Active and Reactive Power Spectra-Based Detection and Separation of Rotor Faults and Low-Frequency Load Torque Oscillations

Mhamed Drif, Heonyoung Kim, Jongwan Kim, Sang Bin Lee, Antonio J.Marques Cardoso

Research output: Contribution to journalArticlepeer-review

18 Citations (Scopus)


Motor current signature analysis (MCSA) based detection of rotor cage faults (RFs) in medium-high voltage induction motors has become common in the field due to its low-cost remote monitoring capability. However, low-frequency load oscillations (LOs) in applications that employ speed reduction couplings can induce frequency components in the vicinity of that of RFs. It has been reported that false-positive or -negative RF indications can be produced with the MCSA, if the frequency components produced by RF and LO overlap. There, recently, have been studies on discriminating RF and LO; however, the case where two harmonic components overlap at the same frequency, which is the most difficult and serious case, has not been investigated. In this paper, a spectrum analysis of instantaneous active and reactive power is applied, and it is shown that RF and LO can be separated for the case where the MCSA fails due to the overlap in the two spectral components. An experimental study performed on a 380 V motor with a speed reduction gear shows that RF and LO can be reliably detected and separated for avoiding false RF indications.

Original languageEnglish
Article number7576687
Pages (from-to)2702-2710
Number of pages9
JournalIEEE Transactions on Industry Applications
Issue number3
Publication statusPublished - 2017 May 1


  • Active power
  • belt pulley
  • coupling
  • electrical fault detection
  • fault diagnosis
  • gears
  • induction motors
  • load oscillation (LO)
  • reactive power
  • rotor cage fault (RF)
  • spectral analysis

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Industrial and Manufacturing Engineering
  • Electrical and Electronic Engineering


Dive into the research topics of 'Active and Reactive Power Spectra-Based Detection and Separation of Rotor Faults and Low-Frequency Load Torque Oscillations'. Together they form a unique fingerprint.

Cite this