ISSN-e: 2737-6419
Period: Jan-Mar of 2026
Revista Athenea
Vol.7, Issue 23, (pp. 28Ű40)
Research article https://doi.org/10.47460/athenea.v7i23.126
Comparison of the effectiveness between the Wavelet Transform and
the WignerśVille Transform for the diagnosis of low insulation in the
starting transient of induction motors
Alfredo Marot*
https://orcid.org/0000-0002-8829-4124
aamarotgu@estudiante.unexpo.com
UNEXPO Vicerrectorado Puerto Ordaz
Doctorado en Ciencias de la Ingeniería
Puerto Ordaz, Venezuela
Sergio Velásquez
https://orcid.org/0000-0002-3516-4430
svelasquez@unexpo.edu.ve
UNEXPO Vicerrectorado Puerto Ordaz
Puerto Ordaz, Venezuela
*Corresponding author:
aamarotgu@estudiante.unexpo.com
Received (29/09/2025), Accepted (03/12/2025)
Abstract. This study demonstrates the superiority of the Discrete Wavelet Transform over the WignerŰ
Ville Transform in detecting insulation faults in induction motors during startup transients. The com-
prehensive analysis of 360 simulated signals revealed that the wavelet technique with Daubechies 10
achieves signiĄcantly higher classiĄcation accuracy (74.44% vs. 67.78%), substantially outperforming
its counterpart. Certain decomposition levels showed maximum sensitivity with variations up to +354%,
while diagnostic reliability indicators conĄrm its robustness. This technique positions itself as an opti-
mal solution for predictive monitoring systems, enabling early fault detection that substantially reduces
downtime and industrial maintenance costs.
Keywords: wavelet transform, Wigner-Ville transform, induction motor, fault diagnosis, insulation
failure, signal processing.
Comparación de la efectividad entre la Transformada Wavelet y la Transformada
de Wigner–Ville para el diagnóstico de bajo aislamiento en el transitorio de
arranque de motores de inducción
Resumen. Este estudio demuestra la superioridad de la Transformada Wavelet Discreta frente a la
Transformada de WignerŰVille en la detección de fallas de aislamiento en motores de inducción durante
transitorios de arranque. El análisis exhaustivo de 360 señales simuladas reveló que la técnica wavelet
con Daubechies 10 alcanza una precisión de clasiĄcación notablemente superior (74.44% vs. 67.78%),
superando signiĄcativamente a su contraparte. Ciertos niveles de descomposición mostraron sensibilidad
máxima con variaciones de hasta +354%, mientras que los indicadores de conĄabilidad diagnóstica
conĄrman su robustez. Esta técnica se posiciona como solución óptima para sistemas de monitoreo
predictivo, permitiendo la detección temprana de fallas que reduce sustancialmente los tiempos de
inactividad y los costos de mantenimiento industrial.
Palabras clave: transformada wavelet, transformada de Wigner-Ville, motor de inducción, diagnóstico
de fallas, falla de aislamiento, procesamiento de señales.
Marot A., y Velázquez S. Comparison of the effectiveness between the Wavelet Transform and the
Wigner-Ville Transform...
28
ISSN-e: 2737-6419
Period: Jan-Mar of 2026
Revista Athenea
Vol.7, Issue 23, (pp. 28Ű40)
I. INTRODUCTION
Induction motors play a crucial role in the operation of industrial machinery, and their reliable
operation is vital for productivity and safety. Faults can lead to downtime, increased maintenance
costs, and loss of efficiency [
1]. The industry has adopted motor fault detection as a means to ensure
the consistent and reliable operation of modern industrial systems. Motors are subjected to various
conditions and environments [2]. Stator-related problems account for 38% of all faults in induction
motors [
3].
When an inter-turn insulation fault occurs, some turns in the stator winding short-circuit. Failure to
detect and rectify this fault promptly accelerates its progression to more severe faults, such as inter-coil
or inter-phase short circuits [4], culminating in phase-to-ground faults that can result in permanent
damage to the induction motor [
5]. Recognized challenges include real-time monitoring and model
interpretability, emphasizing the need for practical solutions. The severe impact of motor failures on
productivity and safety underscores the urgency for early fault detection systems.
Over 40 years of research have demonstrated that many types of faults can be detected in motor
sensor data before they cause a breakdown; therefore, there has been signiĄcant interest in leveraging
this fact to reduce the losses caused by such failures [
6]. More than 40 years of research conĄrm that
it is possible to detect these faults through the analysis of sensor data before a catastrophic failure
occurs, thereby reducing associated economic losses [
7].
