Experiment and Research on Multi-Spectral Detection of Composite Insulators' Typical Defects under the Perspective of Unmanned Aerial Vehicle

Abstract

Currently, due to the lack of researches on how the reliability of ultraviolet(UV) detection based on Unmanned Aerial Vehicle(UAV) is influenced by distance, and the shortage of infrared(IR) and UV image pairs for comparative analysis, the development of multi-spectral intelligent detection technology for composite insulators is greatly restricted. In this paper, composite insulators with 3 kinds of typical defects, including mandrel rot, poor sheath-mandrel bonding and internal metal foreign body, were tested in laboratory respectively by hand-held IR and UV detectors under the perspective of UAV. Results show that, within 15 meters' “High Sensitivity Ranging”, the UV lens can still maintain effective capabilities of defect identification and positioning even if the received defect discharge signal is weak to an approximate interference level; under rated voltage and if sheath is undamaged, only the IR heating of mandrel rot defect can be observed but without UV discharge; if sheath is damaged, the relative position characteristics of 3 typical defects' heating and discharge are different; the complementarity of IR and UV detection endows multi-spectral detection a good reliability under unfavorable camera angles.