Галерея 2846293
⚡ ПОДРОБНЕЕ ЖМИТЕ ЗДЕСЬ 👈🏻👈🏻👈🏻
Галерея 2846293
All Books Conferences Courses Journals & Magazines Standards Authors Citations
Insulator Bunch-Drop Detection and Location
Experimental results of insulator fault detection based on spatial morphological features of aerial images: (a) Original image; (b) Tilt correction image; (c) Located ins... View more
Abstract: Because insulators provide electrical insulation and mechanical support for electric transmission lines, these components are of paramount importance to safe and reliable... View more
Because insulators provide electrical insulation and mechanical support for electric transmission lines, these components are of paramount importance to safe and reliable operations of power systems. However, insulators are often considered to be prone to different faults, e.g., bunch-drop, which demands a novel solution for accurate fault detection and fault location. Current research efforts have primarily focused on the bunch-drop fault of glass insulators, and the study of ceramic insulators has not been reported to date. To this end, this paper proposes an algorithmic solution for the bunch-drop fault detection for both glass and ceramic insulators based on spatial morphological features, which can be integrated into an unmanned aerial vehicle-based inspection system. Color models can be established based on the unique color features of both glass and ceramic insulators. Next, the target areas of the insulators can be identified according to the color determination combined with the insulator's spatial features. The target area is morphologically processed to highlight the fault location, and the rules are established based on the spatial feature differences between the insulators with and without faults. Consequently, the fault location can be accurately identified, and the coordinates can be determined. The performance of the proposed solution is evaluated in comparison with existing solutions. The numerical results demonstrate that the proposed solution can detect the bunch-drop faults of insulators with a better than average detection rate. In addition, the performance is assessed and validated in terms of robustness and real-time performance.
Published in: IEEE Access ( Volume: 6 )
Experimental results of insulator fault detection based on spatial morphological features of aerial images: (a) Original image; (b) Tilt correction image; (c) Located ins... View more
TABLE 1
Comparison of the Time Consumption and Detection Rate of the Algorithms
S. P. Zhang, Z. Yang, X. N. Huang and Y. Z. Wu, "Defects detection and positioning for glass insulator from aerial images", J. Terahertz Sci. Electron. Inf. Technol. , vol. 11, pp. 609-613, Aug. 2013.
T. Hirakawa et al., "Tree-wise discriminative subtree selection for texture image labeling", IEEE Access , vol. 5, pp. 13617-13634, Jul. 2017.
L. Li, L. Sun, W. Kang, J. Guo, C. Han and S. Li, "Fuzzy multilevel image thresholding based on modified discrete grey wolf optimizer and local information aggregation", IEEE Access , vol. 4, pp. 6438-6450, Sep. 2016.
Z. Guan, X. Wang, X. Bian, L. Wang and Z. Jia, "Analysis of causes of outdoor insulators damages on HV and UHV transmission lines in China", Proc. IEEE Elect. Insulation Conf. (EIC) , pp. 227-230, Jun. 2014.
Y. L. Wang and B. Yan, "Vision based detection and location for cracked insulator", Comput. Eng. Des. , vol. 35, no. 2, pp. 583-587, Feb. 2014.
D. Zuo, H. Hu, R. Qian and Z. Liu, "An insulator defect detection algorithm based on computer vision", Proc. IEEE Int. Conf. Inf. Automat. (ICIA) , pp. 361-365, Jul. 2017.
J. Han, J. J. Zhang and B. H. Wang, "Method on recognizing the structure of transmission line based on perceptual organization", Infrared Laser Eng. , vol. 42, no. 12, pp. 3458-3463, Dec. 2013.
W. Wang, Y. Wang, J. Han and Y. Liu, "Recognition and drop-off detection of insulator based on aerial image", Proc. 9th Int. Symp. Comput. Intell. Design (ISCID) , pp. 162-167, 2016.
Y. T. Jiang et al., "The identification and diagnosis of self-blast defects of glass insulators based on multi-feature fusion", Electr. Power , vol. 50, no. 5, pp. 52-58, May 2017.
J. P. Shang, C. X. Li and L. Chen, "Location and detection for self-explode insulator based on vision", J. Electron. Meas. Instrum. , vol. 31, no. 6, pp. 844-849, Jun. 2017.
F. Y. Zhang, "Recognition and research of anomaly map of transmission line inspection based on computer vision", 2015.
Y. Zhai, D. Wang, M. Zhang, J. Wang and F. Guo, "Fault detection of insulator based on saliency and adaptive morphology", Multimedia Tools Appl. , vol. 76, no. 9, pp. 12051-12064, May 2017.