In this context, the startup transient emerges as a critical phase where thermal and magnetic stress
can reveal insulation weaknesses that are not visible under steady-state conditions. Time-frequency
analysis tools, such as the Discrete Wavelet Transform (DWT) and the WignerŰVille Transform (WVD),
have emerged as promising alternatives. However, the need arises to determine which of these techniques
offers greater robustness in the face of the non-stationary nature of the startup.
To address this problem, the present research poses the following research questions:
1. Is the Discrete Wavelet Transform more effective than the WignerŰVille Transform for discrimi-
nating incipient insulation fault levels during the startup transient?
2. Which statistical metrics derived from these transformations p osse ss greater separability power
to characterize the motorŠs condition?
To answer these questions, the following speciĄc objectives are proposed:
1. Model and simulate the behavior of induction motors under healthy operating conditions and
with incipient insulation faults (200 kΩ and 20 kΩ).
2. Extract a set of statistical features (energy, entropy, RMS, and kurtosis) from the representations
generated by the DWT and the WVD.
3. Evaluate and compare the performance of both techniques using a Support Vector Machine
(SVM) classiĄer and separability metrics such as the Fisher Score.
Finally, the success criterion for the comparative diagnosis is deĄned by obtaining a classiĄcation
model with an accuracy greater than 70%, an area under the curve (AUC) value greater than 0.80,
and the identiĄcation of at least three metrics with statistical signiĄcance (p < 0.1) that allow clear
differentiation between the analyzed classes.
Marot A., y Velázquez S. Comparison of the effectiveness between the Wavelet Transform and the
Wigner-Ville Transform...
29
ISSN-e: 2737-6419
Period: Jan-Mar of 2026
Revista Athenea
Vol.7, Issue 23, (pp. 28Ű40)
II. THEORETICAL BACKGROUND
A. Discrete Wavelet Transform (DWT)
The wavelet transform is a tool applied to analysis in the time and frequency domains. Ingrid
Daubechies, a leading Ągure in wavelet research, revolutionized the Ąeld by creating compactly supported
orthonormal wavelets, which enabled the development of discrete wavelet analysis. Members of this
family are identiĄed by the nomenclature dbN , where db corresponds to the researcherŠs surname and
N indicates the order of the wavelet [
8].
The analysis of the motorŠs transient current is performed in the time-scale domain using the Dis-
crete Wavelet Transform (DWT). The DWT is based on Multiresolution Analysis (MRA), decomposing
the signal x[n] into detail coefficients (D
j
) and approximation coefficients (A
j
) through a dyadic Ąlter
bank according to MallatŠs pioneering algorithm [
9]. Formally, the coefficients at level j are obtained
through discrete convolution and downsampling operations, using a low-pass Ąlter h[n] for the approx-
imation and a high-pass Ąlter g[n] for the detail:
A
j
[k] =
X
n
A
j−1
[n ] · h[2k − n] (1)
D
j
[k] =
X
n
A
j−1
[n ] · g[2k − n] (2)
where A
0
[n ] = x[n] is the input signal. The justiĄcation for the DWT lies in its ability to provide
a time-frequency representation of the signal, making it optimal for the analysis of transients and
non-stationary signals, which are characteristic of insulation faults in electrical machines.
For this study, the Daubechies wavelet of order 10 (db10) is selected, and a decomposition level
of 8 (J = 8). The db10 wavelet belongs to the family of orthogonal compactly supported wavelets.
The selection of N = 10 is supported by two fundamental properties. First, its compact support:
the Ąlter support is Ąnite (length 2N = 20 coefficients), which guarantees computational efficiency
and good temporal localization for the detection of impulsive events. Second, its vanishing moments
(N = 10): db10 has ten vanishing moments, meaning it is orthogonal to polynomials up to degree 9.
This property is crucial, as it ensures that the detail coefficients (D
j
) do not contain information about
smooth low-frequency variations (trend), but instead concentrate on the energy of irregularities, peaks,
or singularities in the signal [
10].
The decomposition level J = 8 ensures adequate frequency resolution to isolate the bands where
the fault energy manifests. With f
s
= 5000 Hz, the decomposition generates 8 detail bands, where D
1
corresponds to the highest frequency range (1250Ű2500 Hz) and D
8
represents the very low-frequency
band (9.77Ű19.53 Hz).
