Embedded power transmission line airborne auxiliary inspection tour device based on FPGA

Abstract

The invention discloses an embedded power transmission line airborne auxiliary patrol device based on an FPGA, and belongs to the technical field of line maintenance, and the device comprises the steps: designing a method for weighting a shared parameter and K-Means++ clustering and carrying out compression optimization on the neural network structure model; putting the compressed network model into a high-performance embedded FPGA system with a parallel structure; installing the integrated FPGA system on the unmanned aerial vehicle; collecting an original image in an unmanned flight state; preprocessing the original image to obtain a target image; performing quality evaluation on the target image, and determining whether reshooting is needed or not; and carrying out real-time defect detection on the target image, and transmitting a detection result to a ground background system. The technical scheme of the embodiment of the invention is used for real-time inspection of the unmanned aerial vehicle for the power transmission line, the real-time performance is high, the result is accurate, and compared with the currently adopted manual inspection, the cost is greatly reduced.

Description

technical field [0001] The invention belongs to the technical field of line maintenance, and in particular relates to the maintenance of power transmission lines, and in particular is based on image processing. Background technique [0002] At present, my country has built six power grids across provinces and regions, namely the South, Northwest, East China, Central China, North China, and Northeast China. The total length of transmission lines exceeds 1.15 million kilometers, and transmission lines of 500kV and above have become the main power grids in various regions. Power grid transmission main force. Our country has a vast territory and relatively complex terrain, with many hills and few plains. In addition, the complex and changeable meteorological conditions bring certain difficulties to the construction of cross-regional power grids and ultra-high voltage transmission lines. For maintenance and maintenance, only relying on existing inspection methods and routine test...

AI Technical Summary

Problems solved by technology

[0003] However, due to limitations in technology and inspection mode, inspections by helicopters and UAVs are less efficient in terms of data collection, data validity, and timeliness of analysis and processing. There are the following problems:

[0004] (1) Due to the large space of the helicopter, in order to avoid dangerous contact with the transmission line, a safe inspection distance of 20-30m is generally guaranteed, which is prone to inspection blind spots, resulting in inaccurate inspection results and limited coverage;

[0005] (2) Helicopter patrol inspection mainly adopts two methods of artificial telescope screening and infrared detection and observation, and the detection mode and identification type have certain limitations, which may easily cause missed inspections, resulting in low inspection efficiency;

[0006] (3) Due to the lack of online inspection target positioning and real-time feedback mechanism for helicopter inspection results, there is no maintenance information tracking for the defects detected in a single inspection, so some defects cannot be determined whether they are new or existed before. Repeated screening of defects affects the efficiency of inspection to a certain extent;

[0007] (4) Most of the data analysis collected by UAV inspections is processed offline in the background, and the real-time performance is low, which affects the accuracy of fault diagnosis to a certain extent, and easily causes problems such as fault misjudgment and missed judgment;

[0008] (5) When the image cannot be recognized due to out-of-focus, shaking, overexposure, etc. during the drone inspection process, there is a lack of information supplementary acquisition methods;

Images