Detection of Broken Rotor Bars in Presence of Load Oscillations

In industrial applications, the reliable performance of induction motors (IM) is essential. This paper proposes an improved method to distinguish between broken rotor bar (BRB) faults and load oscillations (LO) by analyzing stator current components in the rotor-flux reference frame. The approach em...

Full description

Saved in:
Bibliographic Details
Main Authors: Klemen Drobnic, Mitja Nemec, Henrik Lavric, Vanja Ambrozic, Rastko Fiser
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10908227/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850068457835462656
author Klemen Drobnic
Mitja Nemec
Henrik Lavric
Vanja Ambrozic
Rastko Fiser
author_facet Klemen Drobnic
Mitja Nemec
Henrik Lavric
Vanja Ambrozic
Rastko Fiser
author_sort Klemen Drobnic
collection DOAJ
description In industrial applications, the reliable performance of induction motors (IM) is essential. This paper proposes an improved method to distinguish between broken rotor bar (BRB) faults and load oscillations (LO) by analyzing stator current components in the rotor-flux reference frame. The approach employs a diagnostic fault indicator based on the amplitude ratio of current oscillations in the direct and quadrature axes, with a threshold established through simulation data and a support vector classifier (SVC). If the detected spectral peak pair corresponds to a BRB, the fault indicator exceeds the threshold. Conversely, peaks caused by LO result in an indicator value significantly below this threshold. The explicit and unambiguous separation between BRB and LO effects allows for the practical recommendation of a unity threshold. Simulations and experimental results validate the method’s robustness, demonstrating accurate fault identification even under challenging conditions, such as lightly loaded motors with a single BRB. The method relies solely on current and voltage measurements, avoiding the need for speed sensors. With low computational demands, it integrates seamlessly into existing motor drive systems. These features make it a practical and cost-effective tool for real-time fault detection and predictive maintenance in industrial settings.
format Article
id doaj-art-0960a3e31fc44014948b108ebd802f68
institution DOAJ
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-0960a3e31fc44014948b108ebd802f682025-08-20T02:48:03ZengIEEEIEEE Access2169-35362025-01-0113400124002710.1109/ACCESS.2025.354670510908227Detection of Broken Rotor Bars in Presence of Load OscillationsKlemen Drobnic0https://orcid.org/0000-0002-2444-2327Mitja Nemec1https://orcid.org/0000-0002-8774-1025Henrik Lavric2https://orcid.org/0009-0000-7746-4522Vanja Ambrozic3https://orcid.org/0000-0002-3468-4581Rastko Fiser4https://orcid.org/0000-0002-0740-4775Faculty of Electrical Engineering, University of Ljubljana, Ljubljana, SloveniaFaculty of Electrical Engineering, University of Ljubljana, Ljubljana, SloveniaFaculty of Electrical Engineering, University of Ljubljana, Ljubljana, SloveniaFaculty of Electrical Engineering, University of Ljubljana, Ljubljana, SloveniaFaculty of Electrical Engineering, University of Ljubljana, Ljubljana, SloveniaIn industrial applications, the reliable performance of induction motors (IM) is essential. This paper proposes an improved method to distinguish between broken rotor bar (BRB) faults and load oscillations (LO) by analyzing stator current components in the rotor-flux reference frame. The approach employs a diagnostic fault indicator based on the amplitude ratio of current oscillations in the direct and quadrature axes, with a threshold established through simulation data and a support vector classifier (SVC). If the detected spectral peak pair corresponds to a BRB, the fault indicator exceeds the threshold. Conversely, peaks caused by LO result in an indicator value significantly below this threshold. The explicit and unambiguous separation between BRB and LO effects allows for the practical recommendation of a unity threshold. Simulations and experimental results validate the method’s robustness, demonstrating accurate fault identification even under challenging conditions, such as lightly loaded motors with a single BRB. The method relies solely on current and voltage measurements, avoiding the need for speed sensors. With low computational demands, it integrates seamlessly into existing motor drive systems. These features make it a practical and cost-effective tool for real-time fault detection and predictive maintenance in industrial settings.https://ieeexplore.ieee.org/document/10908227/Broken rotor barsfault detectioninduction machinemotor current signature analysissupport vector machine classification
spellingShingle Klemen Drobnic
Mitja Nemec
Henrik Lavric
Vanja Ambrozic
Rastko Fiser
Detection of Broken Rotor Bars in Presence of Load Oscillations
IEEE Access
Broken rotor bars
fault detection
induction machine
motor current signature analysis
support vector machine classification
title Detection of Broken Rotor Bars in Presence of Load Oscillations
title_full Detection of Broken Rotor Bars in Presence of Load Oscillations
title_fullStr Detection of Broken Rotor Bars in Presence of Load Oscillations
title_full_unstemmed Detection of Broken Rotor Bars in Presence of Load Oscillations
title_short Detection of Broken Rotor Bars in Presence of Load Oscillations
title_sort detection of broken rotor bars in presence of load oscillations
topic Broken rotor bars
fault detection
induction machine
motor current signature analysis
support vector machine classification
url https://ieeexplore.ieee.org/document/10908227/
work_keys_str_mv AT klemendrobnic detectionofbrokenrotorbarsinpresenceofloadoscillations
AT mitjanemec detectionofbrokenrotorbarsinpresenceofloadoscillations
AT henriklavric detectionofbrokenrotorbarsinpresenceofloadoscillations
AT vanjaambrozic detectionofbrokenrotorbarsinpresenceofloadoscillations
AT rastkofiser detectionofbrokenrotorbarsinpresenceofloadoscillations