Research On Power Line Corona UV Detection Method Based On UAV

The main cause of power line failure includes the insulation equipment defect on the line.Equipment deterioration causes uneven distribution of electric field on the surface,causing corona discharge.By detecting the early weak discharge phenomenon,the degree of equipment defects can be evaluated.This method can achieve the effect of early warning.It is of great significance to maintain the safe and stable operation of the power grid.Corona discharge radiates signals such as light waves,electromagnetic waves,and electric pulses.High-sensitivity photoelectric sensors can be used to find weaker discharge signals.Bydetecting the ultraviolet light in the“solar blind”section,background light interference can be avoided to achieve all-weather corona detection.The UAV airborne corona ultraviolet detection device is used for power line inspection,and its high efficiency and maneuverability can adapt to various harsh environments.Therefore,the corona detection and location method in UAV line patrol is studied in this thesis.The specific works are as follows:(1)According to the characteristics of corona discharge,the factors affecting corona discharge are analyzed.The relationship between corona discharge intensity and ultraviolet light signal is studied.The corona discharge detection system is designed using a“solar blind”type photomultiplier tube(PMT).The corona detection experiment was completed by the discharge experiment platform at different distances,different angles and different humidity conditions.The experimental results were analyzed and results show that the detection system has high sensitivity.The discharge signal can be detected within 40m,and the detection effect is better at 10m?20m.These place can distinguish the degree of corona discharge of different strengths.(2)Corona ultraviolet power detection model is established based on the UV light scattering link model.The methods of monocular and binocular detection and positioning in the drone track have been studied.The position of the corona discharge point is located by two different inspection methods.The experimental scheme is designed and analog light source and corona discharge source are tested for power detection.The detection and positioning accuracy was analyzed through experiments.The results show that when the detection distance is 20m,the position of discharge point can be determined and the error is less than 1.6m.The effectiveness of the positioning method is verified by experiments(3)To evaluate the corona discharge intensity in the process of UAV line patrol,a corona recognition method based on ANFIS for UAV line patrol is designed in this thesis.The characteristics of the received single pulse signal are extracted and analyzed.The effects of different intensity sources on the received pulse signal intensity are studied under different humidity conditions.Finally,the discharge level evaluation model is established.By comparing the ANFIS algorithm with the BP neuro-fuzzy algorithm,the discharge model in this thesis can accurately assess the level strength of the discharge source.