energies

Review
Review of Fault Detection and Diagnosis Techniques for AC
Motor Drives
Muhammed Ali Gultekin * and Ali Bazzi *

Electrical and Computer Engineering Department, University of Connecticut, Storrs, CT 06269, USA
* Correspondence: [email protected] (M.A.G.); [email protected] (A.B.)

Abstract: Condition monitoring in electric motor drives is essential for operation continuity. This
article provides a review of fault detection and diagnosis (FDD) methods for electric motor drives. It
first covers various types of faults, their mechanisms, and approaches to detect and diagnose them.
The article categorizes faults into machine faults, power electronics (PE) faults, DC link capacitor
faults, and sensor faults, and discusses FDD methods. FDD methods for machines are categorized
as statistical methods, machine-learning methods, and deep-learning methods. PE FDD methods
are divided into logic-based, residual-based, and controller-aided methods. DC link capacitor and
sensor faults are briefly explained. Machine and PE faults are listed and presented as tables for
easy comparison and fast referencing. Most papers are selected from the past five years but older
references are added when necessary. Finally, a discussion section is added to reflect on current trends
and possible future research areas.

Keywords: motor drives; condition monitoring; fault detection and diagnosis; fault mechanism;
power electronics; power electronics faults; machine faults

1. Introduction
Electric motors are the powerhouse of the industry with applications ranging from
manufacturing to transportation. With the rise of electric vehicles (EVs) and the push
Citation: Gultekin, M.A.; Bazzi, A.
for electrification, their usage is increasing. Keeping motors and their drives healthy is
Review of Fault Detection and
crucial to maintain operation continuity or service uptime. However, due to environmental
Diagnosis Techniques for AC Motor
conditions, regular wear and tear, installation and manufacturing defects, or overloading,
Drives. Energies 2023, 16, 5602.
https://doi.org/10.3390/en16155602
electric motor drives are subject to failures. These failures can be on the motor side or the
power electronics side. Detection of these faults is an indispensable function for operation
Academic Editors: Ahmed safety, fault tolerance, mission completion, or fast maintenance.
Abu-Siada, Pietro Romano and In the literature, there are several fault detection and diagnosis (FDD) review articles
Gianluca Brando
for electric machines and drives. It is possible to find review articles ranging from widely in-
Received: 2 June 2023 clusive to narrowly focused ones. The focus of these reviews can be certain fault types such
Revised: 10 July 2023 as stator faults [1], bearing faults [2], or sensor faults [3]. Other reviews focus on specific
Accepted: 22 July 2023 FDD methods such as machine learning (ML) [4,5], deep learning (DL) [6], or finite element
Published: 25 July 2023 analysis [7]. The focus can also be a certain type of motor or drive topology, for example,
induction motors (IM) [8], permanent magnet motors [9], or multilevel inverters [10]. For
established researchers in a field, narrowly focused reviews might be more practical as
they touch on topics in greater detail. These reviews are looking to detect a specific fault in
Copyright: © 2023 by the authors. a specific machine using a predefined method. Some examples are [11,12]; wherein [12],
Licensee MDPI, Basel, Switzerland. articles detecting broken rotor bar faults are looked at for IMs, utilizing fault-signature
This article is an open access article
analysis. Similarly, in [11], bearing fault detection studies in IMs using deep-learning
distributed under the terms and
methods are collected. These reviews provide indepth insight into their focus area.
conditions of the Creative Commons
Presenting the latest research with accessible tables is also important. Some reviews
Attribution (CC BY) license (https://
lack these tables for fast referencing [13]. Another point is fault mechanisms are not
creativecommons.org/licenses/by/
necessarily provided in all review articles. Thus, the existing literature may not be suitable
4.0/).

Energies 2023, 16, 5602. https://doi.org/10.3390/en16155602 https://www.mdpi.com/journal/energies

 dition monitoring in IM drives.
Our aim with this review article is to show possible faults in motors and other drive
components, explain their mechanisms in brief, and provide potential approaches to de-
tect and diagnose these faults. We aim to present a big picture without missing the essence
Energies 2023, 16, 5602
of each component. 2 of 22

Methodology
In this review, drive faults are addressed as four categories: machine faults, power
for a researcher trying to get into this field or someone who needs a broad review of
electronics faults, DC link capacitor faults, and sensor faults. Condition monitoring meth-
condition monitoring in IM drives.
ods for
Ourmachine
aim with FDD
thisare divided
review into
article is statistical methods,
to show possible ML-based
faults in motorsmethods,
and otheranddrive
DL-
based methods. Though ML- and DL-based methods contain similar preprocessing
components, explain their mechanisms in brief, and provide potential approaches to detect stages
to
andstatistical
diagnose methods, they We
these faults. are treated as separate
aim to present a bigcategories due to their
picture without difference
missing in later
the essence of
modeling stages
each component. where neural networks (NN) are employed. We realized the categoriza-
tion of some of the methods is not always straightforward and might be harder to put
them into one category. We used our judgement to categorize them as shown in Figure 1.
Methodology
Power electronics
In this review, driveFDD
faultsmethods are divided
are addressed as four into logic-based,
categories: machine residual-based, and
faults, power elec-
controller-aided methods. Logic-based methods do not require a model
tronics faults, DC link capacitor faults, and sensor faults. Condition monitoring methods of the system to
detect failures but may utilize a model or an actual drive to determine
for machine FDD are divided into statistical methods, ML-based methods, and DL-based signal thresholds.
Residual-based
methods. Though methods
ML- and require a system
DL-based model.contain
methods Controller-aided methods alsostages
similar preprocessing use sys-
to
tem models
statistical but thesethey
methods, models are usedasfor
are treated controlcategories
separate as well; there
dueisto
notheir
needdifference
for an additional
in later
FDD model
modeling of the
stages system.
where neural Sensor and DC
networks (NN)link
arecapacitor
employed.FDD are briefly
We realized theexplained and
categorization
critical points and overlaps with other methods are given. The categorization
of some of the methods is not always straightforward and might be harder to put them into process is
depicted in Figure 1.
one category. We used our judgement to categorize them as shown in Figure 1.

Figure
Figure 1.
1. Flowchart
Flowchart of
of FDD-method
FDD-method categorization.
categorization.

Machine FDD methods

Power electronics FDD are summarized
methods and compared
are divided in Tablesresidual-based,
into logic-based, 1–3 for each ofand
the
previously classified
controller-aided categories.
methods. Also, some
Logic-based of thedo
methods common open-source
not require a modeldata repositories
of the system to
containing machines
detect failures and
but may drives
utilize relatedordata
a model are listed
an actual into
drive Table 4 for interested
determine readers.
signal thresholds.
Power electronics
Residual-based faults are
methods presented
require a systemin model.
Table 5.Controller-aided methods also use system
models but these models are used for control as well; there is no need for an additional
FDD model of the system. Sensor and DC link capacitor FDD are briefly explained and
critical points and overlaps with other methods are given. The categorization process is
depicted in Figure 1.
Machine FDD methods are summarized and compared in Tables 1–3 for each of the
previously classified categories. Also, some of the common open-source data repositories
containing machines and drives related data are listed in Table 4 for interested readers.
Power electronics faults are presented in Table 5.
Most citations are from the last five years but some critical references are added
regardless of their publication years. Also, when selecting prior work, we tried to give
 A motor drive system consists of a motor, power electronic converters, sensors, an
a controller unit, as shown in Figure 2. Apart from the main DC–AC inverter, based o
the energy source, the drive may contain DC–DC converters if it is fed from a ba ery o
AC–DC rectifiers if supplied by an AC source. The industry standard for the DC–AC in
Energies 2023, 16, 5602
verter is accepted to be three-phase two-level inverters. If not mentioned specifically, 3 of 22
suc
as multilevel inverters, this topology is considered throughout the article. Multiphas
drives beyond three-phase are out of the scope of our paper.
Any damage
equal weight or failure
to different to these
approaches. Forcomponents posesbearing
example, to detect the riskfaults,
of a vibration
complete system
breakdown. Some parts of the drive are more susceptible to failure
signals are used mainly, and there is more literature on detection using vibration signals. than others. Reliabilit
analyseswe
However, forincluded
motor drivesstudies are
that conducted
make use of forflux,different
speed, orapplications,
current signals.such as ba ery EV
[14,15]. Theseisstudies
The paper organized show that, for
as follows: EV applications,
Section the motor
2 explains possible controller
fault modes (including sen
in different
sors) has a higher number of chances of failing when compared to the motor itself.
parts of a typical motor drive, and briefly explains their manifestation. Sections 3–6 Withi
the power electronics unit, gate drivers and IGBTs have the highest chance of failur
explain the FDD methods for machines, power electronics, sensors, and DC link capacitors,
respectively. Section 7 concludes the paper by comparing FDD approaches, pointing out
Among the components of the motor, bearings, rotor bars, and stator windings exhibit th
the challenges and providing insights into potential research areas.
highest failure rates and cover most of the fault space. Speed transducer failures are als
2.significant.
Fault SpaceThese results match with the EPRI study and IEEE study on motor-failur
2.1. Drive[16].
causes These studies show that bearings, broken rotor bars, and stator winding shor
Topology
circuit failures
A motor drivecorrespond
system consiststo moreof a than
motor,75%
powerof the faults. converters, sensors, and
electronic
To keep this review focused, the most
a controller unit, as shown in Figure 2. Apart from the main common failures
DC–AC willinverter,
be considered.
based onFor moto
faults,
the bearing-,
energy source, stator-
the drive andmay rotor-related
contain DC–DC faults will be discussed.
converters For power
if it is fed from a battery electronic
orfaults,
AC–DC power semiconductor device failures will be discussed. Sensor faults will be con
rectifiers if supplied by an AC source. The industry standard for the DC–AC
siderediswith
inverter acceptedemphasis on speed sensors
to be three-phase two-leveland currentIf sensors.
inverters. Lastly,specifically,
not mentioned DC link capacito
such as multilevel inverters, this topology is considered throughout the article. Multiphase
failures will be discussed.
drives beyond three-phase are out of the scope of our paper.

Figure
Figure 2. 2. General
General architecture
architecture of a motor
of a motor drive drive system.
system.

Any damage or failure to these components poses the risk of a complete system
breakdown. Some parts of the drive are more susceptible to failure than others. Relia-
bility analyses for motor drives are conducted for different applications, such as battery
EVs [14,15]. These studies show that, for EV applications, the motor controller (including
sensors) has a higher number of chances of failing when compared to the motor itself.
Within the power electronics unit, gate drivers and IGBTs have the highest chance of failure.
Among the components of the motor, bearings, rotor bars, and stator windings exhibit
the highest failure rates and cover most of the fault space. Speed transducer failures are
also significant. These results match with the EPRI study and IEEE study on motor-failure
causes [16]. These studies show that bearings, broken rotor bars, and stator winding
short-circuit failures correspond to more than 75% of the faults.
To keep this review focused, the most common failures will be considered. For motor
faults, bearing-, stator- and rotor-related faults will be discussed. For power electronics
faults, power semiconductor device failures will be discussed. Sensor faults will be consid-
ered with emphasis on speed sensors and current sensors. Lastly, DC link capacitor failures
will be discussed.
Energies 2023, 16, 5602 4 of 22

2.2. Machine Faults

2.2.1. Bearing Faults
Bearings are used to support the rotor shaft and provide consistent motion. Due
to reasons like incorrect installation, overloading, overheating, lack of lubrication, etc.,
bearings fail or degrade [16]. Bearing faults are the most common machine faults and they
correspond to almost 30% of the failures [17]. Bearings degrade before they are considered
faulty; thus, the fault precursors progress and deviate slowly. Also, they are not usually
abrupt compared to power electronics short-circuit or open-circuit faults. Thus, the window
for bearing fault detection is longer. Also, there is a spectrum of positions from fully healthy
to defective, so a partially faulty state is possible.
Bearings have inner and outer races with a series of balls or rolling elements in between
these races. Each of these elements can fail and produce different fault results [18]. To
keep the generality of this study, all the different subcategories will be considered as
bearing faults.
When a surface defect occurs on a bearing element, it causes periodic impact forces
that can be detected through the vibration signal during operation. By analyzing the
frequency components of the machine vibration, faults in the bearing components can be
identified and phase relationships between vibration measurements at different locations
on the machine are commonly used for fault detection [19]. A defective bearing creates
periodic vibrations on the shaft and exhibits itself in the stator current spectrum due to air
gap periodic fluctuation.

