energies

Article
A Method of DC Arc Detection in All-Electric Aircraft
Teng Li 1, *, Zhijie Jiao 1, * , Lina Wang 2 and Yong Mu 3
1 School of Electrical Engineering, Beijing Jiaotong University, Beijing 100044, China
2 School of Automation Science and Electrical Engineering, Beihang University, Beijing 100191, China;
[email protected]
3 Tangshan Power Supply Company of State Grid Jibei Electric Power Co., Ltd., Tangshan 063000, China;
[email protected]
* Correspondence: [email protected] (T.L.); [email protected] (Z.J.); Tel.: +86-15601064191 (T.L.);
+86-13811768182 (Z.J.)




Received: 1 July 2020; Accepted: 10 August 2020; Published: 13 August 2020


Abstract: Arc faults in an aircraft’s power distribution system (PDS) often leads to cable and equipment
damage, which seriously threatens the personal safety of the passengers and pilots. An accurate and
real-time arc fault detection method is needed for the Solid-State Power Controller (SSPC), which is a
key protection equipment in a PDS. In this paper, a new arc detection method is proposed based on
the improved LeNet5 Convolutional Neural Network (CNN) model after a Time–Frequency Analysis
(TFA) of the DC currents was obtained, which makes the arc detection more real-time. The CNN is
proposed to detect the DC arc fault for its advantage in recognizing more time–frequency joint details
in the signals; the new structure also combines the adaptive and multidimensional advantages of
the TFA and image intelligent recognition. It is confirmed by experimental data that the combined
TFA–CNN can distinguish arc faults accurately when the whole training database has been repeatedly
trained 3 to 5 times. For the TFA, two kinds of methods were compared, the Short-Time Fourier
Transform (STFT) and Discrete Wavelet Transform (DWT). The results show that DWT is more
suitable for DC arc fault detection. The experimental results demonstrated the effectiveness of the
proposed method.

Keywords: arc fault detection; discrete wavelet transform; convolutional neural network; more electric
aircraft; time–frequency analysis

1. Introduction
An arc fault in the power distribution system (PDS) of an all-electric aircraft (AEA) is difficult
to be traced since it is a transient phenomenon [1]. Furthermore, many arc faults only occur when
the aircraft is flying, accompanying the severe vibration of the frame, the high stress on the cable,
the high altitude, etc. These environmental factors lead to the insulation being declined or the cable
connection loosened, which make it easier to cause arc faults. The arc discharge is also an erratic
phenomenon affected by different power electronic loads and controllers [2]. Different environment
makes these arc faults hard to reproduce during the maintenance after the aircraft has landed [3].
Therefore, it is necessary to obtain an accurate and real-time arc detection method. According to the
cause and location of the arc, aircraft arc faults can be divided into parallel arc faults, ground arc faults,
and series arc faults. Caused by the insulation destruction of the cables, parallel arc fault refers to the
short-circuit arc faults developed between cables, while ground arc faults regularly occur between the
cable and chassis [4]. Series arc faults, generated between the loosened contacts, often appears at the
connections of the plug or series load, and the arc can be regarded as an extra load [5]. When it occurs
at the PDS, it is difficult to be detected because the circuit current is very small. Moreover, conventional