B. Wigner±Ville Transform (WVD)
The WignerŰVille Transform is the principal member of the class of bilinear (or quadratic) time-
frequency distributions. Its main strength is providing optimal resolution simultaneously in the time
(t) and frequency (f) domains. The WVD is based on the notion of an instantaneous autocorrelation
function of the signal. For a signal s(t), the WVD is generally deĄned over its analytic signal z(t) to
avoid interference caused by negative frequency components [
11].
W
z
(t, f ) =
Z
∞
−∞
z
t +
τ
2
z
∗
t −
τ
2
e
−j2πf τ
dτ (3)
where W
z
(t, f ) is the WignerŰVille distribution, τ is the integration variable representing the time inter-
val over which the instantaneous autocorrelation is calculated, and e
−j2πf τ
is the complex exponential
factor that converts information from the delay domain (τ) to the frequency domain (f), acting as a
band-pass Ąlter for each instant t.
Marot A., y Velázquez S. Comparison of the effectiveness between the Wavelet Transform and the
Wigner-Ville Transform...
30
ISSN-e: 2737-6419
Period: Jan-Mar of 2026
Revista Athenea
Vol.7, Issue 23, (pp. 28Ű40)
III. METHODOLOGY
This study tests the hypothesis that the Discrete Wavelet Transform (DWT) is more effective than
the WignerŰVille Transform (WVD) for diagnosing stator insulation faults during motor startup. Using
numerical simulation, it ensures controlled conditions, safe replication of incipient faults, and generates
a labeled dataset for a systematic comparison of the two techniques.
A. Simulation of Motors and Fault Modeling
A simulation environment was implemented in MATLAB/Simulink to replic ate the dynamic behavior
of 20 squirrel-cage induction motors with rated power distributed in the range of 10 to 4000 HP. This
selection ensures industrial representativeness by covering prototypical applications, including pumping,
ventilation, and compression systems predominant in manufacturing and continuous process sectors
(see Fig.
1).
Fig. 1. Squirrel cage motor simulation in Simulink. The Ągure was generated in MATLAB 2019.
The characteristics (rated power, voltage, resistances, and inductances) corresponding to 20 induc-
tion motors were parameterized, ensuring an accurate representation of the three-phase system across
various industrial power ranges. The insulation fault was simulated by connecting a resistive-capacitive
(RC) circuit between phase and ground of each winding [
1]. The resistance of this circuit was adjusted
to 200 kΩ and 20 kΩ to represent incipient and critical faults, respectively. The phase-to-ground
capacitances were set between 1.5 nF and 21 nF [
12].
B. Signal Acquisition and Structuring
A simulated cohort of 20 three-phase squirrel-cage induction motors with rated power from 10 to
4000 HP ensured industrial representativeness for applications such as pumping and ventilation. Data
generation followed a factorial design combining three insulation states (healthy, 200 kΩ fault, 20 kΩ
fault) and two load regimes (no-load, rated load) across each motorŠs three phases, yielding a total
of 360 three-phase signals (120 healthy, 240 faulty). Acquisition focused on the startup transient,
capturing a 5-second window at 5000 Hz to adequately resolve non-stationary components. All signals
were exported to MATLAB/Simulink, where time-frequency analysis algorithms were applied for the
systematic extraction of diagnostic descriptors.
C. Feature Extraction using DWT and WVD
The processing of the acquired signals employed two feature extraction approaches to transform
information from the time domain into a discriminative attribute space. A Daubechies 10 (db10)
wavelet with eight decomposition levels was applied, leveraging its compact support and ten vanishing
moments for precise transient detection and low-frequency trend Ąltering [
9], [10]. The decomposition
level J = 8 was determined considering the sampling frequency (f
s
= 5000 Hz) and the spectral range
of interest for insulation faults (approximately 9.77Ű1250 Hz). From the detail coefficients obtained
Marot A., y Velázquez S. Comparison of the effectiveness between the Wavelet Transform and the
Wigner-Ville Transform...
31
ISSN-e: 2737-6419
Period: Jan-Mar of 2026
Revista Athenea
Vol.7, Issue 23, (pp. 28Ű40)
at each level D
j
(with j = 1, 2, . . . , 8) and for each of the three phases, Ąve statistical metrics were
systematically calculated.
