• DocumentCode
    2682922
  • Title

    Application of an automated PD failure identification system for EMC-assessment strategies of multiple PD defects at HV-insulators

  • Author

    Aschenbrenne, Dirk ; Lapp, Andreas ; Kranz, Hans-Gerd ; Borneburg, D. ; Peier, Dirk

  • Author_Institution
    Lab. of High Voltage Eng., Wuppertal, Germany
  • fYear
    2000
  • fDate
    2000
  • Firstpage
    112
  • Lastpage
    117
  • Abstract
    EMC-assessment of field emission of high voltage insulators can be performed using phase-angle-resolved partial discharge diagnosis. However, with conventional PD-detection systems no satisfying statements about multiple PD defects are possible because of the highly dynamic apparent charge values for different discharge phenomena. This measurement problem can be solved by a new approach. For this purpose phase resolved pulse sequence analysis methods are suitable diagnosis tools without using the apparent charge as a dominating influence. A recently developed feature extraction method based on consecutive Δu/Δφ values shows good classification results. The problem of multiple PD defects, which occur at the same time, is a new challenge for PD diagnosis systems. For the investigation two reference databases are generated. With the database, which takes these multiple PD defects into account, the diagnosis system WinTED of the University of Wuppertal is able to identify with high reliability actual measurements made by the University of Dortmund
  • Keywords
    automatic testing; failure analysis; feature extraction; insulator testing; partial discharge measurement; EMC-assessment strategies; HV-insulators; WinTED; automated PD failure identification system; discharge phenomena; feature extraction method; field emission; multiple PD defects; phase resolved pulse sequence analysis methods; phase-angle-resolved partial discharge diagnosis;
  • fLanguage
    English
  • Publisher
    iet
  • Conference_Titel
    Dielectric Materials, Measurements and Applications, 2000. Eighth International Conference on (IEE Conf. Publ. No. 473)
  • Conference_Location
    Edinburgh
  • Print_ISBN
    0-85296-730-6
  • Type

    conf

  • DOI
    10.1049/cp:20000488
  • Filename
    888097