Energies 2020, 13, 4190; doi:10.3390/en13164190 www.mdpi.com/journal/energies

Energies 2020, 11, x; doi: FOR PEER REVIEW 2 of 15

Energies 2020, 11, x; doi: FOR PEER REVIEW 2 of 15

off, might
Energies introduce
2020, 13, 4190 a series arc fault during the over-current circuit break [6]. Therefore, it is
2 ofnot
14
suitable
off, mightforintroduce
the PDS protection.
a series arc fault during the over-current circuit break [6]. Therefore, it is not
A Solid-State Power Controller (SSPC), an intelligent power controller with a protection
suitable for the PDS protection.
circuit
function,breakers, which
has been are unsuited
widely to switch
the to a low-current circuit off, might introduce a series arc
A Solid-State Power used to protect
Controller (SSPC), PDS
an [7]. The PDS,
intelligent including
power a main
controller Directa Current(DC)
with protection
fault
bus during
and a the
seriesover-current
of SSPCs, is circuit
shown break
in [6].
Figure Therefore,
1; it shows it is
the not suitable
aircraft load for the
and PDS
the protection.
power supply are
function, has been widely used to protect the PDS [7]. The PDS, including a main Direct Current(DC)
A Solid-State
connected under Power
the Controller
control of (SSPC),
the SSPC. anAnintelligent
SSPC is power
used controller
to break with
the a protection
circuit in more function,
critical
bus and a series of SSPCs, is shown in Figure 1; it shows the aircraft load and the power supply are
has been
conditions widely
sinceused
it is to
moreprotect the
reliable PDS
and [7].
fasterThe
to PDS,
switch including
the power a main
off Direct
[8]. Current(DC)
Although an SSPC bus
is and to
able a
connected under the control of the SSPC. An SSPC is used to break the circuit in more critical
series of SSPCs,
protect thesince is
circuit shown
from in
many Figure
faults,1; it shows the aircraft load and the power supply are connected
conditions it is more reliable andsuch as to
faster from shortthe
switch circuits,
poweroverloads, and some
off [8]. Although arc faults
an SSPC [9],toit
is able
under
is hard thetocontrol
protect ofthe
the cables
SSPC. An fromSSPC
a is used
series of to break
arc faultthe
[6].circuit
Thus, inamore criticalmethod
detection conditions
for since itarc
series is
protect the circuit from many faults, such as from short circuits, overloads, and some arc faults [9], it
more reliable and faster to switch the power off [8]. Although an SSPC is able to protect the circuit
isfaults
hardistoneeded.
protect the cables from a series of arc fault [6]. Thus, a detection method for series arc
from many faults, such as from short circuits, overloads, and some arc faults [9], it is hard to protect
faults is needed.
the cables from a series of arc fault [6]. Thus, a detection method for series arc faults is needed.

Figure 1. A Solid-State Power Controller (SSPC)-based aircraft power distribution system (PDS).
Figure 1. A Solid-State Power Controller (SSPC)-based aircraft power distribution system (PDS).
Figure 1. A Solid-State Power Controller (SSPC)-based aircraft power distribution system (PDS).
The structure of the SSPC is shown in Figure 2. Unlike traditional breakers, MOSFETs will not
The structure of the SSPC is shown in Figure 2. Unlike traditional breakers, MOSFETs will
introduce arcs when
The structure switching the circuit off [3]. Therefore, an SSPC has MOSFETs
the advantages of
not introduce arcs of the
when SSPC is shown
switching in Figure
the circuit off2.[3].
Unlike traditional
Therefore, breakers,
an SSPC will not
has the advantages of
compactness,
introduce high
arcshigh
when reliability,
switchingandthefastcircuit
controloff
by[3].
cutting off the power supply within 25 ms [3].ofA
compactness, reliability, and fast control by cuttingTherefore, an supply
off the power SSPC has the
within advantages
25 ms [3]. A DC
DC arc fault
compactness, detection algorithm
high reliability, can be integrated in the SSPC. According to the diagnosis results of
arc fault detection algorithm and fastintegrated
can be control byincutting
the SSPC.off the power supply
According within 25 results
to the diagnosis ms [3]. of
A
the
DC voltage or current of the main circuit measured by the voltage/current measurement modules,
the arc fault or
voltage detection
currentalgorithm
of the main cancircuit
be integrated
measured in the SSPC.
by the According to measurement
voltage/current the diagnosis results
modules, of
the
the algorithm generates
voltage orgenerates driving
current ofdriving
the main signals
circuit to disconnect
measured by the MOSFETs with the power supply when
the algorithm signals to disconnect thethe voltage/current
MOSFETs with the measurement
power supply modules,
when the
the algorithm
the DC arc is detected.
generates driving signals to disconnect the MOSFETs with the power supply when
DC arc is detected.
the DC arc is detected.
Power Source

Power Source
Main SSPC
Circuit
Main Driving SSPC
Logic
Control External
MOSFET
Circuit Signal Circuit
Control Signal
Driving Logic Control
and Control External
MOSFET Signal Short
Circuit
Interfaces Control Signal
Circuit Control
Control
Short （CPU）
and Internal
Voltage Circuit Interfaces Power
Sampling （CPU） Supplier
Internal
Current Control
Resistance Power
Measurement
Voltage
Sampling Supplier
Current
Resistance
Measurement

Load

Load Figure 2. The structure of the SSPC.