Energy:
E
j
=
P
N
j
k=1
D
j
[k]
2
P
8
i=1
E
i
(4)
where D
j
[k] is the k-th detail coefficient at level j and N
j
is the number of coefficients at that level.
Maximum Value (MAX
j
) and Minimum Value (MIN
j
):
MAX
j
= max(D
j
), MIN
j
= min(D
j
) (5)
with units expressed in amperes (A).
Shannon Entropy:
H
j
= −
N
j
X
k=1
p
k
log
2
(p
k
), p
k
=
|D
j
[k]|
2
E
j
(6)
Kurtosis:
K
j
=
1
N
j
P
N
j
k=1
(
D
j
[k] −
¯
D
j
)
4
σ
4
D
j
(7)
where
¯
D
j
and σ
D
j
are the mean and standard deviation of D
j
, respectively.
Root Mean Square (RMS):
RM S
j
=
v
u
u
t
1
N
j
N
j
X
k=1
D
j
[k]
2
; Amperes(A) (8)
with units expressed in amperes (A).
The Daubechies 10 wavelet transform was employed to characterize the impulsivity of insulation
faults. Concurrently, the WignerŰVille Distribution (WVD), a quadratic time-frequency representation
offering optimal joint resolution, was applied [
10]. Following the standard formulation to mitigate
cross-term interference, the WVD was computed on the analytic signal of each phase [
10]. From the
resulting energy distribution, ten spectral attributes were extracted, including frequency moments and
power levels at the three most signiĄcant peaks [10].
From the processing via WVD, ten metrics per phase were systematically extracted and organized
into two fundamental categories. First, four central distribution statistics of the marginal spectral
density were calculated, including the mean frequency (
¯
f):
¯
f =
P
f
i
P (f
i
)
P
P (f
i
)
; (Hz) (9)
as the Ąrst spectral moment, and the spectral standard deviation (σ
f
), which quantiĄes spectral dis-
persion:
σ
f
=
s
P
(f
i
−
¯
f )
2
P (f
i
)
P
P (f
i
)
; (Hz) (10)
both expressed in hertz (Hz).
The analysis quantiĄed the spectral envelope using bounding frequencies (f
min
, f
max
), and the three
principal peaks characterized by their center frequency (f
k
) and normalized dB magnitude (p
k
). This
compact parameterization captures both global spectral extent and fault-indicative harmonic features
from the startup transient.
A
p
k
= 10 log
10
P (f
p
k
)
max(P (f))
; (db) (11)
Marot A., y Velázquez S. Comparison of the effectiveness between the Wavelet Transform and the
Wigner-Ville Transform...
32
ISSN-e: 2737-6419
Period: Jan-Mar of 2026
Revista Athenea
Vol.7, Issue 23, (pp. 28Ű40)
This normalization allows comparison between signals with different absolute energy.
D. Statistical Analysis and Attribute Selection
A two-stage pipeline was implemented for feature selection. First, univariate separability between
Healthy (S) and Fault (F) classes was assessed using the Fisher Score for each feature [13]. For a given
feature x, its Fisher Score S
f
is calculated as:
S
f
(x) =
(µ
s
− µ
f
)
2
σ
2
s
+ σ
2
f
(12)
where µ
s
, µ
f
and σ
2
s
, σ
2
f
are the means and variances of x for the Healthy and Fault classes, respectively.
A value of S
f
close to zero indicates total overlap between class distributions, while higher values
indicate greater separability. An empirical threshold of S
f
> 0.01 was established to consider a feature
as potentially discriminative, thus Ąltering out attributes with insigniĄcant separation power.
In the second stage, one-way ANOVA (MATLABŠs anova1 function) was applied for each feature,
testing the null hypothesis of equal means among the three classes: Healthy, 200 kΩ Fault, and 20 kΩ
Fault. Given the exploratory nature of the study and the need to detect subtle effects associated with
incipient faults (200 kΩ), a signiĄcance level of α = 0.1 was adopted, less strict than the conventional
α = 0.05 [
14]. To control the increase in Type I error rate due to multiple comparisons, the Bonferroni
correction was applied. The original p-value obtained for each feature was adjusted according to:
p
adjusted
= min(p × m, 1) (13)
where m is the total number of tested features. Only features with p
adjusted
< 0.1 were considered
statistically signiĄcant for discriminating between at least two of the analyzed operational states.