2.2.2. Stator- and Rotor-Related Faults

The stator and rotor are the main parts of a machine. The stator has windings and
insulation material between turns and slots. Whereas the rotor can be wound, a squirrel
cage, or can contain permanent magnets. A properly installed motor should have a balanced
rotor with a constant air gap between the stator and the rotor.
The most common failure for the stator is interturn short circuits or interturn failures.
These failures evolve into phase-to-phase or phase-to-ground failures [20]. The reasons for
these short circuits are mechanical and thermal stress on insulation materials leading to
their eventual breakdown [21]. They create a sideband frequency on phase currents [22].
In squirrel-cage rotors, broken rotor bars and cracked end rings are common rotor
faults [23]. Due to thermal stresses during turn-on transients or mechanical vibrations
stemming from load changes, one or more rotor bars can be broken. Broken bars lead to
imbalanced rotor currents which create uneven heating and accelerate aging [12].
Another set of fault types is eccentricity faults which change the generation of flux
within the machine due to varying airgap. There are three main types of eccentricity faults:
static, dynamic, and mixed, as illustrated in Figure 3. In the case of static eccentricity, the
center of the stator and the center of the rotor are not perfectly aligned. As a result, the rotor
rotates at a different location compared to the center of the stator, although the rotational
center remains the same as the rotor’s center. On the other hand, dynamic eccentricity
occurs when the center of rotation aligns with the center of the stator but the center of the
rotor is not aligned with the other two centers. Mixed eccentricity exhibits characteristics
of both static and dynamic eccentricities, with all three centers (stator, rotor, and rotation)
being misaligned. The misalignment of the rotating center modifies the generated flux. In
effect, this flux will create sideband frequencies in the stator currents [8]. The generated
flux pattern differs in each type of eccentricity fault.
16, x FOR Energies
PEER REVIEW
2023, 16, 5602 5 of 22 5 of 22

Figure 3. Static (a),Figure

dynamic (b), and
3. Static mixed (c)
(a), dynamic (b),eccentricities,
and mixed (c) black, orange,black,
eccentricities, and orange,
red dotsand
represent
red dots represent
stator, rotor, and rotational centers
stator, rotor, respectively.
and rotational centersOuter black and
respectively. inner
Outer blackorange circles
and inner represent
orange sta-
circles represent stator
tor and rotor, respectively.
and rotor, respectively.

2.3. Power Electronics Faults

Demagnetization failure is another common fault specific to permanent magnet motors.
It occurs due to high loading conditions or temperature stress, depending on the magnet
In motor drive systems, power electronics inverters act as the actuators for the motor.
type. The inverse magnetic field generated by the stator can demagnetize the core. Core
They generate the required voltage
demagnetization can and current and,
be complete to achieve the required
hence, affects the whole torque or can
core, or speed
be partial. A
references. Although inverter topologies
demagnetized may
core will cause vary by application,
harmonics, fault models
noise, and vibration [24]. and their
manifestations are similar. Failed power devices within inverters can act as open circuits
where there is no2.3.
gate response
Power andFaults
Electronics the switch is open or as short circuits where they act
as a low-impedance path. These
In motor two
drive faults are
systems, similar
power for allinverters
electronics types ofact switches, such asfor
as the actuators Si,the motor.
SiC, or GaN-based MOSFETs or IGBTs.
They generate the required voltage and current to achieve the required torque or speed
references. Although inverter topologies may vary by application,
Although the manifestation of faults is the same, their mechanisms vary with the ma- fault models and their
manifestations are similar. Failed power devices within inverters can act as open circuits
terial. Si and SiC have similar fault mechanisms; however, fault mechanisms for GaN de-
where there is no gate response and the switch is open or as short circuits where they act as
vices are different. Fault mechanisms
a low-impedance path. can betwo
These divided
faultsinto two parts:
are similar for allintrinsic and extrinsic
types of switches, such as Si, SiC,
faults. Extrinsic faults are tied to the packaging
or GaN-based MOSFETs or IGBTs. of the devices. They happen due to differ-
ent thermal reactionsAlthough
within the thelayered structure
manifestation of the
of faults package.
is the Thesemechanisms
same, their include bond-vary with the
wire fatigue, restructure
material. of metallization,
Si and and solder-joint
SiC have similar fatigue.
fault mechanisms; Intrinsic
however, failures
fault hap- for GaN
mechanisms
pen within the semiconductor die andFault
devices are different. are related to thecan
mechanisms physical properties
be divided into twoof the device
parts: intrinsic and
[25,26]. Dielectric breakdown and hot carrier injection are examples of such failure mech-
extrinsic faults. Extrinsic faults are tied to the packaging of the devices. They happen due
anisms. The extrinsic failure mechanisms for GaN are reported to be similar to the Si coun- include
to different thermal reactions within the layered structure of the package. These
bond-wire fatigue, restructure of metallization, and solder-joint fatigue. Intrinsic failures
terparts but intrinsic failure mechanisms, such as delamination, are different [27–29].
happen within the semiconductor die and are related to the physical properties of the
device [25,26]. Dielectric breakdown and hot carrier injection are examples of such failure
2.4. Sensor Faults mechanisms. The extrinsic failure mechanisms for GaN are reported to be similar to the Si
In motor drive systems, but
counterparts sensor feedback
intrinsic failureismechanisms,
essential forsuch
control, estimation,are
as delamination, and con- [27–29].
different
dition monitoring. There are three main sensors in a motor drive system, namely voltage,
2.4. Sensor Faults
current, and speed sensors, except in speed-sensorless drives. All these sensors can fail at
In motor drive systems, sensor feedback is essential for control, estimation, and
any point and with different failure mechanisms. These include temperature drift, in-
condition monitoring. There are three main sensors in a motor drive system, namely
creased noise, increased variance, bias, or zero output. Sensors are made from different
voltage, current, and speed sensors, except in speed-sensorless drives. All these sensors can
materials and with
faildifferent properties,
at any point and withe.g., current
different transformers
failure mechanisms.to These
opticalinclude
sensors. Each
temperature drift,
type of sensor has a different fault mechanism.
increased noise, increased variance, bias, or zero output. Sensors are made from different
materials and with different properties, e.g., current transformers to optical sensors. Each
2.5. DC Link Capacitor Fault
type of sensor has a different fault mechanism.
For motor drives to maintain stability during transients, the power electronic inverter
is interfaced with a DC link capacitor which responds to instantaneous high current de-
mands. Due to their electrochemical nature, capacitors are likely to age. According to in-
dustrial surveys for EVs and motor drive applications, capacitors have the highest suscep-
tibility to failures, being second to semiconductors [30].
Capacitor aging is reflected as reduced capacity and increased equivalent series re-
sistance (ESR). Subject to high temperatures, power cycling, and other environmental con-
Energies 2023, 16, 5602 6 of 22

2.5. DC Link Capacitor Fault

For motor drives to maintain stability during transients, the power electronic inverter is
interfaced with a DC link capacitor which responds to instantaneous high current demands.
Due to their electrochemical nature, capacitors are likely to age. According to industrial
surveys for EVs and motor drive applications, capacitors have the highest susceptibility to
failures, being second to semiconductors [30].
Capacitor aging is reflected as reduced capacity and increased equivalent series re-
sistance (ESR). Subject to high temperatures, power cycling, and other environmental
conditions, capacitors start to degrade and eventually fail [31].

3. Fault Detection and Diagnosis in Electric Machines

In motor drive systems, various signals can be captured and employed for fault
detection and diagnosis (FDD). Common signals obtained from the motor side include
phase currents, terminal voltages, DC link voltage, vibration, temperature, speed, torque,
and flux around the machine. Additionally, signals within the controller unit, such as
reference voltage or currents, gating patterns, etc., can be employed.
For FDD, different studies and applications make use of some of these signals in
various ways. The utilization and processing of these signals vary between studies and
applications. These quantities may be used directly or processed before being utilized.
Various fault-detection techniques can be employed to address the faults discussed
in Section 2. These techniques can be further categorized based on the data handling
methodology employed. Despite the use of similar signals or measurements, distinct
approaches are employed in detecting failures in each study. The techniques used can be
classified into three main categories:
1. Statistical Methods;
2. Machine-Learning (ML)-Based Methods;
3. Deep-Learning (DL)-Based Methods.
It is worth noting that the categorization of the methods is not always straightforward;
some methods may fall into multiple categories while others may not fit into any of these
categories. For example, in [17], the authors used wavelet transform on vibration signals to
feed this information to a convolutional neural network (CNN) which falls into categories
1 and 3. In [32], graph theory is used and merged with k-nearest neighbor methods (kNN),
which makes it hard to categorize. The categorization process is shown in Figure 1.
These methods can be generalized, as shown in Figure 4. Data for these three categories
can be collected from experiments, simulations, or data repositories. These data include
healthy and various faulty conditions. Later, based on the adopted method, fault detection
is performed.
Detection can also be done in various ways. Three of the common ones are illustrated
in Figure 5. Model-based or regression-based methods use certain thresholds for fault
detection. All statistical methods and some ML methods also use this approach. The second
detection method is using classification, which is used by ML and some DL methods. In
detection by classification, each new sample is classified as healthy or as one type of fault.
The third detection method is detection by NNs, which is used by some ML and DL
methods. NNs are trained to give direct health information so no additional detection block
is needed.
 categories 1 and 3. In [32], graph theory is used and merged with k-nearest neighbor meth-
ods (kNN), which makes it hard to categorize. The categorization process is shown in
Figure 1.
These methods can be generalized, as shown in Figure 4. Data for these three catego-
Energies 2023, 16, 5602 ries can be collected from experiments, simulations, or data repositories. These data in-
7 of 22
clude healthy and various faulty conditions. Later, based on the adopted method, fault
detection is performed.

Energies 2023, 16, x FOR PEER REVIEW 7 of 22

Detection can also be done in various ways. Three of the common ones are illustrated
in Figure 5. Model-based or regression-based methods use certain thresholds for fault de-
tection. All statistical methods and some ML methods also use this approach. The second
detection method is using classification, which is used by ML and some DL methods. In
detection by classification, each new sample is classified as healthy or as one type of fault.
The third detection method is detection by NNs, which is used by some ML and DL meth-
ods. NNs are trained to give direct health information so no additional detection block is
Figure 4. Block diagrams representing general operating principles of FDD methods.
needed.
Figure 4. Block diagrams representing general operating principles of FDD methods.

Figure 5.
Figure 5. Detection
Detection methods,
methods, (a)
(a) Threshold
Threshold based,
based, (b)
(b) Classification
Classification based,
based,(c)
(c)NN
NNbased.
based.

