Research On On-Line Insulation Monitoring System And Diagnosis Metnod Of MOA Based On Virtual Instrument

According to the advantages of good protection ability, large discharge capacity, performance reliability, simple and convenient maintenance and so on, Metal Oxide surge Arrester (MOA) becomes the important protective equipment which protects AC electrical equipments from lightning overvoltage and switching overvoltage damaging. However, MOA tends to ageing because of the weather, temperature, lightning strike, switching overvoltage and internal moisture acting on long term, which makes the protective functions losed and has an effect on the operating of system safely when it is serious. Therefore, in order to prevent from accident, it will be of significance to carry on on-line monitoring to insulation condition. After analyzing the actuality of on-line monitoring techniques, this thesis designs an on-line insulation monitoring system of MOA based on VI and LabVIEW, and presents the insulation diagnosis method of MOA based on fuzzy inference.In this thesis, firstly, software and hardware schemes are designed based upon the characteristics of on-line insulation monitoring system of MOA;hardware platform is builded after sensors choosing, data acquisition card selecting and signal preprocessing circuits devising finished; secondly, fundamental resistive current I_1r third harmonic resistive current I_3r and dissipation factor tgσ are established as the monitoring data, which are calculated by harmonic analysis and managed by Access database; lastly, software is designed by LabVIEW.Aimming at lots of uncertain and fuzzy factors between failure reason and fault symptom of MOA, diagnosis method to insulation condition of MOA is put forward based on fuzzy inference. Subtractive clustering method is applied to clustering computing about sample data from laboratory first; Then fuzzy membership functions and rules of fuzzy diagnosis system are established based on clustering results and expert's diagnostic experiences; At last, fuzzy inference system which is builded in MATLAB and experimental data are applied to verify the performance of fuzzy insulation diagnosis system of MOA, and the results indicate that this...