operation and maintenance decisions.

[0017] Accordingly, the present invention also provides a composite insulator skirt aging risk assessment device, comprising: an image acquisition module, a binary mask image generation module, a hydrophobicity level acquisition module, an edge contour and crack location extraction module, a curvature integral value acquisition module, a fractal dimension calculation module, a damage index calculation module, and an aging level determination module. The image acquisition module is used to acquire visible light and multispectral images of the composite insulator to be evaluated. The binary mask generation module is used to input a visible light image into an image segmentation model and generate a binary mask that represents the pixel type of each pixel in the visible light image; the pixel types include water droplets and background. The hydrophobicity level acquisition module is used to identify independent connected regions in a binary mask image, extract the proportion of water-attached area, the number of water droplets, and the average roundness index based on the connected regions, and evaluate the hydrophobicity level based on the proportion of water-attached area, the number of water droplets, and the average roundness index. The edge contour and crack location extraction module is used to segment and extract the edge contour and crack location of the composite insulator skirt from the visible light image; The curvature integral value acquisition module is used to fit a curve based on the point set of the edge contour; and to integrate the absolute value of the principal curvature of the points on the curve to obtain the curvature integral value. The fractal dimension calculation module is used to calculate the fractal dimension of the image of the region where the crack is located using the box counting method. The damage index calculation module is used to identify the reflectance and current brightness value of the composite insulator skirt surface in multiple bands based on multispectral images; and to calculate the damage index of the composite insulator skirt surface based on each reflectance, current brightness value and insulator reference brightness value. The aging level determination module is used to determine the aging level of composite insulators based on hydrophobicity level, curvature integral value, fractal dimension, and damage index.

[0018] As a preferred embodiment, it also includes: an image preprocessing module; The image preprocessing module is used to adjust the visible light image of the composite insulator to a size that matches the image segmentation model before inputting the visible light image into the image segmentation model; The pixel values ​​of the resized visible light image are normalized. The normalized image is converted to the HSV color space to obtain the preprocessed visible light image.

[0019] Accordingly, the present invention also provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor implements the composite insulator skirt aging risk assessment method as described above when executing the computer program.

[0020] Accordingly, the present invention also provides a computer-readable storage medium, the computer-readable storage medium including a stored computer program; wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to perform the composite insulator skirt aging risk assessment method as described above. Attached Figure Description

[0021] To more clearly illustrate the technical solution of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0022] Figure 1 A flowchart illustrating the steps of a method for assessing the aging risk of composite insulator skirts according to an embodiment of the present invention; Figure 2 This is a schematic diagram of a composite insulator skirt aging risk assessment device provided in an embodiment of the present invention. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0024] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the drawings are intended to cover non-exclusive inclusion.

[0025] In the description of the embodiments of this application, technical terms such as "first" and "second" are used only to distinguish different objects and should not be construed as indicating or implying relative importance or implicitly specifying the number, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly defined.

[0026] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0027] In the description of the embodiments in this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0028] In the description of the embodiments of this application, the term "multiple" refers to two or more (including two), similarly, "multiple sets" refers to two or more (including two sets), and "multiple pieces" refers to two or more (including two pieces).

[0029] In the description of the embodiments of this application, unless otherwise expressly specified and limited, technical terms such as "installation," "connection," "joining," and "fixing" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. For those skilled in the art, the specific meaning of the above terms in the embodiments of this application can be understood according to the specific circumstances.

[0030] To address the issues of slow speed and low accuracy in assessing the aging risk of composite insulator skirts in existing technologies, this application provides a method for assessing the aging risk of composite insulator skirts.

[0031] Please refer to Figure 1 The present invention provides a method for assessing the aging risk of composite insulator skirts, comprising the following steps S101-S108: S101: Acquire visible light and multispectral images of the composite insulator to be evaluated; In this embodiment, acquiring visible light images of the composite insulator to be evaluated facilitates subsequent assessment of the hydrophobicity of the skirt, contour deformation analysis, and crack detection based on the visible light images. Acquiring multispectral images of the composite insulator to be evaluated facilitates subsequent calculation of the material damage index by obtaining the reflectivity of different bands on the skirt surface based on the multispectral images. Acquiring these two types of images before proceeding with other steps provides a data basis for subsequent aging level determination.

[0032] In one possible implementation, visible light and multispectral images of the composite insulator to be evaluated can be acquired during a specific time period. For example, the specific time period could be 4:00-6:00. It is understood that when evaluating the hydrophobicity of a material, it is generally necessary to first make the surface of the material to be evaluated uniformly covered with water, and then evaluate its hydrophobicity by observing its water droplet shape, spreading state and other characteristics. In this embodiment, the morning dew formed naturally in the early morning is used to make the surface of the composite insulator covered with water, without the need to make the surface of the composite insulator covered with water through other means, thus saving manpower and resources.

[0033] In one possible implementation, visible light and multispectral images of the composite insulator to be evaluated can be collected by drones, or by fixedly deployed automatic image acquisition devices. It is understood that acquiring visible light and multispectral images through drones or fixed automatic acquisition devices can achieve non-contact, automated, and multi-scenario image acquisition, avoiding the safety risks of manual tower climbing operations, reducing human detection errors, improving acquisition efficiency and data accuracy, providing a reliable data source for subsequent aging feature extraction and risk assessment, and helping to achieve intelligent and routine monitoring of aging risks of composite insulators.

[0034] S102: Input the visible light image into the image segmentation model to generate a binary mask image representing the pixel type of each pixel in the visible light image; wherein, the pixel type includes water droplets and background.

[0035] In this embodiment, after acquiring the visible light image and multispectral image of the composite insulator to be evaluated in step S101, the acquired visible light image is input into the image segmentation model. Based on the image segmentation model, a binary mask image representing the pixel type is obtained. It can be understood that by inputting the preprocessed visible light image into the image segmentation model, water droplets and background can be accurately separated from the complex visible light image, realizing the accurate extraction of the water droplet region and eliminating background interference such as the insulator skirt body and environmental debris. This provides a clear image basis for subsequent identification of connected regions, extraction of water-attached area ratio, number of water droplets and average roundness, etc., ensuring the accuracy of subsequent hydrophobicity level assessment.