3.1. Statistical Approaches

3.1. Statistical Approaches
Statistical methods employ signal-processing techniques to analyze data in time,
Statistical methods employ signal-processing techniques to analyze data in time, fre-
frequency, or time-frequency domains. In the time domain, RMS and residual analyses
quency, or time-frequency domains. In the time domain, RMS and residual analyses are
are often employed. In the frequency domain, the FFT and wavelet transform, or other
often employed. In the frequency domain, the FFT and wavelet transform, or other statis-
statistical approaches like skewness, kurtosis, and special distributions are employed. Many
tical approaches like skewness, kurtosis, and special distributions are employed. Many of
of these utilize direct measurements, so results are easy to interpret and easy to implement
these utilize direct measurements, so results are easy to interpret and easy to implement
for online cases. However, recent studies revolve around ML approaches and the trend is
for online cases. However, recent studies revolve around ML approaches and the trend is
shifting from statistical methods to ML methods as more and more data becomes available.
shifting from statistical methods to ML methods as more and more data becomes availa-
ble. Bearing Faults
3.1.1.
To detect bearing faults, the most commonly used signal is vibration [11,17,18,32–39].
3.1.1. Bearing Faults
Researchers also considered using speed measurement [40], current measurements [41],
To detect bearing
flux measurement [35],faults, the most
and thermal commonly
images of theused signal
motor [42].isBearing
vibration [11,17,18,32–39].
faults are common
Researchers also considered using speed measurement [40], current measurements
for all types of motors including induction motors (IM), permanent magnet synchronous [41],
flux measurement
motors (PMSM), and[35], and thermal
brushless images(BLDC).
DC motors of the motor [42]. Bearing faults are common
for all types of motors including induction motors (IM), permanent magnet synchronous
motors (PMSM), and brushless DC motors (BLDC).
A model-based detection method is proposed in [34] using probabilistic techniques.
It is an online detection method and is suitable for time, frequency, or envelope analysis.
The authors propose to use a particle filter to create the feature vector and estimate a prob-
Energies 2023, 16, 5602 8 of 22

A model-based detection method is proposed in [34] using probabilistic techniques. It

is an online detection method and is suitable for time, frequency, or envelope analysis. The
authors propose to use a particle filter to create the feature vector and estimate a probability
mass function (PMF). If the PMF exceeds the confidence value, the fault is detected. The
system is learned from the data using recursive least squares with a forgetting factor. The
method allows for identification and prognosis but it is not experimentally validated.
In [43], the authors used mean values and standard deviations to detect faulty operation.
Researchers in [40] used a time-frequency method called variational mode decomposition
to detect failures. They utilized angle transformations to enhance detection capabilities.
In [11], authors used entropies and kurtosis of vibration signals to detect failures. Later, they
employed SVM to classify the fault. They incorporated experimental data and an online
dataset to detect failure with varying operating conditions. Fast spectral correlation and
enhanced envelope spectrums methods are investigated in [37], where the cyclic property
of the bearing fault is exploited.

3.1.2. Stator and Rotor Faults

Stator interturn faults, broken motor bars, and eccentricity faults have unique signa-
tures on the current and torque spectra [44,45].
The stator- and rotor-related faults are detected mostly using phase currents [46–56].
Utilizing the motor flux is also another approach [57–62]. Some studies used motor
torque [63] and acoustic signals [64,65] as fault identifiers. The advantage of using flux
signals is to detect failures in their incipient stages. Detecting early-stage failures can be
difficult with current or vibration signals since there are multiple factors affecting them.
Though their stator designs are similar, the driving circuitry for IM and PMSMs are
different. This creates a need for different detection schemes. Also, IM and PMSMs have
different rotor structures and different types of rotor failures.
In [55], the authors proposed the motor current signature analysis (MCSA) method
to detect stator inter-turn faults on IM drives using current signals. MCSA is a method
that utilizes certain sideband frequencies on the line current of motors [23]. By analyzing
the harmonic content of the current signal, certain faults can be detected. Reference [53].
proposes to use the square value of line currents and use multiple signal classification
techniques. The squared current amplifies the fault indicators to better detect broken rotor
bar faults. Apart from current signals, the authors of [60] proposed placing flux sensors to
detect interturn faults in their incipient stages which is not possible to detect using classical
methods. By placing three flux sensors, stray flux can be monitored to detect interturn
failures in IMs utilizing harmonic analysis. Similarly, the authors in [57] used flux sensors
to detect broken rotor-bar failures using a flux spectrum where MCSA is not applicable.
In [58], flux sensors are placed inside the airgap to capture real-time flux data which enabled
fast detection of eccentricity and rotor failures. PMSM and BLDC motor stator failures
can be detected with their current and flux signals. The authors of [47] investigated the
usability of current signals to detect failures on a six-step voltage source-driven BLDC.
They discovered that the third harmonic is the best feature to detect stator failures. In [61],
the third harmonic of the flux signal is also utilized to detect interturn short-circuit faults as
well as their locations. In [48], vibration signals are used on top of the current signals. They
utilized a spectral analysis method called Fast Kurtogram on the vibration signals which
proved to be very accurate for severe faults. In the incipient stage, they employed MCSA
and looked at the third harmonics of the current. Ref. [59] uses fluxes to detect both rotor
eccentricity and demagnetization faults, which cause flux asymmetry. Residual analysis is
done to differentiate demagnetization from eccentricity.
Partial demagnetization is taken into consideration in [56]. Certain harmonic orders
are extracted and an envelope is created to detect faults. A faster and more precise detection
of the demagnetization fault is possible by installing fluxgate sensors, as suggested in [62],
which gives the ability to detect faults in incipient stages.
Energies 2023, 16, 5602 9 of 22

3.2. ML-Based Methods

ML is a broad category that includes logistic regression, support vector machines
(SVM), decision trees, and neural NN. Fuzzy methods can also be considered in the ML
methods. ML methods require feature engineering for high-performance operation or
satisfactory accuracies. They do require training and testing which needs prior data
to operate.

3.2.1. Bearing Faults

The authors in [18] used statistical time features of vibration signals as well as com-
pressed features to make them interpretable. A hierarchical NN is then trained and used
for classification. Detection is done by classification where the classifier chooses between
healthy and multiple faulty states. The researchers in [33] utilized autocorrelation infor-
mation and extracted features from autocorrelations rather than the raw data itself. Then,
a random-forest classifier is used to classify the healthy state from different faulty states.
The authors in [39] used an NN-based filter for vibration signals and then fed the features
to a second NN with a multilayer perceptron (MLP) structure. It is shown that, with the
addition of the filter, the detection and classification accuracy increased dramatically. Apart
from using NNs, ensemble classifiers such as random forest and XGBoost can also be used.
In [66], phase-current signals are processed using discrete wavelet transform and different
ensemble classifiers are tested. This method achieved 99% accuracy for fault detection
with XGBoost. Thermal images were also used to detect bearing faults [42] and have the
advantage of being nonintrusive and noncontact. Collected images are used as 2D signals
and 2D wavelet transform is applied to images to generate features. The dimensionality
reduction technique, namely principal-component analysis, is applied to the feature set to
pick the best features. The authors reached 100% accuracy in detection and classification
using the SVM classifier.

Table 1. Summary of Machine FDD using Statistical Methods.

Method Analysis Type MT 1 Fault Type Used Signals Fault Indicator Svrty? 7 Ref.
Deviation from healthy
Improved Dynamic System
St 2 - Bearing Vibration operation based on No [35]
Model with Particle Filtering
healthy model
Complementary Ensemble
Refined Composite
Empirical Mode
St 2 IM Bearing Vibration Multiscale RMS Yes [36]
Decomposition, Weighted
(RCRMS), Kurtosis
Multiscale Entropy
Sparse Code Shrinkage
Denoising, Fast Frq 3 - Bearing Vibration Ball pass frequency No [38]
Spectral Correlation
Angle Domain Conversion, Recurring frequency
Vibration,
Variational Mode Frq 3 - Bearing components No [40]
Speed
Decomposition in angle domain
Mean value difference in
Basic Statistics St 2 IM Bearing Stray Flux Yes [43]
stary flux measurements
The third harmonic in
Fourier Analysis Frq 3 BLDC Stator IT 4 Current Yes [47]
negative frequency
Amplitude of certain
Current Harmonic Analysis Frq 3 PMSM DeMgt 5 Current No [48]
harmonic orders
Motor Current Vibration, The third harmonic in
Frq 3 BLDC Stator IT 4 Yes [49]
Signature Analysis Current current spectrum
Double multiples of slip
MUSIC Frq 3 IM BRB Current frequency in Yes [54]
Fourier analysis
Energies 2023, 16, 5602 10 of 22

Table 1. Cont.

Method Analysis Type MT 1 Fault Type Used Signals Fault Indicator Svrty? 7 Ref.
Multiple Reference The third harmonic in
Frq 3 IM Stator IT 4 Current Yes [56]
Frames, MCSA current spectrum
Increased amplitude on
Fourier Analysis Frq 3 IM BRB Stray Flux Yes [57]
flux spectrum at
Sensor Measurement Difference in airgap flux
Time IM BRB, Ecc 6 Internal Flux Yes [58]
Difference densities in similar poles
Sensor Measurement Ecc 6 , Normalized changes in
Time PMSM Flux Yes [59]
Difference DeMgt 5 flux measurements
Amplitude increments of
Flux Vector Analysis Frq 3 IM Stator IT 4 Flux frequency in Flux Yes [60]
Vector FFT
Stray Flux Analysis Frq 3 PMSM Stator IT 4 Flux 3rd harmonic of stray flux Yes [61]
Amplitude of certain
Flux Spectrum Analysis Frq 3 PMSM DeMgt 5 Leakage Flux Yes [62]
harmonic orders
1 MT: Machine Type, 2 St: Statistical, 3 Frq: Frequency, 4 IT: Interturn, 5 DeMgt: Demagnetization, 6 Ecc: Eccentricity,
7 Svrty: Severity.

3.2.2. Stator and Rotor Faults

The authors in [49] used line currents to detect interturn failures in VSI-fed IMs. They
extracted features with wavelet transform and later used an SVM-based algorithm. They
successfully detected faults with varying switching frequencies and achieved more than
99% accuracy. Researchers in [52] used a similar approach to use line currents to both
detect the stator failure and also decide the severity of the fault using two MLP models for
detection and severity assignment. They also utilized a multiagent system and multiple
systems to achieve generality. The detection accuracies are up to 100% for different fault
cases. Utilization of torque signal to detect stator faults while deciding severity is shown
in [63]. Startup torques with different severity levels are fed into an NN with one hidden
layer. Accuracies ranging from 88% up to 99.9% are achieved. In [54], a broken rotor bar
fault-detection scheme is proposed using ANNs by utilizing stator currents.
In [50], the authors proposed to use frequency domain features of line currents to
detect stator failures on PMSMs using the kNN approach where classes are the number
of shorted turns. Researchers in [51] showed that broken rotor-bar faults can be detected
using current signals of PMSMs. Time-domain signals are used and statistical features are
extracted. The random-forest classifier resulted in more than 98% accuracy in deciding
healthy or faulty conditions. Discrete wavelet transform on the stator current is used to
detect broken magnet faults and eccentricity faults in [67]. An adaptive filter is developed
to remove the fundamental component of the stator current. Later, an SVM classifier is
employed to classify healthy and faulty conditions.

Table 2. Summary of Machine FDD using ML Methods.

