Design of the Transmission Line Inspection System Based on UAV

In the power line, the fracture of the aluminum stranded wire and the surface damage will gradually develop into an irreversible damage to the conductor and the entire transmission line, and even cause serious power failure. In order to find potential strand breaks and damages and prevent them from further deterioration, this paper proposes a method based on UAV inspection to identify power transmission line breaks and surface defects. First, the transmission line conductor image is obtained through the UAV image acquisition system, the image is enhanced by the 3/4 denoising clustering method, and the Canny edge detection method is used to extract the conductor area contour. Then, an image segmentation recognition method based on clustering and DS evidence fusion theory is proposed, which can simply and effectively identify the state of wire breakage, and establish a library of recognition models under different states of transmission lines. Finally, the transmission line image comparison method based on the AdaBoost algorithm is used for rapid and accurate identification, thereby improving the work efficiency and inspection quality of the transmission line maintenance and operation inspection. The performance of the method is analyzed through a series of experiments, and the results show that the average accuracy of the method can be measured for conductor fracture and surface defect failures of 90.45% and 92.05%, respectively. According to the method proposed in this article, the health of the conductor can be monitored, so as to ensure the safe operation of the power transmission line.

I. Introduction

The reliability, stability and power supply quality of electric energy play a vital role in the country's economic development. my country's transmission lines cover a wide range and the terrain is complex, which poses greater challenges for its normal operation, inspection and maintenance. The traditional manual inspection method has low efficiency, long cycle, high requirements, and low safety, and cannot meet the current transmission line maintenance needs under current national conditions. The drone patrol detection method with drones as the core can observe the transmission line at high altitude through the carried camera equipment and return images for detection. It has the advantages of low cost, easy operation, high safety and flexibility, so it has become the focus of research on transmission line detection in recent years. It is worth noting that most of the domestic inspection methods for UAV inspections still remain at the stage of manual inspection and shooting of videos and images, and there is very little research on computer automatic image recognition, and there is still a big gap compared with foreign research. Referenced [1] P. Nichols and T. J. Summers use the parameters of the ground return transmission line to align the length of the overhead conductors in the static range to evaluate the length, which provides a basis for power inspection. Referenced [2] D. H. Waters, J. Hoffman, E. Hakansson and M. KumosaInt studied the low-speed impact effect of transmission line conductors, and analyzed the low-speed impact of electrical energy that caused the failure of transmission conductors. Referenced [3] C. Rakpenthai and S. Uatrongjit estimate the state of the power system and the temperature of transmission line conductors, and provide important decisions for power inspections. Referenced [4] AAM Farah, MM Afonso, JA Vasconcelos and MAO Schroeder proposed a finite element method for the surface electric field calculation of overhead transmission line conductors. This method can accurately calculate the transmission line parameters and provide a reference for power inspection. Referenced [5] Y. Wu, Y Sui and G. Wang proposed a vision-based UAV sensing system for real-time air object positioning and tracking methods, which provide important ideas for UAV transmission line inspection. Referenced [6] Y. Zhang, X. Yuan, W. Li, and S. Chen first proposed the idea of using UAV images for automatic power line inspection. Referenced [7] V. N. Nguyen, R. Jenssen and D. Roversodu summarized the power line inspection method based on deep learning. Referenced [8] B. Li and C. Chen proposed a layered detection algorithm for transmission lines based on aerial images, which improved the accuracy of transmission line inspections. Referenced [9] Z. Li, Y. Liu, R. Walker, R. Hayward, and J. Zhang proposed a UAV transmission line detection method based on pulse coupled neural filter and improved Hough transform, which provides for transmission line inspection theoretical basis.