Figure 2. The structure of the SSPC.
Researchers in various fields Figurehave been
2. Thestudying
structure arc fault
of the detection, such as renewable energy
SSPC.
power generation [10], electric vehicles [11], DC power distribution systems [12], aerospace [3], etc.
According to the characteristics of the current flowing through the arc, arc faults can be divided into
Energies
DC arc 2020,
faults11,and
x; doi: FOR PEER REVIEW
Alternating www.mdpi.com/journal/energies
Current(AC) arc faults [4]. Two kinds of AC arc fault detection methods
Energies 2020, 11, x; doi: FOR PEER REVIEW www.mdpi.com/journal/energies
Energies 2020, 13, 4190 3 of 14

based on an SSPC in aerospace have been presented [3]. Unlike the AC current, it is not easy to
extinguish the DC arc since it has no zero-crossing point [13], so it is necessary to develop a DC arc
detection method that can be integrated into the SSPC. For the DC arc detection, the parallel and
ground arcs cause large load current surges, which can be more easily detected by the overcurrent
protection. It is more difficult to detect series arcs since it decreases the load current [2]; there remains
to be this lack in the methods to obtain high-accuracy arc fault detection and diagnosis for series DC
arc faults in aircrafts.
Various detection methods have been studied in the past. The proposed detection methods
can be divided into the physical feature methods, including the heat, sound, and light detection
methods [14–16]; time domain feature methods [13,17,18]; frequency domain feature methods [19–21];
time-to-frequency analysis methods [12–14,19]; artificial intelligence methods [9,13,15,20]; the arc
current entropy method [22,23]; and the blind-source separation method [24]. The study of time
domain and frequency domain features based on the arc voltage or current are reliable compared to
the physical feature methods, where the detection accuracy is sensitive to the position of the sensor
installation [25]. Furthermore, as aforementioned, it can realize the real-time detection within dozens of
milliseconds. However, the detection accuracy is usually below 90% since most of them only extracted
some features of the signals as the arc fault detection criteria, such as the amplitude or the variance of
the wavelet coefficient of the arc current [18]. The accuracy is low by using some criteria purely in
certain domains due to the limited characteristic information of the arc current or voltage, even though
multiple criteria was adopted [17,26].
Time–Frequency Analysis (TFA) could present the time and frequency local features at the same
time, or the mapping relationship of the features in the two domains. Therefore, with the TFA it is
easier to distinguish the arc signal from the normal signal [27]. It shows that the Short-Time Fourier
Transform (STFT) with a Hamming window behaves much better than the real-time Fast Fourier
Transform(FFT) for getting a more obvious separation between the normal and fault detection variable
values [18]. Of course, data fusion from multiple detection methods could further improve the accuracy.
For example, the time-domain fluctuation and STFT analysis variables are fused to give a high accuracy
above 95.9% [18]. The Discrete Wavelet Transform (DWT) has also been applied [28]. It has a relatively
higher accuracy compared to the STFT; however, with a lower computation efficiency [18].
Due to the nonstationary and nonlinearity of the arc signals, artificial intelligent methods could
further improve the arc detection accuracy. A series of machine learning methods have been applied to
perform the arc detection, including the support vector machine (SVM) [29], extremely randomized
trees (ET), random forest (RF), K-nearest neighbor (KNN), Gaussian Naive Bayes (GNB), hidden
Markov model (HMM) [30], etc. Each method has its own advantage and disadvantage, such as there
is no need for the KNN to be trained compared to the SVM; however, SVM has a better detection
accuracy [17,29]. Ensemble machine learning (EML) has been performed to achieve a higher accuracy
close to 100%. Three ensemble techniques, including bagging, boosting, and stacking, were studied.
It showed that the stacking ensemble algorithm maintains superior performance compared to other
EML algorithms [17].
Even though the EML has a high accuracy, however, it heavily relies on the ensemble method and
integrated models. It shows that the bagging technique that used SVM as a base learner (E-SVM) only
achieved a 98.5% accuracy using 3000 training datasets [17]. In contrast, deep learning can automatically
extract and choose the features. A deep neural network (DNN) with Fourier coefficients, Mel cepstrum
data, and wavelet features as inputs were used to identify the DC arc faults [31]. Comparing with
other deep learning methods, there are problems, such as the expansion of the number of parameters
and overfitting for reaching a high accuracy, and it is impossible to model the changes in the time
series [32]. Convolution Neural Network (CNN) has shared convolution kernels; there is no pressure
for processing the high dimensional data by using CNN compared to DNN [32]. A domain adaptation
(DA) combined with a deep convolutional generative adversarial network (DCGAN) was proposed to
diagnose the DC series arc faults in a changed scene, where the lightweight CNN classifier was used
Energies 2020, 13, 4190 4 of 14