Method MT 1 FT 2 Used Signals FE 3 FS 4 Classifier Acc 5 Svrty? 6 Ref.

STD 7,
H-MLP 14 IM Bearing Vibration Discriminant Analysis H-MLP 95% Yes [18]
CCA 8
Removing Nonbearing
NN filter 96–
IM Bearing Vibration TD 9 Fault Component MLP Yes [33]
MLP 100%
(RNFC) filter
Recursive Feature 96.7–
RF IM Bearing Vibration St 10 RF 15 Yes [34]
Elimination (RFE) 100%
PCA, Mahalanobis
SVM IM Bearing Thermography DWT 11 , St SVM 16 100% Yes [42]
distance (MD)
Energies 2023, 16, 5602 11 of 22

Table 2. Cont.

Method MT 1 FT 2 Used Signals FE 3 FS 4 Classifier Acc 5 Svrty? 6 Ref.

Stator
SVM IM SIT 18 DWT, St - SVM 99.74% Yes [50]
Current
Stator
kNN PMSM SIT FFT - kNN 99.10% Yes [51]
Current
Stator Feature
RF PMSM BRB STD RF 98.40% No [52]
Current Importance Based
Sampled
Stator
MLP IM SIT Measure- - MLP 100% Yes [53]
Current
ment
88–
NN IM SIT Torque St, FD 12 - NN Yes [63]
100%
Stator DWT, TFD
XGBoost PMSM Bearing 13 - XGBoost 17 99.30% Yes [66]
Current
Ecc 19 ,
Stator
SVM PMSM Broken FD LDA 20 SVM 96%+ Yes [67]
Current
Magnet
1 MT: Machine Type, 2 FT: Fault Type, 3 FE: Feature Extraction, 4 FS: Feature Selection, 5 Acc: Accuracy, 6 Svrty:
Severity, 7 STD: Statistical Time Domain, 8 CCA: Curvilinear Component Analysis, 9 TD: Time Domain, 10 St:
Statistical, 11 DWT: Discrete Wavelet Transform 12 FD: Frequency Domain, 13 TFD: Time-Frequency Domain,
14 : H-MLP: Hierarchical Multilayer Perceptron, 15 RF: Random Forest, 16 kNN: k-Nearest Neighbor, 17 XGBoost:

Extreme Gradient Boosting, 18 SIT: Stator Interturn, 19 Ecc: Eccentricity, 20 LDA: Linear Discriminant Analysis.

3.3. DL-Based Methods

DL methods are a subset of ML methods. The main difference is that there is no feature
engineering in DL approaches and they require much more data compared to ML methods.
Common DL methods can be listed as autoencoders, convolutional neural networks (CNN),
generative adversarial networks (GAN), recurrent neural networks, and reinforcement
learning. A summary of DL methods applied for machine FDD is shown in Table 3.

3.3.1. Bearing Faults

DL-based approaches are extensively studied for bearing faults [6]. Approaches are
taken to utilize autoencoder structures to denoise and detect failures [39]. Autoencoders
first reduce the dimensionality of the signal and attempt to regenerate the original signal.
Another approach is to convert physical signals to images and use strong image process-
ing tools such as CNNs [17,68]. CNNs are a form of multilayer perceptron with more
constraints. Also, they include convolution and pooling layers. The interpretability and
performance of CNNs can also be improved by imposing physical constraints on the model;
hence, physics-informed structures can be formed [37]. A different DL approach used
in bearing fault detection is GAN [69]. With GAN, two parallel networks are trained to
surpass the other network, hence the name adversarial. The purpose of utilizing GANs
in [69] is to generate synthetic data for under-represented classes to increase generality. A
promising and emerging field in fault detection using DL is deep-transfer learning and
domain adaptation [11].
As mentioned earlier, DL requires a large amount of data which is hard to collect
for condition-monitoring applications. Using DL to find domain-invariant features and
transferring the model from one system to the other is a proposed solution for the lack
of data [70].

3.3.2. Stator and Rotor Faults

For stator and rotor faults, the literature is not as rich as bearing faults due to a lack of
open-source datasets. Yet, researchers have been looking into the utilization of this useful
tool to achieve better results. In [71], the detection and classification of one healthy and
five faulty states are investigated, including bearing faults, shorted stator windings, broken
Energies 2023, 16, 5602 12 of 22

rotor bars, bent rotors, and unbalanced rotation. Authors utilized autoencoders and used
generalization techniques, such as denoising autoencoder and dropout. The study shows
more than 97% accuracy across the classes. Researchers in [46] focused on different levels
of stator faults and they used CNN to detect and decide the severity of the fault. The paper
shows accuracies for various layers and activation functions. The authors showed a single
convolution layer with rectified linear unit (ReLU) activation function performs best. To
increase network resistance, they added pooling layers and dropout layers which resulted
in more than 99% accuracy. The authors of [72] tried to tackle the interpretability problem
of DL methods by using a novel deep-SincNet structure. They used current signals to detect
bearing and broken rotor-bar faults with accuracies higher than 99.9%.

Table 3. Summary of Machine FDD using DL Methods.

Used
Architecture Data Source Fault Type Classifier Accuracy Severity? Ref.
Signals
CNN, Physics
CWRU Bearing Vibration Softmax 91.82–99.97% Yes [37]
Informed
Residual
Autoencoder XJTU-SY,
Bearing Vibration Autoregression 97.97% No [39]
with Memory NASA-IMS
Estimator
IM
CNN Stator Inter-turn Stator Current Softmax 99.30% Yes [46]
Experiment
CWRU,
CNN Bearing Vibration - 99.48–100% Multiple Types [68]
Experiments
GAN, Auto
CWRU Bearing Vibration Auto Encoder 99.20% Multiple Types [69]
Encoder
Sparse Auto IM BRB, Bearing,
Vibration DNN 93.5–100% Multiple Faults [71]
Encoder Experiment Stator Winding,
IM Multiple
Deep-SincNet Bearing, BRB Stator Current Softmax 99.93% [72]
Experiment Severities/Faults

3.4. Machine Faults Data Repositories

Table 4 shows a list of four publicly available datasets for electric machine faults. These
datasets are useful for researchers to test and evaluate new FDD algorithms.

Table 4. Open-Source Data Repositories for Bearing Faults.

Dataset Source Acronym Ref.

Case Western Reserve University CWRU [73]
Xi’an Jiaotong University and Changxing Sumyoung
XJTU-SY [74]
Technology Co., Ltd., Changxing, China
NSF I/UCR Center for Intelligent Maintenance Systems NASA-IMS [75]
NASA–FEMTO PRONOSTIA [76]

4. Fault Detection and Diagnosis in Power Electronics

Power electronics faults should be detected quickly to avoid further damage to the
system, especially given that they are abrupt in nature. Therefore, fault-detection schemes
differ from those of machine faults.
The dominant circuit topology for motor drives is accepted to be three-phase
two-level inverters. On the other hand, modular multilevel inverters are gaining popu-
larity; thus, fault-detection methods that utilize multilevel inverters are included here. Also,
device-level detection methods are explored since they are circuit topology independent.
Power electronics faults are classified as open- and short-circuits based on the power
semiconductor device behavior. In this sense, methods that utilize circuit behavior can be
 Energies 2023, 16, 5602 13 of 22

used for drives containing IGBTs, MOSFETs, or other wide band-gap devices. However,
methods exploiting device properties to detect failures cannot be transferred to other types
of devices [77–79]. This is an important point as WBG devices are penetrating the market,
thus increasing the diversity of power electronic devices. Similarly, some methods are
topology- or modulation-dependent. Please refer to Table 5 for a summary of PE-related
FDD methods.
Approaches to detect PE faults can be classified into three main categories:
1. Logic-based methods;
2. Residual-based methods;
3. Controller-aided methods.
4.1. Logic-Based Methods
The first approach is to use sensor measurements directly with a threshold or con-
gies 2023, 16, x FOR PEER REVIEW structing logic functions directly from measurements [77–86] which 13
weofcall
22 logic-based
methods. Logic-based methods investigate the relationship between measurements rather
than building a model as shown in Figure 6.

Figure 6. Block diagram

Figure of logic-based
6. Block diagram ofFDD for PE faults.
logic-based FDD for PE faults.

These methods Theserequire knowledge

methods of circuit operation
require knowledge in healthyinmode.
of circuit operation healthyIn mode.
[77,78],In [77,78], the
the authors exploited the difference in gate capacitances of SiC MOSFET and IGBTIGBT
authors exploited the difference in gate capacitances of SiC MOSFET and for for healthy
healthy and faulty conditions. Detection of OC and SC faults is possible using only the gate voltage.
and faulty conditions. Detection of OC and SC faults is possible using only the
gate voltage.AAsimilar
similarapproach
approachisistaken
takenfor
forGaN
GaNHEMTHEMTdevices
devicesinin[79]
[79]where
wherespecific
specificand very fast
and very fastdetection
detection hardware
hardware is integrated
is integratedwith thethe
with gate driver.
gate AnyAny
driver. deviation fromfrom
deviation phase voltage is
detected and fault flags are generated. In [80], OC faults in an H-Bridge
phase voltage is detected and fault flags are generated. In [80], OC faults in an H-Bridge (HB) were detected
(HB) were detected and localized. The output voltage of the HB and load current are uti- detection
and localized. The output voltage of the HB and load current are utilized. The
methodmethod
lized. The detection can be scalable for multilevel
can be scalable and three-phase
for multilevel structures.
and three-phase For three-phase
structures. For PMSM
drives, the averaged current technique is used in [81]. By looking
three-phase PMSM drives, the averaged current technique is used in [81]. By looking into into the average phase
currents, it is possible to do detection without requiring any system parameters. In [82] OC
the average phase currents, it is possible to do detection without requiring any system
faults in three-phase inverters are investigated. Phase currents are studied under different
parameters. In [82] OC faults in three-phase inverters are investigated. Phase currents are
fault locations and a 3D trajectory mapping is generated. It allowed for the detection and
studied under different fault locations and a 3D trajectory mapping is generated. It al-
localization of OC faults for motor drives. In [84], the authors followed a similar mapping
lowed for the detection and localization of OC faults for motor drives. In [84], the authors
to detect multiple OC faults. For multilevel inverters, a modular approach is taken in [83]
followed a similar mapping to detect multiple OC faults. For multilevel inverters, a mod-
and detection is done by taking one current and one voltage measurement per H-bridge
ular approach is taken in [83] and detection is done by taking one current and one voltage
module. By modifying a regular three-phase inverter by adding an input filter, it is possible
measurement per H-bridge module. By modifying a regular three-phase inverter by add-
to detect OC and SC faults for a three-phase bridge circuit, as shown in [85], as the current
ing an input filter, it is possible to detect OC and SC faults for a three-phase bridge circuit,
trajectory is predictable. In [86], the authors used a single method to detect OC, current,
as shown in [85], as the current trajectory is predictable. In [86], the authors used a single
and speed-sensor faults in PMSM drives. They used three-phase current measurements
method to detect OC, current,
and rotor position.and speed-sensor
The sum of three faults in PMSM
current drives. They
measurements canused three-
isolate current sensor
phase current measurements
faults and rotor
from OC faults. position.
To increase The sum ofthe
robustness, three current
authors measurements
used adaptive thresholds and
can isolate current sensorfeatures.
normalized faults from OC faults. To increase robustness, the authors used
adaptive thresholds and normalized features.