[0036] Preferably, before inputting the visible light image into the image segmentation model, the method further includes: adjusting the visible light image of the composite insulator to a size that matches the image segmentation model; normalizing the pixel values ​​of the resized visible light image; and converting the normalized image to the HSV color space to obtain the preprocessed visible light image.

[0037] Understandably, preprocessing the visible light image before inputting it into the image segmentation model can uniformly adjust the image to an input format suitable for the model. Size adjustment ensures the image meets the model's input requirements, and pixel value normalization improves the model's numerical stability and convergence. Furthermore, converting the normalized image to the HSV color space highlights the color and brightness differences between water droplets and the background, reducing the impact of lighting, environment, and other interference factors on the segmentation results. The preprocessed image facilitates accurate identification and differentiation of water droplet pixels from background pixels by the image segmentation model, effectively improving the generation accuracy of the binary mask image. This provides a reliable data foundation for the accurate extraction of subsequent hydrophobicity-related indicators, thereby enhancing the stability and accuracy of the entire aging risk assessment method.

[0038] In one possible implementation, the image segmentation model can be a U-Net model. The training process of the U-Net model is as follows: A pre-processed composite insulator image is input; the encoder extracts features; the decoder outputs a probability map of the water-covered region; the probability map is compared with the labeled real water-covered regions; the error is calculated; and the model parameters are optimized through backpropagation. Finally, the preliminary segmentation result is post-processed, and an optimized mask of the water-covered region is output, completing the model training. After pre-training the U-Net model, this embodiment will use the trained U-Net model to identify water droplet regions and generate binary mask images. Specifically, the pre-processed image to be evaluated is input into the U-Net model. The encoder of the U-Net model is responsible for extracting image features, while the decoder is responsible for accurately locating and reconstructing the precise contours of the water-covered region, thereby outputting a probability map of the same size as the input image. The value of each pixel in the map represents the probability that the location is a water droplet. Then, a binarization threshold is set to convert the probability map into a binary mask image that is either black or white. Pixels with a probability higher than the binarization threshold are classified as water droplets; otherwise, they are considered background. This means that using the U-Net model for training and water droplet region segmentation allows the encoder to efficiently extract image features and the decoder to accurately reconstruct the contours of the water-covered region, thus achieving automatic identification and precise localization of the water droplet region. During the training phase, the model continuously optimizes its parameters using labeled data, ensuring the accuracy and robustness of the segmentation results in practical applications. By using probability... Figure 2 Value-enhanced binary mask images can effectively distinguish water droplets from the background, eliminate environmental interference, and provide a clear and reliable image basis for subsequent connected component analysis and hydrophobicity level assessment, thereby improving the accuracy and stability of the entire aging assessment process.

[0039] S103: Identify independent connected regions in the binary mask image, extract the proportion of water-attached area, the number of water droplets, and the average roundness index based on the connected regions; evaluate the hydrophobicity level based on the proportion of water-attached area, the number of water droplets, and the average roundness index.

[0040] This embodiment, based on the binary mask image representing pixel type obtained in step S102, identifies independent connected regions in the binary mask image, denoting each independent connected region as a water droplet region. The number of all connected regions is counted to obtain the number of water droplets. The ratio of the total pixel area of ​​all water droplet regions to the total area of ​​the umbrella skirt is calculated and recorded as the water-attached area ratio. Simultaneously, the roundness of each independent connected region is calculated, and the average roundness is obtained by averaging all roundness values. After obtaining the three indicators—water-attached area ratio, number of water droplets, and average roundness—the three indicators are comprehensively judged according to preset evaluation rules to determine the hydrophobicity level of the composite insulator umbrella skirt. For example, the hydrophobicity level of the composite insulator umbrella skirt can be determined using the following methods, as detailed in the table below: Table 1. Example of determining the hydrophobicity level of composite insulator skirts As shown in the table, for example, the average roundness C of water droplets is used as the core judgment index, and the C value range of HC1-HC7 is matched to obtain the initial hydrophobicity level. Then, the proportion of water-covered area RA is used as a secondary index to verify whether the initial hydrophobicity level is correct. If RA is in the same range or adjacent to the initial level of C, then RA verification is passed, and the initial hydrophobicity level determined by the average roundness C remains valid. Next, the number of water droplets N is used as an auxiliary verification index. If N is in the same range or adjacent to the initial hydrophobicity level after RA correction, then the verification is passed, and the level is determined to be the final judgment level.

[0041] Understandably, by identifying connected regions in a binary mask image and extracting indicators such as the percentage of water-covered area, the number of water droplets, and the average roundness, and then assessing the hydrophobicity level accordingly, the water-covered morphology on the umbrella skirt surface can be transformed into quantitative features, enabling an objective and accurate evaluation of hydrophobicity. This step provides key performance indicators for subsequent comprehensive assessment of the umbrella skirt's aging degree, effectively improving the reliability and scientific rigor of aging risk assessment.

[0042] In one possible implementation, before identifying independent connected regions in the binary mask image and extracting the proportion of water-attached area, the number of water droplets, and the average roundness index based on the connected regions, the method further includes: performing an opening operation on the binary mask image to obtain a denoised image after removing small noise points and smoothing the edges of water droplets; performing a closing operation on the denoised image to fill holes in the denoised image with an area smaller than a preset area to obtain a morphologically processed binary mask image.

