Study On AC Surface Discharge Based On Visible Light Image Chromaticity And Emission Spectrum

Features about surface discharge of the equipment largely affect the quality of its insulation performance.It is of great practical significance to study the physical mechanism and diagnostic technology of surface discharge to improve the insulation design of electrical equipment,implement insulation monitoring and prevent insulation accidents.The visible light discharge image taken by ordinary digital cameras contains rich spatial chromaticity and shape information,and the long exposure time can better reflect the statistical law of discharge,but the existing research rarely applies it to the detection of creeping discharge.By observing and analyzing the visible light image and emission spectrum of discharge,the AC surface discharge in both dark environment and sunlight environment are studied respectively,and the regular pattern of chromaticity change in the surface discharge process under dark environment is summarized,and a chromaticity diagnostic indicator of the surface discharge severity in the sunlight environment is proposed,providing new ideas and perspectives for the study of digital images of surface discharge.For the surface discharge in the dark environment,experiments were carried out with a white ceramic plate and a red phenolic resin plate,and chromaticity coordinates at different voltages are computed with the three primary colors RGB data of the visible light image and the visible light band data of the emission spectrum,respectively.This establishes the relationship between the surface discharge emission spectrum and its visible light image chromaticity index,achieving the connection between the microscopic physical radiation process and the macroscopic physical state quantity.It can be found that:(1)Both calculation methods indicate that as the voltage rises,the chromaticity coordinates are approaching the pure blue point along the line between the equipotential white point and the pure blue point,which reflects that the discharge luminescence will show more and more saturated blue component with increasing voltage;(2)the dynamic range of the chromaticity coordinates of the emission spectrum is greater than that of the visible light image;the rate of change of the chromaticity coordinates of the emission spectrum with the voltage is greater than that of the visible light image;(3)The chromaticity coordinates of the visible light image show a more stable and more linear law with the voltage change;(4)The chromaticity coordinates of the emission spectrum are basically unaffected by the medium color while that of the visible light image are obviously affected by the medium color,and the degree of influence is positively correlated with the intensity of the discharge light radiation and can be significantly weakened by substituting some pixels with lower brightness instead of all those pixels of the image into the calculation.The results of the above-mentioned association studies will help to improve the technology of identifying the discharge state with visible light image chromaticity information.For the surface discharge in sunlight environment,in order to overcome the influence of daylight and medium color on the chromaticity index,this paper proposes an analysis method that combines the colorimetric index and the statistical index to detect chromaticity mutation in partial area of the medium surface,which can be used to diagnose surface discharge.(1)Using the chromaticity variable hue,transformed from RGB,to describe the color,which can overcome the influence of exposure intensity.(2)Using the statistic standard score,marked as Z,to measure the degree of hue deviation caused by the surface discharge on the medium,and the identification criterion of hue mutation is set as |Z|>3.0,which overcomes the influence of sunlight and medium color.(3)The area ratio of the region where hue mutation exists,marked as P,is used to measure the development degree of surface discharge.(4)Experiments were carried out with white ceramic plates and red electric boards respectively.For both media,P increases with voltage,and 0<P?0.1,0.1<P?0.5,P>0.5 can correspond to the initial stage,the development stage,the pre-breakdown stage of surface discharge,respectively.This shows that the combination of visible light image chromaticity information and statistic indicators can diagnose the state of surface discharge in sunlight environment.