L. Shi, "The method of image detection on defect insulator in transmission line", 2013.
F. Gao et al., "Recognition of insulator explosion based on deep learning", Proc. 14th Int. Comput. Conf. Wavelet Act. Media Technol. Inf. Process. (ICCWAMTIP) , pp. 79-82, 2017.
Y. Liu, J. Yong, L. Liu, J. Zhao and Z. Li, "The method of insulator recognition based on deep learning", Proc. 4th Int. Conf. Appl. Robot. Power Ind. (CARPI) , pp. 1-5, 2016.
Z. Zhao, G. Xu, Y. Qi, N. Liu and T. Zhang, "Multi-patch deep features for power line insulator status classification from aerial images", Proc. Int. Joint Conf. Neural Netw. (IJCNN) , pp. 3187-3194, 2016.
F. Gao, J. Wang, Z. Kong, J. Wu and N. Feng, "Recognition of insulator explosion based on deep learning", Proc. 14th Int. Comput. Conf. Wavelet Act. Media Technol. Inf. Process. (ICCWAMTIP) , pp. 79-82, 2017.
Y. J. Zhai, D. Wang, Z. B. Zhao and H. Y. Cheng, "Insulator string location method based on spatial configuration consistency feature", Chin. Soc. Elect. Eng. , vol. 37, no. 5, pp. 1568-1577, Mar. 2017.
M. J. Swain and D. H. Ballard, "Color indexing", Int. J. Comput. Vis. , vol. 7, no. 1, pp. 11-32, 1991.
T. Fang, C. Dong, X.-L. Hu and Y. Wang, "Contour extraction and fault detection of insulator strings in aerial images", J. Shanghai Jiaotong Univ. , vol. 47, no. 12, pp. 1818-1822, Dec. 2013.
C. Yao, L.-J. Jin and S.-J. Yan, "Recognition of insulator string in power grid patrol images", J. Syst. Simul. , vol. 24, no. 9, pp. 1818-1822, Sep. 2012.
K. B. Cui, "Research on the key technologies in insulator defect detection based on image", 2016.
X. Y. Zhang, J. B. An and F. M. Chen, "A method of insulator fault detection from airborne images", Proc. 2nd WRI Global Congr. Intell. Syst. , pp. 200-203, 2010.
R. C. Gonzalez, R. E. Woods and S. L. Eddins, Digital Image Processing Using MATLAB, Beijing, China:PHEI, 2013.
IEEE Account
Change Username/Password
Update Address
Purchase Details
Payment Options
Order History
View Purchased Documents
Need Help?
US & Canada: +1 800 678 4333
Worldwide: +1 732 981 0060
Contact & Support
The insulator is one of the most important components for electrical insulation and mechanical support in electric power transmission lines. Insulators are subjected to large mechanical tension and extremely high voltage with long time exposure outdoors. The defects or faults of insulators can directly lead to significant power loss and can even result in large-scale power outages or blackouts. The insulators that are adopted in transmission lines are primarily made of ceramic and glass. The ceramic insulators are polycrystalline heterogeneous materials, and cracks can occur due to mechanical, electrical and external forces. In the case of strikes due to lightening, an electric arc forms a drainage channel in the head ceramic parts, and the ceramic can burst, which causes a bunch-drop accident. Glass insulators are considered to have uniform texture and compact structure. Although the tensile strength is improved by toughening, a bunch-drop can occur due to overloading. In general, such an insulator fault can be identified visually. However, an assessment method by human visual inspection is inefficient and even not feasible in practice for high-voltage transmission lines that span a large geographical area because of the high error rate and long assessment time. In recent years, the development direction of line inspection has trended toward the use of unmanned aerial vehicle (UVA) inspection. Through the processing and analysis of the aerial images captured by the UAV, the faults in the insulators can be efficiently detected and located [1] – [4] .