[0043] In this embodiment, before identifying independent connected regions in the binary mask image, a preprocessing step is performed on the binary mask image. Specifically, the first step is to perform an erosion-dilation process (i.e., opening operation) on the binary mask image. First, a structuring element of appropriate size is selected, and this structuring element is moved pixel by pixel on the binary mask image. At each position, only when the area completely covered by the structuring element consists entirely of water droplet pixels is the pixel at that position set to the corresponding color of the water droplet in the binary mask image. As long as there is even one point in the area covered by the structuring element that is not a water droplet, the process continues. For each water droplet pixel, the pixel at that location is changed to the corresponding color of the background in the binary mask. This removes small, scattered noise points in the binary mask that are not real water droplets. After erosion, a dilation operation is performed. Specifically, the structuring element is used to move pixel by pixel on the image where noise points have been removed. As long as there is even a tiny water droplet pixel in the area covered by the structuring element, the pixel at that location is changed to the color corresponding to the water droplet. This fills in the small gaps and burrs that appear at the water droplet boundaries caused by the erosion stage, making the edges of the water droplets smoother and more even. The second step involves performing a dilation-erosion operation (i.e., closing) on ​​the binary mask image. Specifically, using the same structuring element, the image is moved pixel by pixel. Whenever a droplet pixel is within the area covered by the structuring element, that location is changed to the color corresponding to the droplet. This fills in the tiny, invisible black holes within the droplet area with the droplet's color, connecting the previously scattered and discontinuous droplet areas into a continuous white droplet region. Finally, an erosion operation is performed, again using the same structuring element, moving pixel by pixel. Only when the area covered by the structuring element is entirely composed of droplet pixels is the droplet's color retained; otherwise, it's changed to the background color. This step is to smooth out the excess parts extending from the dilated droplet boundaries, restoring the droplet's shape to its original form, preventing distortion from the previous dilation. In essence, this pre-processing of the binary mask image effectively removes tiny noise points, smooths the edges of the droplet region, and fills in the tiny holes within the droplets, transforming scattered areas into continuous, complete droplet regions. This processing step can eliminate interference and optimize image quality, providing a clear and reliable image basis for subsequent accurate identification of independent connected regions, extraction of water-attached area ratio, water droplet quantity and average roundness, thereby improving the accuracy and stability of hydrophobicity level assessment.

[0044] S104: Segment and extract the edge contour and crack location of the composite insulator skirt from the visible light image; In this embodiment, a visible light image of the composite insulator is first acquired. The U-net target edge detection algorithm is then used to precisely segment the visible light image, extracting the complete edge contour of each composite insulator skirt. This edge contour consists of a series of ordered pixel coordinate points, which clearly delineate the outer boundary of each skirt. Clearly delineating the outer boundary of each skirt is crucial for accurately locating the specific location of the crack. By accurately determining the effective area of ​​each skirt, the skirt area is distinguished from the background and other components, avoiding interference from irrelevant areas in subsequent crack detection and contour analysis. This ensures that crack location and state assessment are performed only within the actual skirt area, improving the accuracy and reliability of the detection results. Simultaneously, within each extracted skirt area, the algorithm performs binarization processing on the located crack area to distinguish it from the surrounding background, thereby extracting the clear area where the crack is located from the image. This facilitates accurate analysis of the crack's location, morphology, and other information in subsequent steps.

[0045] S105: Fit a curve to the point set of the edge contour; integrate the absolute value of the principal curvature of the points on the curve to obtain the curvature integral value; In this embodiment, firstly, the discrete contour point set extracted in step S104 and scattered along the edge of the umbrella skirt is subjected to smoothing and interpolation processing. Smoothing eliminates minor fluctuations and burrs between these discrete points, making the point set more regular. Interpolation adds appropriate points between adjacent discrete points, making the originally scattered point set more closely connected. Through these two processes, the originally discrete contour point set is fitted into a continuous and smooth spatial plane curve without breaks. This curve is the complete edge contour curve of the umbrella skirt. After processing, based on the principles of differential geometry, the curvature value is calculated for each point on this contour curve, obtaining the curvature data corresponding to all points on the curve. Then, along the entire contour curve of the umbrella skirt, the principal curvature of each point is found, the absolute value of each principal curvature is taken, and the absolute values ​​of these principal curvatures are integrated. The final result is the curvature integral, and the formula for calculating the curvature integral is as follows: (1) In the formula, It is the curvature integral, used to quantify the degree of melting deformation of the composite insulator skirt. It is the total length of the outline curve of the composite insulator skirt, in millimeters (mm). , These are the two principal curvatures at a point on the profile curve of the composite insulator skirt, expressed in millimeters, reflecting the degree of curvature at that point. It is understood that this embodiment, by smoothing and interpolating the discrete profile point set, can eliminate burrs and fluctuations at the profile points, fitting scattered points into a continuous, complete, and smooth skirt edge curve, ensuring the accuracy of subsequent calculations. Furthermore, by calculating the curvature and principal curvatures at each point on the profile, and integrating the absolute values ​​of the principal curvatures to obtain the curvature integral, the bending changes at the skirt edge can be converted into quantitative values. This objectively and accurately measures the irregularity and melting deformation of the skirt edge, providing a stable and reliable quantitative basis for subsequent judgments of skirt aging and damage, improving the scientific rigor and accuracy of the evaluation results.

[0046] S106: Calculate the fractal dimension of the image of the region where the crack is located using the box counting method; In one possible implementation, the fractal dimension of the image of the crack location region is calculated using box counting, including: covering the image of the crack location region with grids of a first preset size and a second preset size, respectively; counting the number of first grids required when covering the image of the crack location region with grids of the first preset size, and the number of second grids required when covering the image of the crack location region with grids of the second preset size; determining a first change in the number of grids required to cover the crack when the grid size changes from the first preset size to the second preset size based on the first grid number and the second grid number; calculating a second change in the grid size on a logarithmic scale based on the first preset size and the second preset size; and determining the fractal dimension of the image of the crack location region based on the first change and the second change.

[0047] For example, after obtaining the binarized image of the crack, the formula for calculating its fractal dimension using the box counting method is as follows: (2) In the formula, The fractal dimension of a crack is used to quantify its complexity, spatial distribution density, or propagation state. A higher fractal dimension generally indicates a more complex and denser crack network. For use of size The number of first grid cells required when the grid covers the image of the area where the crack is located. For use of size The number of second grids required when the grid covers the image of the area where the crack is located. Indicates the first preset size. Indicates the second preset size, and , The quantitative relationship between them usually satisfies: , The first change in the number of meshes required to cover the crack is represented when the mesh size changes from a first preset size to a second preset size. This represents the second change in grid size on a logarithmic scale. Understandably, calculating the fractal dimension of a crack region image using box counting transforms the complex morphology of the crack into a quantifiable numerical indicator. A higher fractal dimension indicates a more complex, denser, and irregularly propagating crack network; conversely, a lower fractal dimension indicates a relatively simpler crack morphology. This quantitative indicator provides an objective and stable basis for subsequent assessments of the aging degree and damage level of composite insulator skirts, avoiding the subjectivity of manual crack observation and improving the accuracy and repeatability of the entire aging risk assessment scheme.