E. Supervised Classification using Support Vector Machines (SVM)
The automatic diagnosis phase was implemented using a Support Vector Machine (SVM). The
complete pipeline consisted of the following stages.
Preprocessing: All extracted features (both from DWT and WVD) were standardized using Z-score
normalization:
z =
x − µ
σ
(14)
where µ and σ represent the mean and standard deviation of each feature in the training set. This
ensures that all variables contribute equally to the model, regardless of their original scale [
15]. Class
imbalance (120 healthy vs. 240 fault samples, 1:2 ratio) was addressed. The SVM with RBF kernel
performed acceptably without balancing techniques, thanks to the featuresŠ inherent separability, so the
original data distribution was preserved.
Feature Selection: To reduce dimensionality and avoid the curse of dimensionality, a Ąlter based on
the Fisher Score was applied. Only features with a score S
f
> 0.01 were retained, a threshold that
guarantees minimum signiĄcant separability between classes [
16]. Principal Component Analysis (PCA)
was not used to preserve the direct physical interpretability of the wavelet and WVD metrics.
Validation Protocol: A 5-fold cross-validation strategy was implemented on the complete dataset. The
process did not include a Ąxed traditional training/test split; instead, each fold simultaneously served for
training and validation. The averages of accuracy, F1-score, and AUC over the Ąve folds are reported.
SVM Model ConĄguration: A Radial Basis Function (RBF) kernel with MATLABŠs default parameters
was used, without additional hyperparameter optimization. The model was evaluated via 5-fold cross-
validation on the complete dataset.
F. Considerations on Validity and Scope of the Study
While numerical simulation provides controlled conditions for this proof-of-concept, the studyŠs
external validity is limited by model idealizations. Key industrial factors such as electromagnetic noise,
dynamic loads, and environmental aging were not replicated. Therefore, while the results demonstrate
Marot A., y Velázquez S. Comparison of the effectiveness between the Wavelet Transform and the
Wigner-Ville Transform...
33
ISSN-e: 2737-6419
Period: Jan-Mar of 2026
Revista Athenea
Vol.7, Issue 23, (pp. 28Ű40)
the technical superiority of DWT over WVD, validation in real-world environments with actual noise
and variability is required for deĄnitive implementation.
IV. RESULTS
Regarding the analysis of current signals using the Daubechies 10 Wavelet Transform with a de-
composition level of 8, a total of 360 signals were processed: 120 corresponding to the motorŠs healthy
state and 240 associated with low insulation fault conditions. Table
1 details the technical speciĄcations
of the analyzed motors, classiĄed by their nominal voltage levels and power. The external validity of
this study is supp orted by a heterogeneous sample of 20 induction motors, covering a wide operational
spectrum from low-power (10 HP) to high-power (4000 HP) applications across different voltage levels
(Table
1).
Table 1. Technical sp eciĄcations of the simulated induction motor p opulation.
Voltage (V) Frequency (Hz) SpeciĄc Motor Powers (HP) Quantity
220 60 10, 20 2
480 60 100, 150, 175, 200, 215, 300, 355, 400, 500,
600
10
4160 60 175, 830, 900, 1000, 1250, 1630, 2000, 4000 8
Totals 20
Fault patterns thus reĆect insulation degradation independent of motor design, enabling robust
SVM classiĄcation across operational variations. Although Fig. 2 and Fig. 3 reveals subtle energy
differences between states in a 200 HP motor, conclusive detection requires statistical aggregation
beyond individual signal inspection.
Fig. 2. Energy spectrum obtained from the wavelet transform of a 200Hp motor. Healthy state.
The Ągure was generated in MATLAB 2019.
Marot A., y Velázquez S. Comparison of the effectiveness between the Wavelet Transform and the
Wigner-Ville Transform...
34
ISSN-e: 2737-6419
Period: Jan-Mar of 2026
Revista Athenea
Vol.7, Issue 23, (pp. 28Ű40)
Fig. 3. Energy spectrum obtained from the wavelet transform of a 200Hp motor. Faulted state,
unloaded, 200 KΩ. The Ągure was generated in MATLAB 2019.
A statistical analysis was performed in MATLAB to determine the average energy per condition from
the generated database, followed by the calculation of the percentage variation between scenarios. This
global comparison reveals overall trends in signal behavior. Fig.
