• DocumentCode
    2893753
  • Title

    An Initial Analysis of Fault Diagnosis to Electrical Apparatus Product on Bayesian

  • Author

    Liu, Jiao-min ; Wang, Jing-hong ; Li, Bi ; Zhang, Chang-yong

  • Author_Institution
    Inst. of Electr. Apparatus, Hebei Univ. of Technol., Tianjin
  • fYear
    2006
  • fDate
    13-16 Aug. 2006
  • Firstpage
    2474
  • Lastpage
    2478
  • Abstract
    This paper is based on the fault diagnosis analysis about electrical product of Bayesian network. We represent fault diagnosis as decision problem under ambiguity and immaturity of information. Bayesian network classifiers (BNC) is thus established which is based on the fault diagnosis of, and representing the ambiguity of information as probability description. Pre-process analysis is made to fault information and the attribute technology in term of maximum information entropy approach is brought forth. We realize electrical product fault diagnosis under the ambiguity and immaturity of information; further research and put forward probability study new arithmetic of BNC. We apply Bayesian network classifiers technology to electrical system fault diagnosis, and make it play a better role
  • Keywords
    belief networks; entropy; fault diagnosis; power apparatus; power engineering computing; probability; Bayesian network classifier; decision problem; electrical apparatus product; fault diagnosis; information entropy; preprocess analysis; probability; Artificial intelligence; Bayesian methods; Classification tree analysis; Context modeling; Cybernetics; Data mining; Electrical products; Fault diagnosis; Information analysis; Information entropy; Machine learning; Random variables; Bayesian network classifiers; Electrical Apparatus; Fault diagnosis; Information entropy;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Machine Learning and Cybernetics, 2006 International Conference on
  • Conference_Location
    Dalian, China
  • Print_ISBN
    1-4244-0061-9
  • Type

    conf

  • DOI
    10.1109/ICMLC.2006.258782
  • Filename
    4028480