offline firstly for training the network and then online for the arc fault diagnosis. However, normal and
arc data in the time domain were packed to form the 2D image as the input of the CNN other than the
TFA generation [33]. A new DC arc fault detection algorithm based on a Convolution Neural Network
(CNN) is proposed here. CNN is used to distinguish the arc fault by automatically analyzing the
time–frequency spectrum of the circuit current, which is obtained from the Time–Frequency Analysis
(TFA) of the current signal with full detail representation. There is no need to the extract and choose
the features manually as with some machine learning methods. This method has been applied for
detecting other faults [32]. However, it is the first time its feasibility and accuracy are explored for
detecting a series DC arc in aircraft. The TFA generation methods of the STFT and the DWT were
used. Combined with an improved CNN, the experimental results bearing the arc faults show a high
accuracy for a series arc fault diagnosis, which is close to a 100% accurate identification after the CNN
is trained by the TFA of 104 arc signals. At the same time, the diagnosis time is also fast, which is
around 1 s.

2. Time–Frequency Analysis of the Current Signal

Compared to pure time-domain or frequency-domain representation, the TFA includes information
in both domains for the DC arc. Moreover, there is no need to extract the arc characteristics anymore;
considering the arc is easily disturbed by the noise, the TFA can separate the detected signal from
the surrounding noise intuitively [9]. Therefore, the signal information is more complete without
discarding some details.
Nowadays, a great number of TFA methods have been developed, including the Empirical Mode
Decomposition (EMD) [13], Wigner–Ville Distribution (WVD) [15], and Discrete Wavelet Transform
(DWT) [20]. However, bilinear time–frequency distribution methods represented by WVD are used less
due to the inherent cross-interference items, and the EMD methods are susceptible to mode aliasing [26].
DWT has been paid lots of attention to in the multiresolution analysis because it can identify the
component of a signal on a multi-scale and has been applied in the field of arc fault detection [28].
Here, we will discuss the DWT method by comparing two kinds of wavelet bases.
In the general wavelet basis, the dependent variables of the time–frequency spectrum are the
scale coefficient and the wavelet coefficient, while the dependent variable is the wavelet coefficient in
the orthogonal wavelet basis [34]. As a result, the spectrum of the orthogonal wavelet generation is
selected here to reduce the impact factors of the spectrum generation.
In DWT, scale a and translation b are discrete, and the input signal is continuous [35]. Thus,
the discrete current signal in the database should be reconstructed into a continuous signal before the
DWT. The reconstructive function is

i sin[π · fs · (t − fs )]
n
X i
f (t) = f ( )· (1)
i=1
fs π · fs · (t − fis )

where n is the number of sampling points for each current data; fs is the sampling frequency; t is the
time variable; f (i/fs ) is the value of discrete sampling signal; and f (t) is the reconstructed function.
fs must satisfy the Shannon sampling theorem, so that there will be no spectrum aliasing when the
signal is reconstructed. The wavelet coefficients of the DWT are calculated accordingly:
Z +∞
W f (a, b) = W f (2−j , 2−j k) = f (t)ψ∗ j,k (t)dt j, k ∈ Z (2)
−∞