In the literature, numerous studies have been conducted to address the challenge of insulator fault detection, and a collection of solutions is available. Wang and Yan [5] performed rough segmentation of insulators in a laboratory space to obtain two value images to represent the location and range of the insulators. A mathematical model was developed to calculate the ratio of the effective pixels in the insulator area for bunch-drop fault detection. In [6] – [8] , the identified insulator region was divided into blocks, and insulator bunch-drop fault diagnosis could be conducted using a texture feature quantity. This method performs well in bunch-drop fault diagnosis for the overlapping insulator pieces, but the performance degraded in the cases in which insulator pieces were separated from each other, which could lead to a false diagnosis result. In [9] , the insulator region was divided into a flake area based on the insulator string, and the center of gravity of each insulator string was analyzed. The center of gravity distance between the adjacent insulator pieces could be calculated, and hence, the fault location could be determined based on the difference in the center of gravity distance. However, the performance of this solution can be degraded by the background texture and light, and the system might not be able to obtain the complete insulator area. The studies in [10] and [11] adopted the maximum between-cluster variance and the Adaboost classifier to locate the insulator position and then calculate the relative distance of the adjacent insulator piece for bunch-drop fault detection using the insulator contour. This method is merely applicable to the independent and unobstructed situation among adjacent insulators in aerial images. In [12] , the proposed solution adopted the color determination method to segment the insulators and locate the defective parts through an adaptive morphology. However, such an approach cannot be applied to ceramic insulators because the ceramic insulators can hardly be identified from the background. In [13] and [14] , the sliding window algorithm was developed to match the gray histogram of the template and the captured insulators. The insulator defects can be identified and located based on the distance of the histogram. The performance can be significantly affected by the detection environment and the selection of templates. The studies in [15] and [16] adopted advanced deep learning algorithms (CNN and Faster-RCNN) to identify the location of insulators. However, a large number of image samples and computational time are generally required during the training process, and the on-board GPU is needed to carry out the task. The detection of insulator bunch-drop fault has not been explicitly studied in these studies. Gao et al. [17] adopted a deep learning based algorithm (VGG16) to identify insulators from the complex background using pixel reconstruction. However, the detection of insulator string fault was based on conventional method of finding the center of mass, and is limited to the glass insulators.
It can be observed that most existing studies have focused on glass insulators, and it is firmly assumed that the insulators are independent without obstacles. However, because the shooting distance and angle of aerial images vary over time during the UAV inspection, the insulator pieces in the aerial images can be connected and overlapping. It should be highlighted that the ceramic insulators dominate in electric power transmission lines in many cases, and few studies are available for fault detection of ceramic insulators.
Motivated by the existing solutions (e.g., [12] , [18] ), this paper proposes an algorithmic solution for insulator fault detection that is based on the spatial morphology features of obtained UAV aerial insulator images while considering the unique spatial and color features of glass insulators and ceramic insulators. The basic idea behind the proposed algorithmic solution is illustrated in Fig. 1 and summarized as follows: the color model for image segmentation is established based on the color features of glass and ceramic insulators. The captured insulators with different angles can be processed using the Hough Transform to detect the straight lines, and the insulators can be located. Finally, the located insulator is processed through morphological processing, and the spatial features of the fault location are analyzed; subsequently, the rules can be established. Based on these rules, the fault location coordinates in the insulators can be determined.
Flowchart of the proposed fault detection method. (a) Overall process of the algorithmic solution. Flowchart of the proposed fault detection method. (b) Detailed flowchart of the proposed algorithmic solution.
In summary, the main technical contributions made in this work can be summarized as follows: (1) the proposed solution can accurately identify and locate the bunch-drop faults based on the spatial features of UAV aerial images of both glass and ceramic insulators; and (2) the robustness and real-time performance of the proposed solution is evaluated and validated. The numerical result demonstrates that the proposed solution outperforms the existing solutions.
The remainder of this paper is organized as follows: Section II presents the insulator target detection process. Section III presents the detection and location method of the insulator bunch-drop fault in detail. Section IV reports on a range of experiments that were performed and presents the numerical results. Finally, the study’s conclusions are presented in Section V .
The color histogram was first used for image feature extraction in [19] and is considered to be an efficient method for describing the color features. The histogram can well reflect the composition and distribution of the image colors, i.e., the probability of the appearance of various colors. In [20] and [21] , the Lab color space or HIS color space was used to distinguish the glass insulators based on the empirical threshold. However, it should be noted that the segmentation of insulators from the complex background can be hardly achieved by the use of one threshold in practice. In addition, most of the solutions have merely focused on the glass insulators, and little investigation has been made of ceramic insulators. To accomplish this goal, this work exploits both types of insulators (i.e., glass and ceramic), and obtains the color distribution by sampling the RGB values of 100 glass insulators and ceramic insulators, as shown in Fig. 2 .