4.2. Residual-Based Methods

The second approach is called residual-based methods and they require a model gen-
erated from data or dynamic equations. The source of data can be simulations or experi-
ments. Later, immediate measurements and model output are compared to calculate re-
siduals [87–93]; hence, it is called residual-based methods. The model can be an ML model,
a statistical model, or a dynamical model. Figure 7 shows the model-based approach in
 Energies 2023, 16, 5602 14 of 22

4.2. Residual-Based Methods

The second approach is called residual-based methods and they require a model
generated from data or dynamic equations. The source of data can be simulations or
experiments. Later, immediate measurements and model output are compared to calculate
residuals [87–93]; hence, it is called residual-based methods. The model can be an ML
model, a statistical model, or a dynamical model. Figure 7 shows the model-based approach
EER REVIEW 14 of 22
in FDD and how residuals can be utilized. A fair amount of papers that use ML can be
found in this category, ranging from Bayesian networks to CNNs. As ML methods have an
appearance, they are not applied for PE faults as widely as machine faults.

Figure 7. Block diagramFigure

of residual-based FDD schemes.
7. Block diagram of residual-based FDD schemes.

In [87], the authors used Bayesian networks to detect single and double OC faults in
For multilevel converters,
PMSM CNN
drives. is used
Gate signals andinline
[89] to avoid
currents thefrom
are used feature-extraction
both simulation and process.
experimen-
In the same study, for a four-cell multilevel converter, four voltages and two currents are
tal data to generate features. For single and double OC faults, 100% and 98.9% accuracies
are achieved. In [88], the fuzzy logic method is used with Park’s phase currents. Multiple
measured. With less than 100 ms
ML methods, detection
including SVMs, time,
kNNs, a and
99.7% average
multilayer fault-detection
perceptrons, are compared accuracy
in [90] for
is achieved. Since theEVmultilevel
inverters. converter topology is bidirectional, it is possible to adapt
the approach to motor drives. The usage of CNN allowed the detection of faults in varying
For multilevel converters, CNN is used in [89] to avoid the feature-extraction process.
In the same study, for a four-cell multilevel converter, four voltages and two currents are
load conditions. Using the fastWith
measured. Fourier
less thantransform algorithm,
100 ms detection the proposed
time, a 99.7% method in
average fault-detection [91]
accuracy
extracts fault frequency spectrum features, selects the most critical features through a fea-
is achieved. Since the multilevel converter topology is bidirectional, it is possible to adapt
ture selection method, and employs a random vector functional link network to learn the
the approach to motor drives. The usage of CNN allowed the detection of faults in varying
load conditions. Using the fast Fourier transform algorithm, the proposed method in [91]
faulty knowledge. The method
extracts fault has beenspectrum
frequency tested and demonstrated
features, selects the most high accuracy
critical and ro-a
features through
bustness in identifying faults,
feature evenmethod,
selection underand varying
employsconditions.
a random vector In functional
[93], a machine-learning
link network to learn
algorithm is used to model the stator current of a single switch using drain current, switch
the faulty knowledge. The method has been tested and demonstrated high accuracy and
robustness in identifying faults, even under varying conditions. In [93], a machine-learning
voltage, and temperature.
algorithmThe experimentally
is used to model the statorvalidated
current ofmethod shows
a single switch usingOC and
drain SC faults
current, switch
can be detected with very high accuracies. For three-phase inverters, the authors in faults
voltage, and temperature. The experimentally validated method shows OC and SC [92]
proposed to use residuals from the line and phase voltages to detect OC faults as well as
can be detected with very high accuracies. For three-phase inverters, the authors in [92]
proposed to use residuals from the line and phase voltages to detect OC faults as well as
current-sensor faults.current-sensor
To enrich the faults.detection,
To enrich thethey also incorporated
detection, the polarity
they also incorporated ofofthe
the polarity the
residuals as features. residuals
To calculate residuals,
as features. theresiduals,
To calculate authorsthe used current
authors measurements
used current measurements from
from
two current sensors instead
two currentof sensors
three to reduce
instead cost
of three and complexity.
to reduce cost and complexity.

4.3. Controller-Aided Methods

In motor drive applications, the inverter or motor has some form of closed-loop con-
trol, which can be a version of field-oriented control, direct torque control, or other control
techniques. Other control techniques might employ adaptive or predictive models. In this
Energies 2023, 16, 5602 15 of 22

4.3. Controller-Aided Methods

In motor drive applications, the inverter or motor has some form of closed-loop
control, which can be a version of field-oriented control, direct torque control, or other
control techniques. Other control techniques might employ adaptive or predictive models.
In this case, an already-built control model can also be part of FDD.
The third approach is to make use of adaptive or predictive controller architecture
Energies 2023, 16, x FOR PEER REVIEW 15 of 22 for
within the system for FDD [94–97]. In this approach, estimated states or parameters
control purposes are used for detection as well. This category utilizes estimators and
observers present in the controller, as shown in Figure 8.

Figure 8. Block diagram of controller-aided FDD for PE faults.

Figure 8. Block diagram of controller-aided FDD for PE faults.

Estimators take previous estimations as a reference to detect anomalies. These methods

Estimators take previous estimations as a reference to detect anomalies. These meth-
provide fault-detection capability without requiring additional sensor installations. In [94],
ods provide fault-detection capability without requiring additional sensor installations.
a model-based observer is used to estimate the phase currents, and the estimated variables
In [94], a model-based observer is used to estimate the phase currents, and the estimated
are compared with measurements for three-phase inverters. Similarly, the authors in [95]
variables are compared with measurements for three-phase inverters. Similarly, the au-
also use
thors in observers but
[95] also use they estimate
observers but theyspeed andspeed
estimate flux to detect
and OC
flux to failures.
detect For multilevel
OC failures. For
inverters, model-predictive control is widely used and is also utilized as a fault-detection
multilevel inverters, model-predictive control is widely used and is also utilized as a fault-
method
detectionasmethod
in [96,97].
as in [96,97].
5. Fault Detection and Diagnosis in Sensors
5. Fault Detection and Diagnosis in Sensors
Sensor-fault
Sensor-fault detection
detectionisiscritical
criticalin in
fault-tolerant drives
fault-tolerant andand
drives the the
continuation of healthy
continuation of
operation
healthy operation [98–100]. Unexpected sensor measurements confuse the controllerwell
[98–100]. Unexpected sensor measurements confuse the controller as as as
existing FDD schemes,
well as existing indicating
FDD schemes, a system
indicating fault fault
a system whereas it is it
whereas a sensor fault.
is a sensor Detecting
fault. De-
sensor faults and separating them from system faults is also critical.
tecting sensor faults and separating them from system faults is also critical.
In
Inthe
theliterature,
literature,combined
combinedsensor-failure
sensor-failuredetection
detectionschemes
schemeswithwithPE PEfaults areare
faults reported
re-
ported
for motor fordrive
motorsystems
drive systems [81,86,91,92],
[81,86,91,92], as well asas
well as standalone
standalone detection
detection schemes
schemes [98–
[98–103].
103].
Table 5. Summary of FDD Methods for Power Electronic Faults.
Table 5. Summary of FDD Methods for Power Electronic Faults.
Method Topology Modulation Fault Type Switch Type Used Signals Scalable Ref.
Switch
LogicMethod
Based Topology
Independent Modulation
Independent Fault Type
OC, SC IGBT UsedGate
Signals
voltage Scalable
Yes Ref.[77]
Type
Logic Based Independent Independent OC, SC SiC Gate voltage Yes
Logic Based Independent Independent OC, SC IGBT Gate voltage Yes [77][78]
Logic
Logic Based Independent
Based Independent Independent
Independent OC,
SC SC SiCHEMT
GaN GatePhase
voltage
voltage Yes Yes [78][79]
Logic
Logic Based
Based Independent
H-Bridge Independent
LS-SPWM SC
OC GaN HEMT
Any PhaseVoltage,
Output voltageload current Yes Yes [79][80]
Inverter
H-Bridge Output Voltage, load
Logic Based LS-SPWM OC Any Yes [80]
Logic Based Inverter
3-phase
SVM PWM
OC, Current
Any current
3-phase currents No [81]
Inverter Sensor
3-phase OC, Cur-
Logic Based SVM PWM Any 3-phase currents No [81]
Logic Based * Inverter
3-phase SPWM– rentOC
Sensor Any 3-phase currents No [82]
Inverter Expandable
3-phase SPWM–Ex-
Logic Based * OC Any 3-phase currents No [82]
Inverter pandable
Current and Voltage Per
Logic Based CHB MLI SPWM OC Any Yes [83]
Leg
3-phase
Logic Based SPWM OC Any 2 phase currents No [84]
Inverter
Modified 3-
Energies 2023, 16, 5602 16 of 22

Table 5. Cont.

Method Topology Modulation Fault Type Switch Type Used Signals Scalable Ref.
Logic Based CHB MLI SPWM OC Any Current and Voltage Per Leg Yes [83]
3-phase
Logic Based SPWM OC Any 2 phase currents No [84]
Inverter
Modified
Input current, gate signals,
Logic Based 3-phase SVM PWM OC SC Any No [85]
gate voltage
Inverter
OC, Speed,
3-phase
Logic Based * SVM PWM and Current Any 3-phase current, speed No [86]
Inverter
Sensor
3-phase
Residual Based SPWM OC Any 3-phase currents, Gate signals No [87]
Inverter
3-phase
Residual Based SVM PWM OC Any 3-phase currents No [88]
Inverter
Module voltage, load current,
Residual Based MLI NA OC Any Yes [89]
circulating current
3-phase
Residual Based NA OC Any 3-phase currents No [90]
Inverter
Async.-Sync
3-phase OC, Current
Residual Based Hybrid Any 3-phase currents No [91]
Inverter Sensor
Modulation
3-phase OC, Current 3-line voltage, 3-phase
Residual Based NA Any No [92]
Inverter Sensor voltage
Si, SiC Device voltage, device
Residual Based Independent Independent OC, SC Yes [93]
MOSFET current, case temperature
Controller 3-phase OC, Current
SVM PWM Any 3-phase currents No [94]
Aided Inverter Sensor
Controller 3-phase 3-phase voltage, 3-phase
SPWM OC Any No [95]
Aided Inverter current
Controller
MLI MPC Based OC Any Module voltage Yes [96]
Aided
Controller Module input voltage,
MLI NA OC Any Yes [97]
Aided module current
* These methods can also be considered in residual-based methods.

Combined methods detect and isolate sensor faults from PE faults. In [81], the symme-
try of three-phase currents in the drive is exploited. The authors showed that the phase
shift between healthy phases changes upon a sensor fault. Similarly, in [86], the summation
of phase currents is tracked to determine the sensor fault. A neural network approach is
taken in [91], which allowed the detection of voltage and current-sensor faults. FFT-based
features are used to detect and diagnose different fault modes of sensors. The authors
in [92] worked in systems where only two current sensors are employed and by utilizing
phase and line voltage differences.
As for standalone sensor FDD schemes, to detect current sensor faults, in [98], a
model-free approach is proposed where the asymmetry between phase-current measure-
ments is used for field-oriented controlled (FOC) drives. Again, for FOC drives [101],
researchers used delayed signals to detect current and speed-sensor failures. In [99], the
authors proposed multiple estimation schemes. Any discrepancy between estimated values
and measured values can be detected and, instead of measured, estimated values are used
for fault-tolerant operation. NN-based sensor fault-detection schemes are also proposed
in [100,103] and an extreme learning machine is used in [102] for the detection and classifi-
cation of the type of faults. But, similar to PE faults, due to the lack of available data and the
presence of simpler alternatives, ML literature is not as rich in sensor fault detection area.
Energies 2023, 16, 5602 17 of 22

6. DC Link Capacitor Fault Detection

To detect DC link capacitor failures, its condition should be monitored in real-time.
One method to do it is by injecting current or voltage during normal operation as proposed
in [104–106]. The injected current or voltage creates a fluctuation in the capacitor power.
Monitoring and analyzing this fluctuation allows for real-time ESR monitoring. The authors
in [104] injected pulse-type signals whereas, in [105,106], a low-frequency signal is injected.
The signal injection has the downside of disturbing normal operation and increasing THD.
For this reason, researchers investigated capacitor voltage, capacitor current, and ripple
voltage on the DC link capacitor to estimate ESR and capacitance. In [107,108], capacitor
current and voltage are used to calculate capacitance directly whereas, in [109], only ripple
voltage is used. Extracting ripple voltage requires additional instrumentation and might be
challenging if the DC voltage is high. Similarly, in [110], the voltage and current of the DC
capacitor are processed using a short-time Fourier transform. Later, ESR and capacity are
calculated using mathematical models.