where Ψ* j,k (t) means the conjugate of Ψj,k (t); Wf (2−j , 2−j k) is the wavelet coefficient in two dimensions;
a is 2−j and b is 2−j k; and Ψj,k (t) is the orthonormal basis generated by the orthogonal wavelet Ψ(t),
which is shown as follows:
1 t − 2−j k
ψ j,k (t) = √ ψ( ) j, k ∈ Z (3)
2− j 2− j
Energies 2020, 13, 4190 5 of 14

where Ψ[(t−2−j k)/2−j ] is the result of Ψ(t) after t is magnified by dividing 2−j firstly and then shifted to
the right by 2−j k.
In Equation (2), the wavelet coefficient is a measurement of the similarity between the wavelet
basis and the original signal Ψj,k (t) [35]. In other words, Equation (2) only expresses the relationship
between the wavelet coefficient and the relative frequency of the signal on the wavelet basis, but not
the relationship between the wavelet coefficient and the actual frequency of the signal.
The actual frequency f a can be calculated by the wavelet center frequency f c and the sampling
frequency f s , which is shown as follows:
fc
fa = fs (4)
c·a
where ca is the scale sequence and c is a constant. The constant c should be determined to get f a .
According to Shannon sampling theorem, f a is in the range of (0, fs /2), thus the corresponding range
of the scale sequence ca is (2fc , ∞). Suppose s is the length of ca. When the sampling frequency and
wavelet basis are fixed, the value of f a depends on the value of ca, and the smaller ca is, the larger f a is.
The maximum value fs /2 of f a corresponds to the minimum value c/s of ca. Thus, it is easy to obtain
that the value of c is 2 × fc × s.
1
ta = n (5)
fs
The time-independent variable sequence ta that only depends on f s and n can be calculated
by Equation (5), where ta is the time corresponding to certain sampled point and n is the certain
sampling points. Derived from Equation (1) to Equation (5), the time–frequency spectrum is established
according to the wavelet coefficients at different times and frequencies derived from the current signals,
as shown in Figure 3. When n is fixed, the resolution of the time–frequency spectrum depends on scale
a, in which the larger the j, the higher the resolution. On the other hand, the amount of calculation
is positively related to j. Therefore, the value of j cannot be set, neither as infinity nor infinitesimal.
Through a large number of tests, it is more appropriate that the value of j is 8, so that the length of m
is 256.
From Equation (2), the wavelet coefficients merely depend on the reconstructive signal f (t) when
the wavelet basis is constant, and the wavelet coefficients of the same signal will be different under
different wavelet bases. The commonly used wavelet basis in the field of arc detection is db4, db10,
and bior1.5 [4]. Since bior1.5 is not an orthogonal wavelet, db4 and db10 are only discussed here as the
wavelet bases of DWT.
One fault datum is selected randomly from the database, and the time–frequency spectra with
db4 and db10 are formed, respectively, to determine the optimal wavelet basis. The results can be seen
from Figure 3a,b, corresponding, respectively, to the outcome with db10 and db4.
From Figure 3a,b, the maximum value of the spectrum appears at around 2.5 s in Part 1, and at this
moment, it can be observed on the experimental platform that there is an obvious arc between the two
electrodes. There are some other peaks in Part 3, shown in the spectrum after the main peak in Part 1,
which belong to the arc nuisance noise. However, the peaks in the low frequency bands shown in Part
2 are the artefacts caused by the wavelet transform. Figure 3c is the 2D form of Figure 3b, and it shows
that the peaks in Part 2 are at the edges of the image, which do not affect the recognition of the arc fault.
Compare Figure 3a,b: it indicates the wavelet basis of db4 has a better performance since the peak
in Part 1 has a larger value. As a consequence, the db4 wavelet is more suitable for the representation
of the DC arc when the DWT is applied.
The normal data, which were measured in the same environment as the fault data in Figure 3b,
was selected from the database, and its time–frequency spectrum is shown in Figure 4. It can be
concluded that the time–frequency spectrum of the normal data is flatter than the spectrum presented
in Figure 3b based on the db4 wavelet basis, which means the magnitude of the normal current is far
less than the arc current and the normal current fluctuates uniformly. Furthermore, the influence of the
db4 wavelet basis on the normal data is not obvious.
Energies 2020, 13, 4190 6 of 14
Energies 2020, 11, x; doi: FOR PEER REVIEW 6 of 15