[0048] S107: Identify the reflectance of the composite insulator skirt surface in multiple bands based on multispectral images, calculate the current brightness value based on the reflectance; calculate the damage index of the composite insulator skirt surface based on each reflectance, the current brightness value, and the reference brightness value of the composite insulator. In one possible implementation, the reflectance of the composite insulator skirt surface in multiple bands is identified based on multispectral images, and the current brightness value is calculated based on the reflectance. Based on each reflectance, the current brightness value, and the composite insulator's reference brightness value, a damage index of the composite insulator skirt surface is calculated, including: determining the normalized difference of the reflectance of the skirt surface in two bands based on the reflectance of the composite insulator skirt surface in a preset first band and a preset second band; wherein the wavelength of the first band is shorter than the wavelength of the second band; determining the relative degree of color change of the composite insulator skirt surface based on the current brightness value and the insulator's reference brightness value; and calculating the damage index of the composite insulator skirt surface based on the normalized difference of reflectance and the relative degree of color change of the composite insulator skirt surface.

[0049] Specifically, ash generated in wildfire environments can chemically react with the silicone rubber sheds or core rods of insulators, leading to structural damage and performance degradation. This damage is usually gradual, and long-term accumulation can cause insulation failure and decreased mechanical strength. Therefore, it is necessary to quantify the surface damage based on multispectral diagnostic indices. For example, the formula for calculating the damage index of composite insulator shed surfaces is as follows: (3) In the formula, This indicates the reflectivity of the composite insulator skirt surface at a preset first wavelength of 450nm. This indicates the reflectivity of the composite insulator skirt surface in the preset second wavelength band of 850nm. Indicates the current brightness value. This indicates the reference brightness value for composite insulators, which is typically greater than or equal to 85. Characterizing the relative degree of color change on the surface of the composite insulator skirt. This refers to the normalized difference in reflectance of the umbrella skirt surface in two bands. The 0.7 and 0.3 in the formula are two weighting coefficients used to allocate the contribution ratio of the two indicators in the final damage index FPI.

[0050] It is understood that this embodiment obtains the reflectivity of the umbrella skirt surface in different bands through multispectral images and calculates the damage index (FPI) by combining the brightness value. This can transform the complex state of the umbrella skirt surface, such as aging, pollution, and discoloration, into a unified and quantifiable numerical index, providing a reliable quantitative basis for subsequent aging risk level determination and improving the accuracy and repeatability of the assessment results.

[0051] S108: Determine the aging level of composite insulators based on hydrophobicity level, curvature integral value, fractal dimension, and damage index.

[0052] In this embodiment, four evaluation indicators are first obtained: the hydrophobicity level of the composite insulator, the integral value of curvature corresponding to the skirt profile, the fractal dimension of the crack region, and the damage index of the skirt surface. The hydrophobicity level characterizes the change in the hydrophobic properties of the skirt surface; the integral value of curvature quantifies the melting deformation and irregularity of the skirt edge; the fractal dimension characterizes the complexity and propagation state of the crack network; and the damage index comprehensively reflects the degree of aging, contamination, and color change of the skirt surface. Subsequently, these four indicators are used as inputs, and the aging state of the composite insulator is comprehensively judged according to preset evaluation rules or threshold ranges to determine its corresponding aging level. This transforms multi-dimensional and dispersed detection and evaluation results into intuitive and actionable engineering conclusions, allowing maintenance personnel to directly judge the health status of the composite insulator based on the aging level and quickly formulate targeted maintenance strategies, avoiding the tedious process of analyzing multiple independent indicators one by one.

[0053] In one possible implementation, the aging level of the composite insulator is determined based on the hydrophobicity level, curvature integral value, fractal dimension, and damage index, including: Based on the aging index calculation formula, the aging index is calculated according to the hydrophobicity level, curvature integral value, fractal dimension, and damage index; the calculation formula for the aging index is: (4) In the formula, The aging index, Hydrophobicity level The weighting coefficients corresponding to the hydrophobicity level are: The value is the integral of curvature. These are the weighting coefficients corresponding to the integral value of curvature. For fractal dimension, These are the weighting coefficients corresponding to the fractal dimension. The damage index, These are the weighting coefficients corresponding to the damage index. This represents the maximum value of the curvature integral achieved under extreme melting deformation of the umbrella skirt. This represents the maximum value of the fractal dimension achieved under severe crack conditions.

[0054] The following is a further explanation of formula (4). This formula aims to normalize the original feature indicators of different dimensions, such as hydrophobicity level and curvature integral value, to the [0,1] interval, to ensure that the weight ratio of each indicator is reasonable during weighted fusion, and finally to achieve the quantitative comparability of CAI. Specifically, Characterization will The mapping is a linear sequence of 0–6. Dividing by 6 is to normalize the term to the interval [0,1], so that... When this term is 1, it indicates that hydrophobicity is completely lost; curvature integral value Reflecting the degree of contour deformation, assuming extreme melting conditions Possibly reach Left and right, divided by This makes the value close to 1.0 in extreme cases, facilitating integration with other metrics on comparable scales; fractal dimension The theoretical baseline is 1.0 (no cracks), assuming a severely cracked network. Can be increased to about, This represents the dimensionality increment caused by the crack. Divide by When for At that time, the value of this item was 1.0.

[0055] In one possible implementation, the aging level of the composite insulator can be determined based on the aging index.

[0056] Understandably, this embodiment uses the aging index calculation formula to weight and fuse four key indicators—hydrophobicity level, curvature integral value, fractal dimension, and damage index—according to their respective weight coefficients, resulting in a unified aging index (CAI). The aging level is then determined based on this index. This process transforms the originally scattered multi-dimensional detection results into a quantifiable and comparable comprehensive value, objectively and comprehensively reflecting the overall aging status of composite insulators in terms of hydrophobicity, edge melting deformation, crack complexity, and surface damage. This avoids the one-sidedness of single-indicator evaluation, improves the accuracy and scientific nature of aging risk assessment, and provides a stable and reliable quantitative basis for subsequent operation and maintenance decisions.

[0057] In one possible implementation, the aging level of the composite insulator is determined based on the aging index, as shown in Table 2 below: Table 2. Standard for Classification of Aging Grades of Composite Insulators Specifically, when 0 ≤ CAI ≤ 0.3, the CAI level is determined as follows: When 0.3 < CAI ≤ 0.6, the CAI level is determined to be... When CAI > 0.6, the CAI level is determined to be... .

