Investigation On Online Monitoring Technology Of Zinc Oxide Arresters

With the development of power system,higher requirements are put forward for the security and stability of power system operation.The traditional periodic outage detection can not fully meet our needs,so the maintenance mode gradually develops to real-time online monitoring.The arrester,operating as the protector in power system,is always under the action of power frequency voltage for a long time,and influenced by environmental factors,harmonics of power grid,over-voltage,internal moisture and other factors.Its insulation performance of arrester will decline in varying degrees in the course of operation,and will lose its protective effect on the system and even explode in serious cases.The operation status of the arrester can be monitored in real time through on-line monitoring.The leakage current and its resistive component of zinc oxide arrester can be used as the main characteristic parameters to analyze the insulation state of arrester.In this paper,an on-line monitoring system for zinc oxide arrester is designed by means of the full current method,fundamental wave method and harmonic method.The system uses sensors to extract features,uses high-speed A/D sampling controlled by FPGA to collect data,and achieves a maximum of 16 channels of 200 kSPS sampling;uses GPS timing device to acquire precise time for the system,and achieves a timing acquisition and real-time data and historical data by controlling and displaying the sampling data through the touch screen of the host computer.Display;FTP server is established,and data can be read through LAN/WAN.A high voltage test environment for 10 kV lightning arrester has been built in the laboratory.The maximum applied voltage can reach over 20 kV AC and 30 kV DC.Applying 5 kV-13.6 kV to the arrester,the eigenvalue changes of the arrester under different voltages were measured.In addition,lightning arrester usually operates outdoors,and its on-line monitoring will be affected by the changes of environmental conditions.In this paper,experiments are carried out under various conditions,including temperature from 20 to 60 degrees,relative humidity from 20 to 80 percent and pollution degree from 20 g/L to 160g/L.The arrester was accelerated aging and heated for 100 hours at 200 C.The insulation condition of the arrester was deteriorated to a great extent.Because the characteristic value of arrester changes more complex,in order to better judge the insulation state of zinc oxide arrester,this paper establishes a prediction model based on BP neural network.Its input is the third harmonic component of surface contamination,temperature,relative humidity,total current,resistive current and resistive current.After training and testing,the relative error of the model is less than 1%.The results show that the model can be used to predict the insulation state of zinc oxide arrester.By removing the interference of environmental factors on the judgment of the insulation state of arresters,the insulation state of arresters can be accurately judged.