Research On The Ultraviolet Imaging Diagnosis Method Of Faulted Insulator And Its Influence Factors

Ultraviolet imaging(UV)technology has been widely used in the discharge detection of faulty insulators.To solve the shortage of quantifying the discharge intensity by "photon number" in the traditional UV imaging method,an UV imaging diagnosis method of faulted insulator is proposed in this thesis,and the experimental study and analysis of the influencing factors on UV imaging detection were carried out.Firstly,combined with Thomson discharge theory and the streamer discharge theory,the principle of insulator corona discharge is analyzed.The principle and process of the pollution flashover discharge of insulators are also discussed to demonstrate the necessity of UV imaging detection for polluted insulators.Then,the working and imaging principles of the solar-blind UV imager are analyzed,and the effectiveness of detecting the insulator discharge by UV imaging is verified theoretically.The noise degradation and recovery model of insulator UV image is analyzed.NSE,MSE,SNR,and PSNR are used as the objective evaluation criteria of image denoising.Through the adaptive median filter algorithm,the ideal denoising effect is achieved.The UV image is enhanced by Gamma correction with appropriate correction coefficient.While ensuring the background recognition degree of the image,the UV spot area is made more prominent,which is favorable for effective identification and discrimination of the discharge area.An image parameter quantization method for fault insulator discharge intensity is proposed.Firstly,threshold segmentation by Otsu method is performed on the image and then followed by the binary open operations of mathematical morphology.Finally,Canny edge detection operator is used to fulfill the edge detection and contour extraction of UV spot areas.The image parameters such as spot area,perimeter,and equivalent diameter are calculated as characteristic parameters and quantification criteria for insulator discharge intensity.Finally,combined with the UV image diagnosis method proposed in this thesis,the artificial contamination flash experiments are carried out.Experimental results show that the effect of ambient humidity is not constant,it changes from weakened discharge to enhanced discharge with increasing humidity.Relatively high ambient humidity will weaken the UV signal received by the UV imager while enhancing the discharge intensity,causing detection errors.As the detection distance increases,the detected UV signal decreases exponentially.In practical applications,it should be as close as possible to the detection object to improve the accuracy under the premise of ensuring safety.The detection angle has no significant effect on the UV imaging detection and the error is negligible.The detected spot area increases with the increase of the UV imager gain.But if the gain is too high,it will cause a large error.Therefore,the gain should be set within a reasonable range to ensure the detection accuracy.The influence of environmental factors on the detection of insulator discharge by UV imaging is analyzed through the experiments,and feasible suggestions for practical discharge detection are given in this thesis.