7. Conclusions, Challenges, and Future Work

This paper provided an overview of the faults and relevant FDD methods with empha-
sis given to recent studies over the past five years. We observed that statistical methods are
used to detect all types of machine faults, including bearing, BRB, stator winding, demagne-
tization, etc. Mathematical analysis, simulations, and simple experiments provide enough
information to build these FDD systems. Knowing the fault mechanism and building an
FDD around this information still holds its value. Many practical papers are utilizing
these mechanisms. A challenge, however, is to use the same or available signals to detect
multiple faults. Detecting and diagnosing multiple faults leads to the usage of ML- and
DL-based methods.
We can also observe that ML methods are becoming more popular to detect all types of
faults but the availability of data is the key. Being able to simulate different operating con-
ditions and various faulty conditions empowers the ML methods. However, the trade-off
with the computation time and model complexity is still there to decide on which method
to use. Also, it might be misleading to check the success of an ML method simply from its
detection accuracy. Readers should be aware that the detection accuracy is decided based
on select operating conditions in a given system. When implemented in another system,
these algorithms might not produce the same results in terms of accuracy.
In contrast, deep-learning methods require much more data than ML methods and
their applications are mostly to detect bearing faults. This is, again, related to the data avail-
ability issue. We conclude that there is no clear advantage to using DL methods over ML or
statistical methods in terms of accuracy, the number of used signals, and computational
cost. The usage of DL methods for FDD on a simple system needs justification but there is
room for fault prognosis using DL methods. Also, DL might be useful when information
from multiple drives is put together to create a generalized method.
Researchers are exploring condition monitoring for machines and drives extensively. A
challenge in condition monitoring is to develop generalizable fault-detection schemes. Many
systems might appear to be similar but the difference in the actual products and unique operat-
ing conditions make generalization a challenging task. In this sense, time-frequency statistical
methods, in general, are tied to fault mechanisms which makes them more generalizable.
On the other hand, ML and DL methods rely heavily on data availability and quality;
when data is available, they perform exceptionally well. Open-source data repositories for
machine drive systems accelerate FDD using ML and DL. As more researchers share their
experimental data, the research community will grow and the generalization problem can
be tackled. Some open-source experimental data repositories are summarized in Table 5.
As ML and DL are used increasingly, the inherent problems of these methods become
apparent in the results. Biased or imbalanced data sets, not having data for unanticipated
faults, or multiple types of faults happening concurrently, are such problems. There are
Energies 2023, 16, 5602 18 of 22

also promising methods to overcome these obstacles. Transfer learning and adaptation
methods are emerging and showing impressive results [66].
On the power electronics side, device physics is exploited, as well as topology and
modulation, to detect faults. As new WBG devices became more common in applications,
FDD methods that consider those specific devices should be developed. SiC MOSFETs
are becoming more common but GaN HEMTs require more work on their fault detection.
Also, as transportation and defense industries are going towards more reliable solutions,
paralleling devices have become more popular. This aspect also needs more research.
The abrupt nature of power electronics faults pushed researchers to logic-based meth-
ods which are, in general, faster and application specific. Model-based approaches are also
fast and can be applied to different scenarios.
Usage of MLIs (cascaded H-bridge or NPC inverters) for drives requires revisiting
some established methods. Established methods rely on special properties of generated
harmonics in the current and voltage spectra. MLIs produce different current and voltage
spectra than those of classical inverters, which is worth investigating.

Author Contributions: Conceptualization, M.A.G. and A.B.; methodology, M.A.G. and A.B.; writing—
original draft preparation, M.A.G.; writing—review and editing, A.B.; visualization, M.A.G.; supervision,
A.B. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Alshorman, O.; Alshorman, A. A review of intelligent methods for condition monitoring and fault diagnosis of stator and rotor
faults of induction machines. Int. J. Electr. Comput. Eng. 2021, 11, 2820–2829. [CrossRef]
2. Khan, M.A.; Asad, B.; Kudelina, K.; Vaimann, T.; Kallaste, A. The Bearing Faults Detection Methods for Electrical Machines—The
State of the Art. Energies 2023, 16, 296. [CrossRef]
3. Li, D.; Wang, Y.; Wang, J.; Wang, C.; Duan, Y. Recent advances in sensor fault diagnosis: A review. Sens. Actuators A Phys. 2020,
309, 111990. [CrossRef]
4. Kumar, P.; Hati, A.S. Review on Machine Learning Algorithm Based Fault Detection in Induction Motors. Arch. Comput. Methods
Eng. 2021, 28, 1929–1940. [CrossRef]
5. Gonzalez-Jimenez, D.; Del-Olmo, J.; Poza, J.; Garramiola, F.; Sarasola, I. Machine learning-based fault detection and diagnosis of
faulty power connections of induction machines. Energies 2021, 14, 4886. [CrossRef]
6. Zhang, S.; Zhang, S.; Wang, B.; Habetler, T.G. Deep Learning Algorithms for Bearing Fault Diagnostics—A Comprehensive
Review. IEEE Access 2020, 8, 29857–29881. [CrossRef]
7. Liang, X.; Member, S.; Ali, M.Z.; Member, S. Induction Motors Fault Diagnosis Using Finite Element Method: A Review. IEEE
Trans. Ind. Appl. 2020, 56, 1205–1217. [CrossRef]
8. Liu, Y.; Bazzi, A.M. A review and comparison of fault detection and diagnosis methods for squirrel-cage induction motors: State
of the art. ISA Trans. 2017, 70, 400–409. [CrossRef]
9. Xu, X.; Qiao, X.; Zhang, N.; Feng, J.; Wang, X. Review of intelligent fault diagnosis for permanent magnet synchronous motors in
electric vehicles. Adv. Mech. Eng. 2020, 12, 1–14. [CrossRef]
10. He, J.; Yang, Q.; Wang, Z. On-line fault diagnosis and fault-tolerant operation of modular multilevel converters—A comprehensive
review. CES Trans. Electr. Mach. Syst. 2021, 4, 360–372. [CrossRef]
11. Hassan, O.E.; Amer, M.; Abdelsalam, A.K.; Williams, B.W. Induction motor broken rotor bar fault detection techniques based on
fault signature analysis—A review. IET Electr. Power Appl. 2018, 12, 895–907. [CrossRef]
12. Neupane, D.; Seok, J. Bearing fault detection and diagnosis using case western reserve university dataset with deep learning
approaches: A review. IEEE Access 2020, 8, 93155–93178. [CrossRef]
13. Riera-Guasp, M.; Antonino-Daviu, J.A.; Capolino, G.A. Advances in electrical machine, power electronic, and drive condition
monitoring and fault detection: State of the art. IEEE Trans. Ind. Electron. 2015, 62, 1746–1759. [CrossRef]
14. Tang, Q.; Shu, X.; Zhu, G.; Wang, J.; Yang, H. Reliability study of bev powertrain system and its components—A case study.
Processes 2021, 9, 762. [CrossRef]
15. Shu, X.; Guo, Y.; Yang, W.; Wei, K.; Zhu, Y.; Zou, H. A Detailed Reliability Study of the Motor System in Pure Electric Vans by the
Approach of Fault Tree Analysis. IEEE Access 2020, 8, 5295–5307. [CrossRef]
16. Singh, G.K.; Al Kazzaz SA, S. Induction machine drive condition monitoring and diagnostic research—A survey. Electr. Power
Syst. Res. 2003, 64, 145–158. [CrossRef]
17. Neupane, D.; Kim, Y.; Seok, J. Bearing Fault Detection Using Scalogram and Switchable Normalization-Based CNN (SN-CNN).
IEEE Access 2021, 9, 88151–88166. [CrossRef]
Energies 2023, 16, 5602 19 of 22