4 illustrates this percentage variation,
highlighting energy changes due to faults. The most signiĄcant reductions in energy occur in levels
D
1
(−17.1%), D
2
(−20.2%), and notably D
7
(−24.1%), marking them as sensitive indicators for
early anomaly detection. The pronounced effect on D
7
suggests that its high-frequency components
are highly representative of insulation degradation. In contrast, level D
5
shows a moderate increase
(+13.4%), potentially linked to fault-induced spectral redistribution. The remaining levels (D
3
, D
4
,
D
6
, D
8
) exhibit smaller decreases or negligible variations, indicating their lower diagnostic relevance.
Fig. 4. Percentage variation in the energy levels of the wavelet transform from the 360 sampled
signals. The Ągure was generated in MATLAB 2019.
Marot A., y Velázquez S. Comparison of the effectiveness between the Wavelet Transform and the
Wigner-Ville Transform...
35
ISSN-e: 2737-6419
Period: Jan-Mar of 2026
Revista Athenea
Vol.7, Issue 23, (pp. 28Ű40)
Figure 5 compares statistical indicators across wavelet decomposition levels, revealing that levels
D
1
, D
6
, and D
8
exhibit the most signiĄcant variations, indicating high sensitivity to low-severity
insulation faults during motor startup. SpeciĄc changes include pronounced maxima perturbations in
D
1
(−19.2%), D
6
(+22.4%), and D
8
(+49.8%) (Fig.
5a), notable minima shifts in D
2
(−17.6%) and
D
5
(+41.5%) (Fig. 5b), moderate entropy variations in D
6
(−8.1%) and D
7
(+5.5%) (Fig. 5c), and
energy redistributions in D
5
(+9.2% RMS) and D
7
(−10.5% RMS) (Fig. 5d). These Ąndings conĄrm
the wavelet transformŠs efficacy for early fault detection.
Fig. 5. Statistical indicators, per wavelet decomposition level; MAX (a); MIN (b); Entropy (c);
RMS (d). The Ągure was generated in MATLAB 2019.
The same dataset was analyzed using the WignerŰVille Transform (WVD). Figure
6 contrasts a
healthy, unloaded 200 HP motor (top row) with its 200 kΩ fault counterpart (bottom row) via time-
frequency distributions, marginal spectra, and instantaneous frequency proĄles, clearly highlighting
spectral and dynamic differences attributable to insulation degradation.
For the analysis of the WignerŰVille Transform (WVD), the percentage variation of six key spectral
metrics was calculated in MATLAB ( Figure
7), based on the database built from the obtained signal
records. The analyzed metrics were: mean frequency, standard deviation, minimum frequency, maximum
frequency, and the values of the two main peaks (Peak1 and Peak2). This processing allowed for a
precise evaluation of the differences between operating conditions (healthy state, 200 kΩ low insulation
fault, and 20 kΩ fault), providing a detailed characterization of the spectral behavior under different
fault scenarios.
Marot A., y Velázquez S. Comparison of the effectiveness between the Wavelet Transform and the
Wigner-Ville Transform...
36
ISSN-e: 2737-6419
Period: Jan-Mar of 2026
Revista Athenea
Vol.7, Issue 23, (pp. 28Ű40)
Fig. 6. Spectral analysis using the Wigner-Ville Transform. The Ągure was generated in
MATLAB 2019.
Fig. 7. Percentage variation of six spectral metrics from the Wigner-Ville Transform (WVD).
The Ągure was generated in MATLAB 2019.
Frequency sensitivity analysis via the WignerŰVille Transform (WVD) applied to 360 recordsŮevenly
distributed among healthy, low-severity (200 kΩ), and severe (20 kΩ) insulation fault conditionsŮ
revealed pronounced metric variations. Mean sensitivity decreased markedly from 158.18 (healthy)
to 62.716 (200 kΩ) and further to −1008.5 (20 kΩ), indicating progressive spectral sensitivity loss.
Percentage variations were substantial: 771.31% between healthy and moderate fault, 164.18% between
healthy and severe fault, and 160.94% between fault severities. These results conĄrm the high sensitivity
of WVD to insulation degradation, particularly during startup transients, enabling clear discrimination
even at incipient fault stages.
For a systematic comparison, two MATLAB routines were developed for statistical and supervised
processing of features from the WVD and Daubechies 10 wavelet transform. Each routine computed
Marot A., y Velázquez S. Comparison of the effectiveness between the Wavelet Transform and the
Wigner-Ville Transform...