[0058] It is understood that this embodiment directly maps continuous aging index values ​​to intuitive aging levels by using a preset aging index CAI threshold range, allowing maintenance personnel to quickly grasp the aging status of insulators without having to interpret complex index values.

[0059] In one possible implementation, after determining the aging level of the composite insulator, corresponding maintenance measures can be implemented based on the aging level. For details, please refer to Table 3 below: Table 3. Determination of Aging Level and Maintenance Strategy for Composite Insulators When the aging level is During this period, annual inspections are conducted, and composite insulators are checked on an annual cycle. When the aging level reaches [value missing], [the inspection is carried out]. Quarterly monitoring and surface repair are performed, with composite insulators monitored quarterly and the shed surfaces repaired accordingly. When the aging level is... If necessary, the composite insulator should be replaced immediately.

[0060] Understandably, after determining the aging level of the composite insulator, this embodiment maps different aging levels to specific and actionable maintenance measures, directly transforming the results of the aging risk assessment into targeted operation and maintenance action plans. This allows the entire aging risk assessment plan to be transformed from technical analysis into engineering practice, significantly improving the efficiency and scientific nature of operation and maintenance decisions and ensuring the stable operation of the power system.

[0061] Accordingly, please refer to Figure 2 The present invention provides a composite insulator skirt aging risk assessment device, comprising: an image acquisition module 201, a binary mask image generation module 202, a hydrophobicity level acquisition module 203, an edge contour and crack location extraction module 204, a curvature integral value acquisition module 205, a fractal dimension calculation module 206, a damage index calculation module 207, and an aging level determination module 208.

[0062] Image acquisition module 201 is used to acquire visible light and multispectral images of the composite insulator to be evaluated; Binary mask generation module 202 is used to input a visible light image into an image segmentation model and generate a binary mask that represents the pixel type of each pixel in the visible light image; wherein, the pixel type includes water droplets and background; The hydrophobicity level acquisition module 203 is used to identify independent connected regions in a binary mask image, extract the water-attached area ratio, water droplet quantity, and average roundness index based on the connected regions, and evaluate the hydrophobicity level based on the water-attached area ratio, water droplet quantity, and average roundness index. The edge contour and crack location extraction module 204 is used to segment and extract the edge contour and crack location of the composite insulator skirt from the visible light image. The curvature integral value acquisition module 205 is used to fit a curve based on the point set of the edge contour; and to integrate the absolute value of the principal curvature of the points on the curve to obtain the curvature integral value. Fractal dimension calculation module 206 is used to calculate the fractal dimension of the image of the region where the crack is located by box counting method; The damage index calculation module 207 is used to identify the reflectance and current brightness value of the composite insulator skirt surface in multiple bands based on the multispectral image; and to calculate the damage index of the composite insulator skirt surface based on each reflectance, the current brightness value and the insulator reference brightness value. The aging level determination module 208 is used to determine the aging level of composite insulators based on hydrophobicity level, curvature integral value, fractal dimension and damage index.

[0063] As a preferred embodiment, the system further includes: an image preprocessing module; the image preprocessing module is used to adjust the visible light image of the composite insulator to a size that matches the image segmentation model before inputting the visible light image into the image segmentation model; normalize the pixel values ​​of the resized visible light image; and convert the normalized image to the HSV color space to obtain the preprocessed visible light image.

[0064] Understandably, preprocessing the visible light image before inputting it into the image segmentation model can uniformly adjust the image to an input format suitable for the model. Size adjustment ensures the image meets the model's input requirements, and pixel value normalization improves the model's numerical stability and convergence. Furthermore, converting the normalized image to the HSV color space highlights the color and brightness differences between water droplets and the background, reducing the impact of lighting, environment, and other interference factors on the segmentation results. The preprocessed image facilitates accurate identification and differentiation of water droplet pixels from background pixels by the image segmentation model, effectively improving the generation accuracy of the binary mask image. This provides a reliable data foundation for the accurate extraction of subsequent hydrophobicity-related indicators, thereby enhancing the stability and accuracy of the entire aging risk assessment method.

[0065] As a preferred embodiment, it also includes: a morphological processing module; the morphological processing module is used to perform opening operations on the binary mask before identifying independent connected regions in the binary mask and extracting the proportion of water-attached area, the number of water droplets, and the average roundness index based on the connected regions, to obtain a denoised image after removing small noise points and smoothing the edges of water droplets; and to perform closing operations on the denoised image to fill holes in the denoised image with an area smaller than a preset area, to obtain a morphologically processed binary mask.

[0066] In this embodiment, before identifying independent connected regions in the binary mask image, a preprocessing step is performed on the binary mask image. Specifically, the first step is to perform an erosion-dilation process (i.e., opening operation) on the binary mask image. First, a structuring element of appropriate size is selected, and this structuring element is moved pixel by pixel on the binary mask image. At each position, only when the area completely covered by the structuring element consists entirely of water droplet pixels is the pixel at that position set to the corresponding color of the water droplet in the binary mask image. As long as there is even one point in the area covered by the structuring element that is not a water droplet, the process continues. For each water droplet pixel, the pixel at that location is changed to the corresponding color of the background in the binary mask. This removes small, scattered noise points in the binary mask that are not real water droplets. After erosion, a dilation operation is performed. Specifically, the structuring element is used to move pixel by pixel on the image where noise points have been removed. As long as there is even a tiny water droplet pixel in the area covered by the structuring element, the pixel at that location is changed to the color corresponding to the water droplet. This fills in the small gaps and burrs that appear at the water droplet boundaries caused by the erosion stage, making the edges of the water droplets smoother and more even. The second step involves performing a dilation-erosion operation (i.e., closing) on ​​the binary mask image. Specifically, using the same structuring element, the image is moved pixel by pixel. Whenever a droplet pixel is within the area covered by the structuring element, that location is changed to the color corresponding to the droplet. This fills in the tiny, invisible black holes within the droplet area with the droplet's color, connecting the previously scattered and discontinuous droplet areas into a continuous white droplet region. Finally, an erosion operation is performed, again using the same structuring element, moving pixel by pixel. Only when the area covered by the structuring element is entirely composed of droplet pixels is the droplet's color retained; otherwise, it's changed to the background color. This step is to smooth out the excess parts extending from the dilated droplet boundaries, restoring the droplet's shape to its original form, preventing distortion from the previous dilation. In essence, this pre-processing of the binary mask image effectively removes tiny noise points, smooths the edges of the droplet region, and fills in the tiny holes within the droplets, transforming scattered areas into continuous, complete droplet regions. This processing step can eliminate interference and optimize image quality, providing a clear and reliable image basis for subsequent accurate identification of independent connected regions, extraction of water-attached area ratio, water droplet quantity and average roundness, thereby improving the accuracy and stability of hydrophobicity level assessment.