(a)

(b)

Energies 2020, 11, x; doi: FOR PEER REVIEW 7 of 15

concluded that the time–frequency spectrum of (c) the normal data is flatter than the spectrum
presented
Figure
Figure3.in3. Figure 3b based spectrum
Time–frequency
Time–frequency on the db4
spectrum ofwavelet
of the the arcbasis, which
current
arc current based means
based
on the the
on the
db10 magnitude
anddb10 and db4
db4 wavelet:of(a)the normal
wavelet:
time–
current is far
(a) frequency less
time–frequency than
spectrum the arc
spectrum current and
of the arcofcurrent
the arc the normal
current
based based
on db10; current
(b) on fluctuates uniformly.
db10; (b) time-frequency
time-frequency Furthermore,
arc currentofthe
spectrum of the spectrum
influence
thebased ofonthe
arc current db4
db4; (c)wavelet
based on db4;basis
two-dimensional onspectrum
the normal
(c) two-dimensional data
of the arcis notofobvious.
current
spectrum based
the arcon db4. based on db4.
current

From Figure 3a,b, the maximum value of the spectrum appears at around 2.5 s in Part 1, and at
this moment, it can be observed on the experimental platform that there is an obvious arc between
the two electrodes. There are some other peaks in Part 3, shown in the spectrum after the main peak
in Part 1, which belong to the arc nuisance noise. However, the peaks in the low frequency bands
shown in Part 2 are the artefacts caused by the wavelet transform. Figure 3c is the 2D form of Figure
3b, and it shows that the peaks in Part 2 are at the edges of the image, which do not affect the
recognition of the arc fault.
Compare Figure 3a,b: it indicates the wavelet basis of db4 has a better performance since the
peak in Part 1 has a larger value. As a consequence, the db4 wavelet is more suitable for the
representation of the DC arc when the DWT is applied.
The normal data, which were measured in the same environment as the fault data in Figure 3b,
was selected from the database, and its time–frequency spectrum is shown in Figure 4. It can be

Energies 2020, 11, x; Figure

doi: FOR4.PEER REVIEW
Time–frequency www.mdpi.com/journal/energies
spectrum of the normal current based on db4.
Figure 4. Time–frequency spectrum of the normal current based on db4.

3. Construction and Training of the CNN

CNN is a feedforward neural network, which has been successfully applied to image
classification [16], face recognition [34], semantic segmentation [35], and object detection [36]. CNN
can be viewed as a composite model that is supervised. At the internal of the CNN, each hidden
Energies 2020, 13, 4190 7 of 14
Figure 4. Time–frequency spectrum of the normal current based on db4.