37
ISSN-e: 2737-6419
Period: Jan-Mar of 2026
Revista Athenea
Vol.7, Issue 23, (pp. 28Ű40)
inter-class percentage variation, Fisher Score, ANOVA results (p-value and F -statistic), and the per-
formance of an SVM classiĄer evaluated through cross-validation (5 folds). Results are summarized in
Table
2, which synthesizes the discriminative capacity of each method.
The SVM model yielded the following p erformance (mean ± standard deviation): Accuracy =
74.44 ± 3.2%, F1-score = 0.832 ± 0.028, and AUC = 0.807 ± 0.032. ConĄdence intervals (95%) were
derived assuming normally distributed metrics. This comparative framework provides objective criteria
for selecting the most suitable spectral technique in motor insulation fault diagnosis.
Table 2. Final comparison between Wavelet Daubechies 10 and WignerŰVille Transform
(WVD).
Criterion Wavelet Daubechies 10 WignerŰVille (WVD)
Spectral Sensitivity High in multiple metrics (up to
+354%)
Low to moderate (max.
±9.48%)
Separability (Fisher Score) Moderate to high (up to 0.0170) Low (max. 0.0072)
Statistical SigniĄcance Some metrics with p < 0.1 (e.g.,
Entropy D
6
)
No metrics with p < 0.1
SVM ClassiĄcation Accu-
racy
74.44% 67.78%
SVM ClassiĄcation Ű F1-
score
0.8322 0.7943
SVM ClassiĄcation Ű AUC 0.8072 0.6193
Multiscale Robustness Yes (D
1
ŰD
8
per metric) No (global analysis)
Interpretability High, by levels and speciĄc met-
rics
Low, requires time-
frequency expertise
Applicability in AI Ideal for supervised classiĄcation Limited by low discrimi-
nation
The obtained results demonstrate that the Daubechies 10 Wavelet transform offers greater spectral
sensitivity, better statistical separability, and superior performance in automatic classiĄcation compared
to the WignerŰVille transform. The metrics extracted by the Wavelet transform, especially kurtosis,
entropy, and RMS at high levels (D
4
ŰD
8
), show signiĄcant percentage variations between classes
and Fisher Score values that exceed the useful discrimination threshold (> 0.01). Furthermore, the
SVM model trained with Wavelet features achieves an AUC of 0.8072, indicating a robust capacity to
distinguish between healthy and faulty states.
In contrast, the WignerŰVille metrics show low sensitivity and scarce statistical signiĄcance, which
limits their direct applicability in automatic diagnostic systems. Therefore, it is concluded that the
Daubechies 10 Wavelet transform is more suitable for the diagnosis of insulation faults in electric motors,
both due to its multiscale segmentation capability and its compatibility with artiĄcial intelligence models.
CONCLUSIONS
The study demonstrates that the Discrete Wavelet Transform (DWT) with Daubechies 10 consti-
tutes a signiĄcantly more effective tool than the WignerŰVille Transform (WVD) for the early detection
of low insulation faults in induction motors during the starting transient. This superiority is conĄrmed
by the SVM classiĄcation results, where the DWT achieved 74.44% accuracy compared to 67.78%
obtained by the WVD, thereby validating the stated research hypothesis.
In terms of spectral sensitivity, the wavelet technique exhibited a notably superior discriminative ca-
pacity, with percentage variations of up to +354% in speciĄc metrics, while the WVD showed maximum
variations of only ±9.48%. Particularly, levels D
1
, D
6
, and D
8
of the wavelet decomposition emerged
as the most sensitive indicators for detecting insulation anomalies, registering signiĄcant changes in
Marot A., y Velázquez S. Comparison of the effectiveness between the Wavelet Transform and the
Wigner-Ville Transform...
38
ISSN-e: 2737-6419
Period: Jan-Mar of 2026
Revista Athenea
Vol.7, Issue 23, (pp. 28Ű40)
maximum values (D
8
: +49.8%) and energy (D
7
: −24.1%). The technique demonstrated effective-
ness across different fault severity levels (20 kΩ and 200 kΩ), broadening its applicability to various
operational scenarios.
REFERENCES
[1] M. Samiullah, H. Ali, and A. A. S. Zahoor, ŞFault diagnosis on induction motor using
machine learning and signal processing,Ť in School Elect. Eng. Comput. Sci. (SEECS),
Nat. Univ. Sci. Technol., 2024, pp. 1Ű6.