[0067] As a preferred embodiment, the damage index calculation module specifically includes: a band reflectance acquisition unit, a normalized difference calculation unit, a relative brightness change calculation unit, and a damage index calculation unit. A band reflectance acquisition unit is used to identify the reflectance of the composite insulator skirt surface in a preset first band and a preset second band based on a multispectral image; wherein the wavelength of the first band is shorter than the wavelength of the second band. The normalized difference calculation unit is used to determine the normalized difference of the reflectance of the umbrella skirt surface in the two bands based on the reflectance of the first band and the reflectance of the second band. The relative brightness change calculation unit is used to determine the relative brightness change of the surface color of the composite insulator skirt based on the current brightness value and the insulator reference brightness value. The damage index calculation unit is used to calculate the damage index of the composite insulator skirt surface based on the normalized difference and the relative degree of color change of the composite insulator skirt surface.

[0068] Specifically, ash generated in wildfire environments can chemically react with the silicone rubber sheds or core rods of insulators, leading to structural damage and performance degradation. This damage is usually gradual, and long-term accumulation can cause insulation failure and decreased mechanical strength. Therefore, it is necessary to quantify the surface damage based on multispectral diagnostic indices. For example, the formula for calculating the damage index of composite insulator shed surfaces is as follows: (5) In the formula, This indicates the reflectivity of the composite insulator skirt surface at a preset first wavelength of 450nm. This indicates the reflectivity of the composite insulator skirt surface in the preset second wavelength band of 850nm. Indicates the current brightness value. This indicates the reference brightness value for composite insulators, which is typically greater than or equal to 85. Characterizing the relative degree of color change on the surface of the composite insulator skirt. This refers to the normalized difference in reflectance of the umbrella skirt surface in two bands. The 0.7 and 0.3 in the formula are two weighting coefficients used to allocate the contribution ratio of the two indicators in the final damage index FPI.

[0069] It is understood that this embodiment obtains the reflectivity of the umbrella skirt surface in different bands through multispectral images and calculates the damage index (FPI) by combining the brightness value. This can transform the complex state of the umbrella skirt surface, such as aging, pollution, and discoloration, into a unified and quantifiable numerical index, providing a reliable quantitative basis for subsequent aging risk level determination and improving the accuracy and repeatability of the assessment results.

[0070] As a preferred embodiment, the fractal dimension calculation module 206 specifically includes: a grid coverage unit, a grid quantity statistics unit, a change calculation unit, and a fractal dimension determination unit.

[0071] A grid coverage unit is used to cover the image of the area where the crack is located using a grid of a first preset size and a grid of a second preset size, respectively. The grid quantity counting unit is used to count the number of first grids required when covering the image of the crack location area with a grid of the first preset size, and the number of second grids required when covering the image of the crack location area with a grid of the second preset size. The change calculation unit is used to determine, based on the first grid number and the second grid number, the first change in the number of grids required to cover the crack when the grid size changes from the first preset size to the second preset size; and to calculate the second change in the grid size on a logarithmic scale based on the first preset size and the second preset size. The fractal dimension determination unit is used to determine the fractal dimension of the image of the region where the crack is located based on the first change and the second change.

[0072] For example, after obtaining the binarized image of the crack, the formula for calculating its fractal dimension using the box counting method is as follows: (6) In the formula, The fractal dimension of a crack is used to quantify its complexity, spatial distribution density, or propagation state. A higher fractal dimension generally indicates a more complex and denser crack network. For use of size The number of first grid cells required when the grid covers the image of the area where the crack is located. For use of size The number of second grids required when the grid covers the image of the area where the crack is located. Indicates the first preset size. Indicates the second preset size, and , The quantitative relationship between them usually satisfies: , The first change in the number of meshes required to cover the crack is represented when the mesh size changes from a first preset size to a second preset size. This represents the second change in grid size on a logarithmic scale. Understandably, calculating the fractal dimension of a crack region image using box counting transforms the complex morphology of the crack into a quantifiable numerical indicator. A higher fractal dimension indicates a more complex, denser, and irregularly propagating crack network; conversely, a lower fractal dimension indicates a relatively simpler crack morphology. This quantitative indicator provides an objective and stable basis for subsequent assessments of the aging degree and damage level of composite insulator skirts, avoiding the subjectivity of manual crack observation and improving the accuracy and repeatability of the entire aging risk assessment scheme.

[0073] As a preferred option, the aging level determination module 208 specifically includes: an aging index calculation unit and an aging level determination unit.

[0074] The aging index calculation unit is used to calculate the aging index based on the aging index calculation formula, according to the hydrophobicity level, curvature integral value, fractal dimension and damage index. For example, the formula for calculating the aging index is: (7) In the formula, The aging index, Hydrophobicity level The weighting coefficients corresponding to the hydrophobicity level are: The value is the integral of curvature. These are the weighting coefficients corresponding to the integral value of curvature. For fractal dimension, These are the weighting coefficients corresponding to the fractal dimension. The damage index, These are the weighting coefficients corresponding to the damage index. This represents the maximum value of the curvature integral achieved under extreme melting deformation of the umbrella skirt. This represents the maximum value of the fractal dimension achieved under severe crack conditions.

[0075] Understandably, this embodiment uses the aging index calculation formula to weight and fuse four key indicators—hydrophobicity level, curvature integral value, fractal dimension, and damage index—according to their respective weight coefficients, resulting in a unified aging index (CAI). The aging level is then determined based on this index. This process transforms the originally scattered multi-dimensional detection results into a quantifiable and comparable comprehensive value, objectively and comprehensively reflecting the overall aging status of composite insulators in terms of hydrophobicity, edge melting deformation, crack complexity, and surface damage. This avoids the one-sidedness of single-indicator evaluation, improves the accuracy and scientific nature of aging risk assessment, and provides a stable and reliable quantitative basis for subsequent operation and maintenance decisions.