Histogram distribution of two material insulators in RGB color space. (a) Distribution histogram of glass insulators in RGB color space. (b) Distribution histogram of ceramic insulators in RGB color space.
The pseudo-code of color determination process for glass insulators is presented in Algorithm 1 .
R=image(:,:,1),G=image(:,:,2)
, B=image(:,:,3)
R \,\, = \,\, image(i,j,1),G \,\, = \,\, image(i,j,2), \,\, B=image(i,j,3)
bool_{1} =(78\le R\le 173),bool_{2} =(115\le B\le 175)
,
if bool_{1} ~and ~bool_{2} ~and~ bool_{3}
then
This color model can segment the insulator region from the complex background. The segmentation results of the glass and ceramic insulators are presented in Fig. 3 .
Results of color determination. (a) The segmentation results for the glass insulator. (b) The segmentation results for the ceramic insulator.
The insulator image can be segmented from the complex background based on color features, but when there is noise, it deteriorates the image quality for further analysis. In this instance, the median filter [22] is adopted to filter the lone noise while maintaining the edges of the images based on a nonlinear signal processing technique. Simultaneously, it also calculates the connected domain of the filtered image, removes the smaller part of the connected domain, and further strengthens the target area. The results of the process are presented in Fig. 4 .
Noise filtering process. (a) Original image. (b) Median filter. (c) Removal of the smaller part of the connected domain.
In reality, because the insulator images are often captured with different angles during inspection, the Hough transformation [23] is adopted to detect the straight lines, as well as to correct the images, thereby facilitating the follow-up operation, as shown in Fig. 5 . The target image can be identified using the line detection of the Hough transformation, and the location of the longest line segment is obtained. In light of the slope of the straight line, the angle of the insulator is determined, and then, the target image is rotated to perform a tilt correction.
Tilt correction. (a) Hough transformation detection of lines. (b) Tilt correction. (c) Correction of original image.
Fig. 6 illustrates the structure and components of the insulator. The insulator is composed of a vertical arrangement along the center axis and a number of insulator pieces that have the same shape and color, which indicates consistency in the form of space [18] .
To clearly observe the insulator features, the reverse operation of the insulator image is conducted through a tilt correction. The pseudo-code of spatial feature analysis is presented in Algorithm 2 .
X_{dis}, X_{max}, X_{min}, X_{maxx}, X_{minx}, X_{m}, X_{i}
, p_{x} =0
; q_{x} =0
; t_{r} =1
, t_{x} =1
if X_{dis} (1,t-4:t-1)\le X_{dis} (1,t)\le X_{dis} (1,t+4:t+1)
if X_{m} (1,t_{r})>X_{m} (1,t_{r} -1)+2
then
p_{x} =p_{x} +1
, X_{max} (1,p_{x})=X_{dis} (1,t)
, X_{maxx} (1,p_{x})=t
if X_{m} (1,t_{r})>X_{m} (1,t_{r} -1)+2
then
q_{x} =q_{x} +1
, X_{min} (1,q_{x})=X_{dis} (1,t)
, X_{minx} (1,q_{x})=\,\,t
figure: plot(1:n, Xdis), plot(X_{maxx}, X_{max})
,
Projection curves of insulators in the direction of X/Y axis. (a) Projection curves of normal insulator in the direction of X/Y axis. (b) Projection curves of fault glass insulator in the direction of X/Y axis. (c) Projection curves of fault ceramic insulator in the direction of X/Y axis.
It can be observed from the projection curves that the insulator projection curves exhibit obvious regular features.
The projection curves in the direction of the X axis are mostly equal amplitude oscillations. Each peak corresponds to an insulator piece, and the valley corresponds to the center of the steel cap. The distance between the peak and valley alternately appears which corresponds to the isometric arrangement of the insulator pieces. In addition, the projection curve in the direction of the Y axis is a single peak curve that has a certain width. In fact, the position of the wave peak corresponds to the center axis of the insulator.
Based on the spatial features, the insulator can be located using its projection curve through the following steps:
determination of the insulator position in the Y axis, which is similar to step #1.
determination of the insulator location in accordance with the position coordinates obtained by steps (1) and (2) .
Finally, the experimental result clearly demonstrates the accuracy of the insulator location, as presented in Fig. 8 .
The results of locating the insulator. (a) Original images. (b) Located images.
For insulators, the spacing between adjacent insulator pieces is equal. After a bunch-drop occurs, t
Обнажение Беаты
Нескромные индианки
Загорала у озера топлесс