The Study Of Partial Discharge And Pattern Recognition Of Three-phase Common GIS

Gas Insulated Switchgear (GIS) are widely used in power systems due to a seriesof advantages, gas insulated switchgear has coaxial and three-phase common structure,the coaxial GIS study abroad more fully, the cavity structure and common GIS shaft hasa big difference because of its three-phase conductor cavity in three-phase commondrum GIS,so the research of the three-phase common GIS is important to ensure its safeoperation.GIS in the event of partial discharge chamber will produce a large amount ofelectromagnetic waves, the paper coaxial conductor parallel plate waveguide and anexample of the transverse electromagnetic, transverse wave propagation characteristics,were studied transverse waves. the propagation characteristics of electromagnetic waveswithin the GIS were studied during the propagation of electromagnetic wave reflectionGIS cavity through Ansoft simulation software, decay, leakage characteristics wereanalyzed. PD propagation characteristics of electromagnetic waves generated within GISpartial discharge signals detected defect has important reference value.In three-phase common GIS partial discharge detection cartridge, this paperdesigned a four typical GIS insulation defects within the three-phase model of cartridge,using a UHF partial discharge detection method, produced by Germany’s ultra-PDSGHigh-frequency partial discharge measuring instrument, the local co-barrel within athree-phase fault GIS typical insulation defects were tested discharge volume. Test toobtain a large number of partial discharge patterns of various types of PD patterns werestudied and analyzed, Bureau of test data obtained to put maps and GIS Drum favor of athree-phase PD identified.In total three-phase common GIS PD type drum pattern recognition, this paper puton a variety of defects Bureau obtained characteristic parameters were extracted.According extracted skewness, steepness, PD peak points, such as the relationshipbetween the number of each characteristic parameters using fuzzy K-nearest neighboralgorithm and Fisher recognition algorithms respectively Bureau put four typicalinsulation defects were identified, and achieved better recognition results.Finally, the three-phase of the GIS typical insulation defects drum patternrecognition algorithm to identify online monitoring prepared by LabVIEW software platform for GIS typical three-phase defect type of drum pattern recognition. Thesoftware has a data input, graphical display, alarm function, the actual test results showthat the design phase of GIS drum line detection software function better insulation forthe three-phase common GIS PD defect types with a more high recognition rate.