[0076] The aging level determination unit is used to determine the aging level of composite insulators based on the aging index.

[0077] For example, the aging level of a composite insulator can be determined based on an aging index using a pre-defined table, as shown in the table below: Table 4. Example Table of Aging Grade Classification Standards for Composite Insulators As shown in the table above, when 0 ≤ CAI ≤ a, the CAI level is determined as follows: When a < CAI ≤ b, determine the CAI level as follows: When CAI > b, determine the CAI level as follows: For example, the value of a can be 0.3 and the value of b can be 0.6. The values ​​of a and b can be set according to the actual situation. For example, in some more stringent scenarios, the value of b can be set to 0.55. This embodiment does not make specific limitations on this.

[0078] It is understood that this embodiment directly maps continuous aging index values ​​to intuitive aging levels by using a preset aging index CAI threshold range, allowing maintenance personnel to quickly grasp the aging status of insulators without having to interpret complex index values.

[0079] As a preferred embodiment, the system also includes a maintenance measure determination module. This module, after determining the aging level of the composite insulator, can implement corresponding maintenance measures based on that aging level. Please refer to the table below for details: Table 5. Examples of Aging Level Determination and Maintenance Strategies for Composite Insulators As shown in the table above, when 0 ≤ CAI ≤ a, the corresponding CAI level is... The corresponding maintenance measure is measure c. When a < CAI ≤ b, the corresponding CAI level is... The corresponding maintenance measure is measure d. When CAI > b, the corresponding CAI level is... The corresponding maintenance measure is measure e. For example, measure c could be to perform annual inspections, that is, to inspect the composite insulators on an annual cycle, or to inspect the composite insulators on a 1.5-year cycle. Measure d could be to perform quarterly monitoring and surface repair, that is, to monitor the composite insulators on a quarterly cycle and repair the surface of the sheds. Measure e could be to immediately replace the composite insulators. In specific operations, measures c, d, and e can correspond to different measures, and operators can set them freely. For example, in some more stringent scenarios, measure c can be set to inspect the composite insulators every six months, or measure d can be set to immediately replace them. This embodiment does not make specific limitations on this.

[0080] Understandably, after determining the aging level of the composite insulator, this embodiment maps different aging levels to specific and actionable maintenance measures, directly transforming the results of the aging risk assessment into targeted operation and maintenance action plans. This allows the entire aging risk assessment plan to be transformed from technical analysis into engineering practice, significantly improving the efficiency and scientific nature of operation and maintenance decisions and ensuring the stable operation of the power system.

[0081] As a preferred embodiment, the curvature integral value acquisition module 205 includes a curvature integral value calculation unit, which is used to calculate the curvature integral value according to the formula.

[0082] For example, firstly, the extracted discrete contour point set scattered along the edge of the umbrella skirt undergoes smoothing and interpolation processing. Smoothing eliminates minor fluctuations and jagged edges between these discrete points, making the point set more regular. Interpolation adds appropriate points between adjacent discrete points, making the originally scattered point set more closely connected. Through these two processes, the originally discrete contour point set is fitted into a continuous and smooth spatial plane curve without breaks. This curve is the complete edge contour curve of the umbrella skirt. After processing, based on the principles of differential geometry, the curvature value is calculated for each point on this contour curve, obtaining the curvature data corresponding to all points on the curve. Then, along the entire contour curve of the umbrella skirt, the principal curvature of each point is found, the absolute value of each principal curvature is taken, and the absolute values ​​of these principal curvatures are integrated. The final result is the curvature integral, and the formula for calculating the curvature integral is as follows: (8) In the formula, It is the curvature integral, used to quantify the degree of melting deformation of the composite insulator skirt. It is the total length of the outline curve of the composite insulator skirt, in millimeters (mm). , These are the two principal curvatures at a point on the profile curve of the composite insulator skirt, expressed in millimeters, reflecting the degree of curvature at that point. It is understood that this embodiment, by smoothing and interpolating the discrete profile point set, can eliminate burrs and fluctuations at the profile points, fitting scattered points into a continuous, complete, and smooth skirt edge curve, ensuring the accuracy of subsequent calculations. Furthermore, by calculating the curvature and principal curvatures at each point on the profile, and integrating the absolute values ​​of the principal curvatures to obtain the curvature integral, the bending changes at the skirt edge can be converted into quantitative values. This objectively and accurately measures the irregularity and melting deformation of the skirt edge, providing a stable and reliable quantitative basis for subsequent judgments of skirt aging and damage, improving the scientific rigor and accuracy of the evaluation results.

[0083] It is understood that the above-described device embodiments correspond to the method embodiments of the present invention, and can implement the composite insulator skirt aging risk assessment method provided by any of the above-described method embodiments of the present invention.

[0084] It should be noted that the device embodiments described above are merely illustrative, and some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the device embodiments provided by this invention, the connection relationships between modules indicate that they have communication connections, which can specifically be implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0085] Based on the above embodiments of the composite insulator skirt aging risk assessment method, another embodiment of the present invention provides a terminal device, which includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the composite insulator skirt aging risk assessment method of any embodiment of the present invention.

[0086] For example, in this embodiment, the computer program can be divided into one or more modules, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the terminal device.

[0087] The terminal device may be a desktop computer, laptop, handheld computer, or cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0088] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the terminal device, connecting all parts of the terminal device via various interfaces and lines.

[0089] Based on the above-described method embodiments, another embodiment of the present invention provides a computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to execute the composite insulator skirt aging risk assessment method described in any of the above-described method embodiments of the present invention.