18. Prieto, M.D.; Cirrincione, G.; Espinosa, A.G.; Ortega, J.A.; Henao, H. Bearing fault detection by a novel condition-monitoring
scheme based on statistical-time features and neural networks. IEEE Trans. Ind. Electron. 2013, 60, 3398–3407. [CrossRef]
19. Immovilli, F.; Bellini, A.; Rubini, R.; Tassoni, C. Diagnosis of bearing faults in induction machines by vibration or current signals:
A critical comparison. IEEE Trans. Ind. Appl. 2010, 46, 1350–1359. [CrossRef]
20. Elbouchikhi, E.; Amirat, Y.; Feld, G.; Benbouzid, M. Generalized likelihood ratio test based approach for stator-fault detection in a
PWM inverter-fed induction motor drive. IEEE Trans. Ind. Electron. 2019, 66, 6343–6353. [CrossRef]
21. Siddique, A.; Yadava, G.S.; Singh, B. A review of stator fault monitoring techniques of induction motors. IEEE Trans. Energy
Convers. 2005, 20, 106–114. [CrossRef]
22. Jung, J.H.; Lee, J.J.; Kwon, B.H. Online diagnosis of induction motors using MCSA. IEEE Trans. Ind. Electron. 2006, 53, 1842–1852.
[CrossRef]
23. Milkovic, D. Brief review of motor current signature analysis. HDKBR INFO Mag. 2015, 5, 14–26.
24. Toliyat, H.A.; Nandi, S.; Choi, S.; Meshgin-Kelk, H. Electric Machines: Modeling, Condition Monitoring, and Fault Diagnosis; CRC
Press: Boca Raton, FL, USA, 2012.
25. Chen, W.; Zhang, L.; Pattipati, K.; Bazzi, A.M.; Joshi, S.; Dede, E.M. Data-Driven Approach for Fault Prognosis of SiC MOSFETs.
IEEE Trans. Power Electron. 2020, 35, 4048–4062. [CrossRef]
26. Morozumi, A.; Yamada, K.; Miyasaka, T.; Sumi, S.; Seki, Y. Reliability of power cycling for IGBT power semiconductor modules.
IEEE Trans. Ind. Appl. 2003, 39, 665–671. [CrossRef]
27. Franke, J.; Zeng, G.; Winkler, T.; Lutz, J. Power cycling reliability results of GaN HEMT devices. In Proceedings of the International
Symposium on Power Semiconductor Devices and ICs 2018, Chicago, IL, USA, 13–17 May 2018; pp. 467–470. [CrossRef]
28. Meneghini, M.; Fabris, E.; Ruzzarin, M.; De Santi, C.; Nomoto, K.; Hu, Z.; Li, W.; Gao, X.; Jena, D.; Xing, H.G.; et al. Degradation
Mechanisms of GaN-Based Vertical Devices: A Review. Phys. Status Solidi (A) Appl. Mater. Sci. 2020, 217, 1900750. [CrossRef]
29. Xu, C.; Yang, F.; Ugur, E.; Pu, S.; Akin, B. Performance Degradation of GaN HEMTs Under Accelerated Power Cycling Tests.
CPSS Trans. Power Electron. Appl. 2018, 3, 269–277. [CrossRef]
30. Falck, J.; Felgemacher, C.; Rojko, A.; Liserre, M.; Zacharias, P. Reliability of Power Electronic Systems. IEEE Ind. Electron. Mag.
2018, 12, 24–35. [CrossRef]
31. Zhao, Z.; Davari, P.; Lu, W.; Wang, H.; Blaabjerg, F. An Overview of Condition Monitoring Techniques for Capacitors in DC-Link
Applications. IEEE Trans. Power Electron. 2021, 36, 3692–3716. [CrossRef]
32. Wang, T.; Liu, Z.; Lu, G.; Liu, J. Temporal-Spatio Graph Based Spectrum Analysis for Bearing Fault Detection and Diagnosis.
IEEE Trans. Ind. Electron. 2021, 68, 2598–2607. [CrossRef]
33. Huang, X.; Wen, G.; Dong, S.; Zhou, H.; Lei, Z.; Zhang, Z.; Chen, X. Memory Residual Regression Autoencoder for Bearing Fault
Detection. IEEE Trans. Instrum. Meas. 2021, 70, 3515512. [CrossRef]
34. Zarei, J.; Tajeddini, M.A.; Karimi, H.R. Vibration analysis for bearing fault detection and classification using an intelligent filter.
Mechatronics 2014, 24, 151–157. [CrossRef]
35. Roy, S.S.; Dey, S.; Chatterjee, S. Autocorrelation Aided Random Forest Classifier-Based Bearing Fault Detection Framework. IEEE
Sens. J. 2020, 20, 10792–10800. [CrossRef]
36. Zhang, B.; Sconyers, C.; Byington, C.; Patrick, R.; Orchard, M.E.; Vachtsevanos, G. A probabilistic fault detection approach:
Application to bearing fault detection. IEEE Trans. Ind. Electron. 2011, 58, 2011–2018. [CrossRef]
37. Minhas, A.S.; Kankar, P.K.; Kumar, N.; Singh, S. Bearing fault detection and recognition methodology based on weighted
multiscale entropy approach. Mech. Syst. Signal Process. 2021, 147, 107073. [CrossRef]
38. Shen, S.; Lu, H.; Sadoughi, M.; Hu, C.; Nemani, V.; Thelen, A.; Webster, K.; Darr, M.; Sidon, J.; Kenny, S. A physics-informed deep
learning approach for bearing fault detection. Eng. Appl. Artif. Intell. 2021, 103, 104295. [CrossRef]
39. Li, J.; Yu, Q.; Wang, X.; Zhang, Y. An enhanced rolling bearing fault detection method combining sparse code shrinkage denoising
with fast spectral correlation. ISA Trans. 2020, 102, 335–346. [CrossRef]
40. Tang, G.; Wang, Y.; Huang, Y.; Liu, N.; He, J. Compound Bearing Fault Detection Under Varying Speed Conditions With Virtual
Multichannel Signals in Angle Domain. IEEE Trans. Instrum. Meas. 2020, 69, 5535–5545. [CrossRef]
41. Barcelos, A.S.; Marques Cardoso, A.J. Current-based bearing fault diagnosis using deep learning algorithms. Energies 2021,
14, 2509. [CrossRef]
42. Choudhary, A.; Goyal, D.; Letha, S.S. Infrared Thermography-Based Fault Diagnosis of Induction Motor Bearings Using Machine
Learning. IEEE Sens. J. 2021, 21, 1727–1734. [CrossRef]
43. Frosini, L.; Harlisca, C.; Szabo, L. Induction machine bearing fault detection by means of statistical processing of the stray flux
measurement. IEEE Trans. Ind. Electron. 2015, 62, 1846–1854. [CrossRef]
44. Tang, J.; Chen, J.; Dong, K.; Yang, Y.; Lv, H.; Liu, Z. Modeling and evaluation of stator and rotor faults for induction motors.
Energies 2019, 13, 133. [CrossRef]
45. Tang, J.; Yang, Y.; Chen, J.; Qiu, R.; Liu, Z. Characteristics analysis and measurement of inverter-fed induction motors for stator
and rotor fault detection. Energies 2019, 13, 101. [CrossRef]
46. Skowron, M.; Orlowska-Kowalska, T.; Wolkiewicz, M.; Kowalski, C.T. Convolutional neural network-based stator current
data-driven incipient stator fault diagnosis of inverter-fed induction motor. Energies 2020, 13, 1475. [CrossRef]
47. Lee, S.T.; Hur, J. Detection technique for stator inter-turn faults in BLDC motors based on third-harmonic components of line
currents. IEEE Trans. Ind. Appl. 2017, 53, 143–150. [CrossRef]
Energies 2023, 16, 5602 20 of 22

48. Cruz SM, A.; Cardoso AJ, M. Diagnosis of stator inter-turn short circuits in DTC induction motor drives. IEEE Trans. Ind. Appl.
2004, 40, 1349–1360. [CrossRef]
49. Wang, C.; Delgado Prieto, M.; Romeral, L.; Chen, Z.; Blaabjerg, F.; Liu, X. Detection of Partial Demagnetization Fault in PMSMs
Operating under Nonstationary Conditions. IEEE Trans. Magn. 2016, 52, 3–6. [CrossRef]
50. Shifat, T.A.; Hur, J.W. An Effective Stator Fault Diagnosis Framework of BLDC Motor Based on Vibration and Current Signals.
IEEE Access 2020, 8, 106968–106981. [CrossRef]
51. Akhil Vinayak, B.; Anjali Anand, K.; Jagadanand, G. Wavelet-based real-time stator fault detection of inverter-fed induction
motor. IET Electr. Power Appl. 2020, 14, 82–90. [CrossRef]
52. Pietrzak, P.; Wolkiewicz, M. On-line detection and classification of pmsm stator winding faults based on stator current symmetrical
components analysis and the knn algorithm. Electronics 2021, 10, 1786. [CrossRef]
53. Quiroz, J.C.; Mariun, N.; Mehrjou, M.R.; Izadi, M.; Misron, N.; Mohd Radzi, M.A. Fault detection of broken rotor bar in LS-PMSM
using random forests. Meas. J. Int. Meas. Confed. 2018, 116, 273–280. [CrossRef]
54. Palacios RH, C.; Da Silva, I.N.; Goedtel, A.; Godoy, W.F.; Lopes, T.D. Diagnosis of Stator Faults Severity in Induction Motors
Using Two Intelligent Approaches. IEEE Trans. Ind. Inform. 2017, 13, 1681–1691. [CrossRef]
55. Singh, G.; Naikan VN, A. Detection of half broken rotor bar fault in VFD driven induction motor drive using motor square
current MUSIC analysis. Mech. Syst. Signal Process. 2018, 110, 333–348. [CrossRef]
56. Ramu, S.K.; Raj Irudayaraj, G.C.; Subramani, S.; Subramaniam, U. Broken rotor bar fault detection using Hilbert transform and
neural networks applied to direct torque control of induction motor drive. IET Power Electron. 2020, 13, 3328–3338. [CrossRef]
57. Park, Y.; Yang, C.; Kim, J.; Kim, H.; Bin Lee, S.; Gyftakis, K.N.N.; Panagiotou, P.A.; Kia, S.H.; Capolino, G.-A. Stray flux monitoring
for reliable detection of rotor faults under the influence of rotor axial air ducts. IEEE Trans. Ind. Electron. 2019, 66, 7561–7570.
[CrossRef]
58. Mirzaeva, G.; Saad, K.I. Advanced Diagnosis of Rotor Faults and Eccentricity in Induction Motors Based on Internal Flux
Measurement. IEEE Trans. Ind. Appl. 2018, 54, 3961–3970. [CrossRef]
59. Park, Y.; Yang, C.; Bin Lee, S.; Lee, D.-M.; Fernandez, D.; Reigosa, D.; Briz, F. Online detection and classification of rotor and load
defects in PMSMs Based on Hall sensor measurements. IEEE Trans. Ind. Appl. 2019, 55, 3803–3812. [CrossRef]
60. Gyftakis, K.N.; Cardoso AJ, M. Reliable Detection of Stator Interturn Faults of Very Low Severity Level in Induction Motors. IEEE
Trans. Ind. Electron. 2021, 68, 3475–3484. [CrossRef]
61. Gurusamy, V.; Bostanci, E.; Li, C.; Qi, Y.; Akin, B. A Stray Magnetic Flux-Based Robust Diagnosis Method for Detection and
Location of Interturn Short Circuit Fault in PMSM. IEEE Trans. Instrum. Meas. 2021, 70, 3500811. [CrossRef]
62. Goktas, T.; Zafarani, M.; Lee, K.W.; Akin, B.; Sculley, T. Comprehensive Analysis of Magnet Defect Fault Monitoring Through
Leakage Flux. IEEE Trans. Magn. 2017, 53, 8201010. [CrossRef]
63. Maraaba, L.; Al-Hamouz, Z.; Abido, M. An efficient stator inter-Turn fault diagnosis tool for induction motors. Energies 2018,
11, 653. [CrossRef]
64. Glowacz, A.; Glowacz, W.; Glowacz, Z.; Kozik, J. Early fault diagnosis of bearing and stator faults of the single-phase induction
motor using acoustic signals. Meas. J. Int. Meas. Confed. 2018, 113, 1–9. [CrossRef]
65. Glowacz, A. Fault diagnosis of single-phase induction motor based on acoustic signals. Mech. Syst. Signal Process. 2019,
117, 65–80. [CrossRef]
66. Toma, R.N.; Kim, J.M. Article bearing fault classification of induction motors using discrete wavelet transform and ensemble
machine learning algorithms. Appl. Sci. 2020, 10, 5251. [CrossRef]
67. Heydarzadeh, M.; Zafarani, M.; Nourani, M.; Akin, B. A Wavelet-Based Fault Diagnosis Approach for Permanent Magnet
Synchronous Motors. IEEE Trans. Energy Convers. 2019, 34, 761–772. [CrossRef]
68. Wen, L.; Li, X.; Gao, L.; Zhang, Y. A New Convolutional Neural Network-Based Data-Driven Fault Diagnosis Method. IEEE Trans.
Ind. Electron. 2018, 65, 5990–5998. [CrossRef]
69. Mao, W.; Liu, Y.; Ding, L.; Li, Y. Imbalanced fault diagnosis of rolling bearing based on generative adversarial network:
A comparative study. IEEE Access 2019, 7, 9515–9530. [CrossRef]
70. Zhuang, F.; Qi, Z.; Duan, K.; Xi, D.; Zhu, Y.; Zhu, H.; Xiong, H.; He, Q. A Comprehensive Survey on Transfer Learning. Proc. IEEE
2021, 109, 43–76. [CrossRef]
71. Sun, W.; Shao, S.; Zhao, R.; Yan, R.; Zhang, X.; Chen, X. A sparse auto-encoder-based deep neural network approach for induction
motor faults classification. Meas. J. Int. Meas. Confed. 2016, 89, 171–178. [CrossRef]
72. Abid, F.; Sallem, M.B.; Braham, A. Robust Interpretable Deep Learning for Intelligent Fault Diagnosis of Induction Motors. IEEE
Trans. Instrum. Meas. 2020, 69, 3506–3515. [CrossRef]
73. Case Western Reserve University Bearing Data Center Seeded Fault Test Data. (n.d.). Available online: https://engineering.case.
edu/bearingdatacenter (accessed on 24 July 2023).
74. Xi’an Jiaotong University—Sumyoung Technology (XJTU-SY) Bearing Datasets. (n.d.). Available online: https://biaowang.tech/
xjtu-sy-bearing-datasets/ (accessed on 24 July 2023).
75. Lee, J.; Qiu, H.; Yu, G.; Lin, J.; Services, R.T.; Lee, H.; Qiu, G.; Yu, J.L. Bearing Data Set. IMS, University of Cincinnati. 2007. Available
online: https://www.nasa.gov/content/prognostics-center-of-excellence-data-set-repository (accessed on 24 July 2023).
76. Repository, N.P.D. (n.d.). FEMTO Bearing Data Set. Available online: https://www.nasa.gov/content/prognostics-center-of-
excellence-data-set-repository (accessed on 24 July 2023).
Energies 2023, 16, 5602 21 of 22

