• DocumentCode
    2598106
  • Title

    A method for fault line detecting based on cluster in resonant grounded systems

  • Author

    Li Bin ; Shu Hongchun

  • Author_Institution
    Kunming Univ. of Sci. & Technol., Kunming, China
  • fYear
    2009
  • fDate
    6-7 April 2009
  • Firstpage
    1
  • Lastpage
    4
  • Abstract
    In resonant grounded systems, when single phase-to-earth fault occurring, transient components of zero-sequence current in fault lines are composed of that in healthy lines. So there is a great difference between health lines and fault lines of transient information. This paper gives a new diagnostic method based on cluster. A cluster is an assemblage constituted by a set of objects with the character of similarity each other. But the objects in different clusters usually are not similar. In the process of cluster, the more similar the variables are, the more precedent they merged. Not more similar, merged later. The degree of similarity of different variables merged last is the lowest. When single phase-to-earth fault occurs, the transient information of zero-sequence current is seen as variables of cluster. A detecting criterion is built up by calculating and analyzing the statistics of cluster. The process of cluster can be shown though scatter plot and binary tree diagram. Simulation results show that the proposed method is convenient, symbolic and reliable.
  • Keywords
    fault location; power distribution faults; fault line detecting; resonant grounded systems; single phase-to-earth fault; zero-sequence current; Assembly; Cascading style sheets; Circuit faults; Coils; Current transformers; Fault detection; Inductance; Resonance; Statistical analysis; Statistical distributions; Cluster Analysis; Fault Line Selecting; Resonant Grounded System; Single Phase-to-Earth Fault;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Sustainable Power Generation and Supply, 2009. SUPERGEN '09. International Conference on
  • Conference_Location
    Nanjing
  • Print_ISBN
    978-1-4244-4934-7
  • Type

    conf

  • DOI
    10.1109/SUPERGEN.2009.5347955
  • Filename
    5347955