[0090] The modules / units integrated in the device / terminal equipment, if implemented as software functional units and sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0091] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method of assessing the aging risk of a composite insulator shed, characterized by, include: Acquire visible light and multispectral images of the composite insulator to be evaluated; The visible light image is input into an image segmentation model to generate a binary mask image representing the pixel type of each pixel in the visible light image; wherein, the pixel type includes water droplets and background; Identify independent connected regions in a binary mask image, and extract the percentage of water-covered area, the number of water droplets, and the average roundness index based on the connected regions; evaluate the hydrophobicity level based on the percentage of water-covered area, the number of water droplets, and the average roundness index. The edge contour and crack location of the composite insulator skirt are segmented and extracted from the visible light image; A curve is fitted based on the set of points of the edge contour; the absolute value of the principal curvature of the points on the curve is integrated to obtain the curvature integral value; The fractal dimension of the image of the region where the crack is located is calculated using the box counting method; The reflectance of the composite insulator skirt surface in multiple bands is identified based on multispectral images, and the current brightness value is calculated based on the reflectance. The damage index of the composite insulator skirt surface is calculated based on each reflectance, the current brightness value, and the insulator reference brightness value. The aging level of the composite insulator is determined based on the hydrophobicity level, the curvature integral value, the fractal dimension, and the damage index.

2. The composite insulator shed aging risk assessment method of claim 1, wherein, Before inputting the visible light image into the image segmentation model, the method further includes: Adjust the visible light image of the composite insulator to a size that matches the image segmentation model; The pixel values ​​of the resized visible light image are normalized. The normalized image is converted to the HSV color space to obtain the preprocessed visible light image.

3. The method for assessing the aging risk of composite insulator skirts as described in claim 2, characterized in that, The reflectance of the composite insulator skirt surface in multiple bands is identified based on multispectral images, and the current brightness value is calculated based on the reflectance. Based on the reflectivity, current brightness value, and insulator reference brightness value, the damage index of the composite insulator skirt surface is calculated, including: Based on the reflectivity of the composite insulator skirt surface in a preset first band and the reflectivity in a preset second band, the normalized difference of the reflectivity of the skirt surface in the two bands is determined; wherein, the wavelength of the first band is less than the wavelength of the second band. The relative degree of color change on the surface of the composite insulator skirt is determined based on the current brightness value and the insulator reference brightness value; The damage index of the composite insulator skirt surface is calculated based on the normalized difference of reflectivity and the relative change in color of the composite insulator skirt surface.

4. The method for assessing the aging risk of composite insulator skirts as described in claim 3, characterized in that, Before identifying independent connected regions in a binary mask image and extracting the percentage of water-covered area, the number of water droplets, and the average roundness index based on these connected regions, the process also includes: An opening operation is performed on the binary mask image to obtain a denoised image after removing tiny noise points and smoothing the edges of water droplets. The denoised image is subjected to a closing operation to fill holes in the denoised image with an area smaller than a preset area, resulting in a morphologically processed binary mask image.

5. The method for assessing the aging risk of composite insulator skirts as described in claim 4, characterized in that, The fractal dimension of the image of the region where the crack is located is calculated using the box counting method, including: The image of the crack location area is covered by grids of the first preset size and the second preset size, respectively; The number of first grids required when covering the image of the crack location area with a grid of the first preset size, and the number of second grids required when covering the image of the crack location area with a grid of the second preset size; Based on the first grid number and the second grid number, determine the first change in the number of grids required to cover the crack when the grid size changes from the first preset size to the second preset size; Based on the first preset size and the second preset size, calculate the second change in the grid size at the logarithmic scale; The fractal dimension of the image of the region where the crack is located is determined based on the first change and the second change.

6. The method for assessing the aging risk of composite insulator skirts as described in claim 5, characterized in that, The aging level of the composite insulator is determined based on the hydrophobicity level, the curvature integral value, the fractal dimension, and the damage index, including: Based on the aging index calculation formula, the aging index is calculated according to the hydrophobicity level, curvature integral value, fractal dimension and damage index. The formula for calculating the aging index is as follows: In the formula, The aging index, Hydrophobicity level The weighting coefficients corresponding to the hydrophobicity level are: The value is the integral of curvature. These are the weighting coefficients corresponding to the integral values ​​of curvature. For fractal dimension, These are the weighting coefficients corresponding to the fractal dimension. The damage index, These are the weighting coefficients corresponding to the damage index. This represents the maximum value of the curvature integral achieved under extreme melting deformation of the composite insulator skirt. This represents the maximum value of the fractal dimension achieved under severe crack conditions. The aging level of composite insulators is determined based on the aging index.

7. A composite insulator skirt aging risk assessment device, characterized in that, include: The system includes an image acquisition module, a binary mask generation module, a hydrophobicity level acquisition module, an edge contour and crack location extraction module, a curvature integral value acquisition module, a fractal dimension calculation module, a damage index calculation module, and an aging level determination module. The image acquisition module is used to acquire visible light and multispectral images of the composite insulator to be evaluated. The binary mask generation module is used to input a visible light image into an image segmentation model and generate a binary mask that represents the pixel type of each pixel in the visible light image; the pixel types include water droplets and background. The hydrophobicity level acquisition module is used to identify independent connected regions in a binary mask image, extract the proportion of water-attached area, the number of water droplets, and the average roundness index based on the connected regions, and evaluate the hydrophobicity level based on the proportion of water-attached area, the number of water droplets, and the average roundness index. The edge contour and crack location extraction module is used to segment and extract the edge contour and crack location of the composite insulator skirt from the visible light image; The curvature integral value acquisition module is used to fit a curve based on the point set of the edge contour; and to integrate the absolute value of the principal curvature of the points on the curve to obtain the curvature integral value. The fractal dimension calculation module is used to calculate the fractal dimension of the image of the region where the crack is located using the box counting method. The damage index calculation module is used to identify the reflectance and current brightness value of the composite insulator skirt surface in multiple bands based on multispectral images; and to calculate the damage index of the composite insulator skirt surface based on each reflectance, current brightness value and insulator reference brightness value. The aging level determination module is used to determine the aging level of composite insulators based on hydrophobicity level, curvature integral value, fractal dimension, and damage index.

8. The composite insulator skirt aging risk assessment device as described in claim 7, characterized in that, Also includes: Image preprocessing module; The image preprocessing module is used to adjust the visible light image of the composite insulator to a size that matches the image segmentation model before inputting the visible light image into the image segmentation model; The pixel values ​​of the resized visible light image are normalized. The normalized image is converted to the HSV color space to obtain the preprocessed visible light image.

9. A terminal device, characterized in that, The method includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor, when executing the computer program, implements the composite insulator skirt aging risk assessment method as described in any one of claims 1-6.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program; wherein, when the computer program is executed, it controls the device on which the computer-readable storage medium is located to perform the composite insulator skirt aging risk assessment method as described in any one of claims 1-6.

Images