77. Rodríguez-Blanco, M.A.; Claudio-Sánchez, A.; Theilliol, D.; Vela-Valdés, L.G.; Sibaja-Terán, P.; Hernández-González, L.;
Aguayo-Alquicira, J. A failure-detection strategy for IGBT based on gate-voltage behavior applied to a motor drive system. IEEE
Trans. Ind. Electron. 2011, 58, 1625–1633. [CrossRef]
78. Climaco-Arvizu, O.; Hernández-González, L.; Rodríguez-Blanco, M.A. Fault detection for SiC-Mosfet based on the behavior of
gate signal. In Proceedings of the SDEMPED 2015: IEEE 10th International Symposium on Diagnostics for Electrical Machines,
Power Electronics and Drives, Guarda, Portugal, 1–4 September 2015; pp. 71–76. [CrossRef]
79. Lyu, X.; Li, H.; Abdullah, Y.; Wang, K.; Hu, B.; Yang, Z.; Liu, J.; Wang, J.; Liu, L.; Bala, S. A Reliable Ultrafast Short-Circuit
Protection Method for E-Mode GaN HEMT. IEEE Trans. Power Electron. 2020, 35, 8926–8933. [CrossRef]
80. Kumar, M. Open Circuit Fault Detection and Switch Identification for LS-PWM H-Bridge Inverter. IEEE Trans. Circuits Syst. II
Express Briefs 2021, 68, 1363–1367. [CrossRef]
81. El Khil, S.K.; Jlassi, I.; Marques Cardoso, A.J.; Estima, J.O.; Mrabet-Bellaaj, N. Diagnosis of Open-Switch and Current Sensor
Faults in PMSM Drives through Stator Current Analysis. IEEE Trans. Ind. Appl. 2019, 55, 5925–5937. [CrossRef]
82. Li, K.; Cheng, S.; Yu, T.; Wu, X.; Xiang, C.; Bilal, A. An On-Line Multiple Open-Circuit Fault Diagnostic Technique for Railway
Vehicle Air-Conditioning Inverters. IEEE Trans. Veh. Technol. 2020, 69, 7026–7039. [CrossRef]
83. Lamb, J.; Mirafzal, B. Open-Circuit IGBT Fault Detection and Location Isolation for Cascaded Multilevel Converters. IEEE Trans.
Ind. Electron. 2017, 64, 4846–4856. [CrossRef]
84. Trabelsi, M.; Boussak, M.; Benbouzid, M. Multiple criteria for high performance real-time diagnostic of single and multiple
open-switch faults in ac-motor drives: Application to IGBT-based voltage source inverter. Electr. Power Syst. Res. 2017,
144, 136–149. [CrossRef]
85. Farhadi, M.; Fard, M.T.; Abapour, M.; Hagh, M.T. DC-AC Converter-Fed Induction Motor Drive with Fault-Tolerant Capability
under Open- and Short-Circuit Switch Failures. IEEE Trans. Power Electron. 2018, 33, 1609–1621. [CrossRef]
86. Jlassi, I.; Cardoso AJ, M. A Single Method for Multiple IGBT, Current, and Speed Sensor Faults Diagnosis in Regenerative PMSM
Drives. IEEE J. Emerg. Sel. Top. Power Electron. 2020, 8, 2583–2599. [CrossRef]
87. Cai, B.; Zhao, Y.; Liu, H.; Xie, M. A Data-Driven Fault Diagnosis Methodology in Three-Phase Inverters for PMSM Drive Systems.
IEEE Trans. Power Electron. 2017, 32, 5590–5600. [CrossRef]
88. Yan, H.; Xu, Y.; Cai, F.; Zhang, H.; Zhao, W.; Gerada, C. PWM-VSI Fault Diagnosis for a PMSM Drive Based on the Fuzzy Logic
Approach. IEEE Trans. Power Electron. 2018, 34, 759–768. [CrossRef]
89. Kiranyaz, S.; Gastli, A.; Ben-Brahim, L.; Al-Emadi, N.; Gabbouj, M. Real-Time Fault Detection and Identification for MMC Using
1-D Convolutional Neural Networks. IEEE Trans. Ind. Electron. 2019, 66, 8760–8771. [CrossRef]
90. Moosavi, S.S.; Kazemi, A.; Akbari, H. A comparison of various open-circuit fault detection methods in the IGBT-based DC/AC
inverter used in electric vehicle. Eng. Fail. Anal. 2019, 96, 223–235. [CrossRef]
91. Gou, B.; Xu, Y.; Xia, Y.; Deng, Q.; Ge, X. An Online Data-Driven Method for Simultaneous Diagnosis of IGBT and Current Sensor
Fault of Three-Phase PWM Inverter in Induction Motor Drives. IEEE Trans. Power Electron. 2020, 35, 13281–13294. [CrossRef]
92. Li, Z.; Wheeler, P.; Watson, A.; Costabeber, A.; Wang, B.; Ren, Y.; Bai, Z.; Ma, H. A Fast Diagnosis Method for Both IGBT Faults
and Current Sensor Faults in Grid-Tied Three-Phase Inverters with Two Current Sensors. IEEE Trans. Power Electron. 2020,
35, 5267–5278. [CrossRef]
93. Yang, Q.; Gultekin, M.A.; Seferian, V.; Pattipati, K.; Bazzi, A.M.; Palmieri, F.A.N.; Rajamani, R.; Joshi, S.; Farooq, M.; Ukegawa, H.
Incipient Residual-Based Anomaly Detection in Power Electronic Devices. IEEE Trans. Power Electron. 2022, 37, 7315–7332.
[CrossRef]
94. Jlassi, I.; Estima, J.O.; El Khil, S.K.; Bellaaj, N.M.; Cardoso AJ, M. A Robust Observer-Based Method for IGBTs and Current Sensors
Fault Diagnosis in Voltage-Source Inverters of PMSM Drives. IEEE Trans. Ind. Appl. 2017, 53, 2894–2905. [CrossRef]
95. Maamouri, R.; Trabelsi, M.; Boussak, M.; M’Sahli, F. Fault Diagnosis and Fault Tolerant Control of a Three-Phase VSI Supplying
Sensorless Speed Controlled Induction Motor Drive. Electr. Power Compon. Syst. 2018, 46, 2159–2173. [CrossRef]
96. Zhou, D.; Yang, S.; Tang, Y. A Voltage-Based Open-Circuit Fault Detection and Isolation Approach for Modular Multilevel
Converters with Model-Predictive Control. IEEE Trans. Power Electron. 2018, 33, 9866–9874. [CrossRef]
97. Chai, M.; Gorla NB, Y.; Panda, S.K. Fault Detection and Localization for Cascaded H-Bridge Multilevel Converter with Model
Predictive Control. IEEE Trans. Power Electron. 2020, 35, 10109–10120. [CrossRef]
98. Salmasi, F.R. A Self-Healing Induction Motor Drive With Model Free Sensor Tampering and Sensor Fault Detection, Isolation,
and Compensation. IEEE Trans. Ind. Electron. 2017, 64, 6105–6115. [CrossRef]
99. Chakraborty, C.; Verma, V. Speed and current sensor fault detection and isolation technique for induction motor drive using axes
transformation. IEEE Trans. Ind. Electron. 2015, 62, 1943–1954. [CrossRef]
100. Skowron, M.; Teler, K.; Adamczyk, M.; Orlowska-Kowalska, T. Classification of Single Current Sensor Failures in Fault-Tolerant
Induction Motor Drive Using Neural Network Approach. Energies 2022, 15, 6646. [CrossRef]
101. Tran, C.D.; Palacky, P.; Kuchar, M.; Brandstetter, P.; Dinh, B.H. Current and Speed Sensor Fault Diagnosis Method Applied to
Induction Motor Drive. IEEE Access 2021, 9, 38660–38672. [CrossRef]
102. Gou, B.; Xu, Y.; Xia, Y.; Wilson, G.; Liu, S. An Intelligent Time-Adaptive Data-Driven Method for Sensor Fault Diagnosis in
Induction Motor Drive System. IEEE Trans. Ind. Electron. 2019, 66, 9817–9827. [CrossRef]
103. Dybkowski, M.; Klimkowski, K. Artificial neural network application for current sensors fault detection in the vector controlled
induction motor drive. Sensors 2019, 19, 571. [CrossRef]
Energies 2023, 16, 5602 22 of 22

104. Sun, P.; Gong, C.; Du, X.; Luo, Q.; Wang, H.; Zhou, L. Online Condition Monitoring for Both IGBT Module and DC-Link Capacitor
of Power Converter Based on Short-Circuit Current Simultaneously. IEEE Trans. Ind. Electron. 2017, 64, 3662–3671. [CrossRef]
105. Li, T.; Chen, J.; Cong, P.; Dai, X.; Qiu, R.; Liu, Z. Online Condition Monitoring of DC-Link Capacitor for AC/DC/AC PWM
Converter. IEEE Trans. Power Electron. 2022, 37, 865–878. [CrossRef]
106. Abo-Khalil, A.G.; Al-Qawasmi, A.R.; Eltamaly, A.M.; Yu, B.G. Condition monitoring of dc-link electrolytic capacitors in PWM
power converters using OBL method. Sustainability 2020, 12, 3719. [CrossRef]
107. Wechsler, A.; Mecrow, B.C.; Atkinson, D.J.; Bennett, J.W.; Benarous, M. Condition monitoring of DC-link capacitors in aerospace
drives. IEEE Trans. Ind. Appl. 2012, 48, 1866–1874. [CrossRef]
108. Seferian, V.; Bazzi, A.; Hajj, H. Condition Monitoring of DC-link Capacitors in Grid-tied Solar Inverters Using Data-Driven
Techniques. In Proceedings of the 2020 IEEE Energy Conversion Congress and Exposition (ECCE), Detroit, MI, USA, 11–15
October 2020; pp. 5318–5323. [CrossRef]
109. Sundararajan, P.; Sathik MH, M.; Sasongko, F.; Tan, C.S.; Tariq, M.; Simanjorang, R. Online Condition Monitoring System for
DC-Link Capacitor in Industrial Power Converters. IEEE Trans. Ind. Appl. 2018, 54, 4775–4785. [CrossRef]
110. Laadjal, K.; Sahraoui, M.; Cardoso AJ, M. On-Line Fault Diagnosis of DC-Link Electrolytic Capacitors in Boost Converters Using
the STFT Technique. IEEE Trans. Power Electron. 2021, 36, 6303–6312. [CrossRef]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.