• DocumentCode
    761993
  • Title

    An integrated on-line motor protection system

  • Author

    Farag, S.F. ; Bartheld, R.G. ; Habetler, T.G.

  • Author_Institution
    Siemens Energy & Autom. Inc., Tucker, GA, USA
  • Volume
    2
  • Issue
    2
  • fYear
    1996
  • Firstpage
    21
  • Lastpage
    26
  • Abstract
    Motor protection equipment is typically aimed at protecting the motor insulation against overheating. Additional protection can be gained by applying expensive equipment based on sensing vibration and/or the current spectra of the motor to predict incipient failures. These systems, which have only recently been introduced, have limited application because of the cost and expertise required to operate them. This article proposes a single, low-cost, microprocessor-based unit to protect the motor against-or to alert the user to-the majority of motor failure modes. The proposed system includes thermal modeling for overload protection and variable ground fault protection. Mechanical failure prediction using current spectral analysis is also provided using a frequency selective expert system and a neural network to distinguish potential fault conditions from unknown online operating conditions
  • Keywords
    diagnostic expert systems; electric current measurement; failure analysis; machine testing; microcomputer applications; motor protection; neural nets; power engineering computing; reliability; thermal analysis; vibration measurement; current spectra monitoring; current spectral analysis; failure modes; frequency selective expert system; integrated online motor protection system; mechanical failure prediction; microprocessor; neural network; overload protection; potential fault conditions; thermal modeling; unknown online operating conditions; variable ground fault protection; vibration sensing; Automation; Costs; Expert systems; Frequency; Insulation; Micromotors; Neural networks; Protection; Spectral analysis; Vibrations;
  • fLanguage
    English
  • Journal_Title
    Industry Applications Magazine, IEEE
  • Publisher
    ieee
  • ISSN
    1077-2618
  • Type

    jour

  • DOI
    10.1109/2943.485759
  • Filename
    485759