Detection And Assessment Of Contaminated Suspension Insulator Discharge Based On Ultraviolet Imaging

Contaminated suspension insulator surface discharge will radiate ultraviolet (UV) signal, solar-blind UV imaging is a non-contact and visual discahrge detection method, firstly, the correlation of UV imaging parameters and electric pulse sigal is analyzed, then mainly three problems of composite insulator discharge intensity estimation, ceramic insulator contamination condition assessment, insulation condition assessment and flashover prewarning are studied in this paper, the main work and achievements are as follows.The discharge optical radiation mechanism, spectrum distribution and the UV signal attenuation problem in air is analyzed, then the correlation of optical radiation power and discharge power is studied, and the solar blind UV imager structure and working principle is analyzed, then a method based on vedio analysis and digital image processing is proposed to quantify the discharge character,and the facula area parameter is defined, then the affect law of gain to the facula area is studied and gave a theoretical explanation.A method of quasi-sychronous acquisition and analysis electric pulse signal and UV imaging parameters is proposed, then the response characteristics of facula area to the electrical pulse signals is studied and gave a theoretical analysis, the correlation of large facula area frames with the number of spark (arc) discharge number in certain time is analyzed, the relation of facula area with the electrical pulse amplitude is studied and get the corresponding curves.The relation of image size in UV images with the detection distance is studied under different gains, it showes that the UV channel image and visible images size increase or decrease with the same ratio to the observed distance, this paper put forward a method to estimate the discharge imaging regional area with visible channels image information, and deduced the calculation formulation, then the relation curves of imaging regional area with the discharge pulse amplitude was obtained, this method resolve the problem that the facula area vary with the observe distance.Using least squares support vector regression machine realized the discharge intensity estimation.Based on the artifical climate chamber, the contaminated insulator discharge UV imaging characters under different polllution degrees and humidity are analyzed,and then this paper put forward two UV imaging parameters of average relative facula area and large facula area number in one minute to characterize the discharge feature. A contamination condition assessment model is established with fuzzy logic inference system, and realized the contamination contiditon assessment.The single ceramic insulator UV image feature within the procedure from weak discharge to the finally flashover with the gradual increased voltage is studied, and then this paper put forward the relative facula diameter parameter, according to this parameter and dishcarge phenominon, the insulator insulation condition is divided into four grades, and realizes the insulation condition assessment and flashover prewarning.