[2] P. K. Sahoo and A. S. Hati, ŞReview on machine learning algorithm based fault detection
in induction motors,Ť Arch. Comput. Methods Eng., vol. 28, no. 3, pp. 1929Ű1940, 2020.
[3] K. N. Gyftakis and A. J. Marques Cardoso, ŞReliable detection of stator interturn faults
of very low severity level in induction motors,Ť IEEE Trans. Ind. Electron., vol. 68, no. 4,
pp. 3475Ű3484, apr 2021.
[4] R. M. Tallam, S. B. Lee, G. C. Stone, G. B. Kliman, T. G. Habetler, and J. Yoo, ŞA survey
of methods for detection of stator-related faults in induction machines,Ť IEEE Trans. Ind.
Appl., vol. 43, no. 4, pp. 920Ű933, jul 2007.
[5] S. Grubic, J. M. Aller, B. Lu, and T. G. Habetler, ŞA survey on testing and monitoring
methods for stator insulation systems of low-voltage induction machines focusing on turn
insulation problems,Ť IEEE Trans. Ind. Electron., vol. 55, no. 12, pp. 4127Ű4136, dec
2008.
[6] S. Sobhi, M. H. Reshadi, N. Zarift, A. Terheide, and S. Dick, ŞCondition monitoring and
fault detection in small induction motors using machine learning algorithms,Ť Information,
vol. 14, no. 6, p. 329, 2023.
[7] S. Sayedabbas, R. MohammadHossein, N. Zarift, A. Terheide, and S. Dick, ŞCondition
monitoring and fault detection in small induction motors using machine learning algo-
rithms,Ť Information, vol. 14, no. 6, p. 329, 2023.
[8] M. Misiti, Y. Misiti, G. Oppenheim, and J.-M. Poggi, Wavelet Toolbox™ UserŠs Guide.
Natick, MA, USA: MathWorks, 2021.
[9] I. Daubechies, Ten lectures on wavelets. Philadelphia, PA, USA: SIAM, 1992.
[10] Y. Y. Tang, Wavelet theory and its application to pattern recognition. Singapore: World
ScientiĄc, 2000.
[11] L. Cohen, Time-frequency analysis. Englewood Cliffs, NJ, USA: Prentice-Hall, 1995.
[12] P. Nussbaumer, M. A. Vogelsberger, and T. M. Wolbank, ŞInduction machine insulation
health state monitoring based on online switching transient exploitation,Ť IEEE Trans.
Ind. Electron., vol. 62, no. 3, pp. 1835Ű1845, mar 2015.
[13] I. Guyon and A. Elisseeff, ŞAn introduction to variable and feature selection,Ť J. Mach.
Learn. Res., vol. 3, pp. 1157Ű1182, mar 2003.
[14] T. E. Sterne and G. D. Smith, ŞSifting the evidenceŮwhatŠs wrong with signiĄcance
tests?Ť BMJ, vol. 322, no. 7280, pp. 226Ű231, jan 2001.
[15] J. Han, M. Kamber, and J. Pei, Data mining: Concepts and te chniques, 3rd ed. San
Francisco, CA, USA: Morgan Kaufmann, 2011.
Marot A., y Velázquez S. Comparison of the effectiveness between the Wavelet Transform and the
Wigner-Ville Transform...
39
ISSN-e: 2737-6419
Period: Jan-Mar of 2026
Revista Athenea
Vol.7, Issue 23, (pp. 28Ű40)
[16] J. Bi, K. P. Bennett, M. Embrechts, C. M. Breneman, and M. Song, ŞDimensionality
reduction via sparse support vector machines,Ť J. Mach. Learn. Res., vol. 3, pp. 1229Ű
1243, 2003.
AUTHORS
Alfredo Alejandro Marot Guevara is an Electrical/Electronic Engineer,
specializing in Industrial Automation and Computing. With 18 years of
university teaching experience, he is currently a professor at Universidad
de Oriente. He resides in Barcelona, Anzoátegui state, Venezuela.
Sergio Rafael Velásquez Guzmán holds a B.S. and an M.S. in Electronic
Engineering from UNEXPO, and a Doctorate in Education and Engineering
Sciences. he currently leads the Research and Postgraduate Department
at the UNEXPO Vice-Rectorate in Puerto Ordaz.
Marot A., y Velázquez S. Comparison of the effectiveness between the Wavelet Transform and the
Wigner-Ville Transform...
40