3.
3. Construction
Construction and
and Training of the
Training of the CNN
CNN
CNN
CNN is is aa feedforward
feedforward neural neural network,
network, which which has been successfully
has been successfully applied
applied to image
to image
classification [16], face recognition [34], semantic segmentation [35], and
classification [16], face recognition [34], semantic segmentation [35], and object detection [36]. CNNobject detection [36]. CNNcan
can be viewed as a composite model that is supervised. At the internal
be viewed as a composite model that is supervised. At the internal of the CNN, each hidden layer of the CNN, each hidden
layer
servesserves
as the as the visible
visible layer for layer
the for
nextthe next hidden
hidden layer [37].layerThe[37]. The connection
connection betweenbetween the two
the two adjacent
adjacent layers is sparse, except for the fully connected layer. Connections
layers is sparse, except for the fully connected layer. Connections between the neurons in the same between the neurons in
the
layersame layer
are not are notThe
allowed. allowed.
neuronsTheof neurons
each layer ofcalculate
each layer thecalculate
output ofthe theoutput of thethrough
upper layer upper layer
their
weight-sharing, which reduces the complexity of the network and the number of weights number
through their weight-sharing, which reduces the complexity of the network and the [37]. Thus,of
weights [37]. Thus, the calculation
the calculation speed is efficient. speed is efficient.
Many
Many CNN structures have
CNN structures have been
been proposed,
proposed, including
including somesome classical
classical structures:
structures: LeNet5
LeNet5 [36],[36],
AlexNet [37], and VGG [38]. AlexNet and VGG have a complicated architecture in which numerous
AlexNet [37], and VGG [38]. AlexNet and VGG have a complicated architecture in which numerous
parameters need
parameters needtotobebe iteratively
iteratively calculated.
calculated. Hence,Hence, lots
lots of of samples
samples are demanded
are demanded to train tothetrain the
network
network before the parameter identification. Besides, numerous parameters
before the parameter identification. Besides, numerous parameters lead to an increased computation lead to an increased
computation
and extendedand extended
calculation calculation
time. time.
As a result, As a result,
a complex CNN a complex
structureCNN is notstructure
suitable foris not suitable for
arc detection in
arc detection in SSPC. In this paper, a CNN structure is based on
SSPC. In this paper, a CNN structure is based on LeNet5, which usually includes two convolution LeNet5, which usually includes
two convolution
layers, two pooling layers,
layers,twoa full-connection
pooling layers, alayer,full-connection
and an output layer, andHowever,
layer. an outputone layer. However,
pooling layer
one pooling layer is added here as the first layer, to downsize the
is added here as the first layer, to downsize the sample from a 2000 × 256 TFA image to 100 sample from a 2000 × 256×TFA128.
image to 100 × 128. Therefore, the computation time is decreased greatly.
Therefore, the computation time is decreased greatly. The structure is shown in Figure 5 and the relatedThe structure is shown in
Figure 5 and the related
parameters are shown in Table 1. parameters are shown in Table 1 .

Figure 5. The structure of the CNN.

Energies 2020, 11, x; doi: FOR PEER REVIEW www.mdpi.com/journal/energies
Table 1. CNN parameters.

Parameters Value
The size of input vector 256 × 2000
The number of input channels 1
The number of feature maps 2
Filter size 63 × 35, 18 × 18
Pooling size 2 × 20, 2 × 2, 2 × 2
Activation function Sigmoid

Two feature maps of the normal and arc data were used at the same time for extracting the
information. The flow chart of the algorithm is shown in Figure 6. The circuit current is transformed
into a time–frequency spectrum as input for the CNN. The result of the CNN indicates whether the arc
fault occurs in the circuit.
The load in the circuit was changed while the other conditions were kept constant. The normal
current and arc current were measured at the same circumstance where the normal current was
measured many times due to the different load each time. There were a total of 104 signals. The normal
current data take the proportion of 25.96% in the database while the arc current data take 74.04%.
There are only two ideal results of the CNN: 1 and 0, where 1 stands for the arc fault condition and 0
 The number of input channels 1
The number of feature maps 2
Filter size 63 × 35, 18 × 18
Pooling size 2 × 20, 2 × 2, 2 × 2
Energies 2020, 13, 4190 Activation function Sigmoid 8 of 14

Two feature maps of the normal and arc data were used at the same time for extracting the
represents the normal condition. We extracted 5 fault signals and 5 normal signals randomly from
information. The flow chart of the algorithm is shown in Figure 6. The circuit current is transformed
the database as the test database, and the other 94 signals from the training database. In order for the
into a time–frequency spectrum as input for the CNN. The result of the CNN indicates whether the
CNN to learn sufficiently, the time–frequency spectrum of the training data from the training database
arc fault occurs in the circuit.
had been put into the CNN several times.

Figure 6. The flow chart of the algorithm.

Figure 6. The flow chart of the algorithm.
4. Results and Discussion
The load in the circuit was changed while the other conditions were kept constant. The normal
4.1. Experimental
current and arc current Results were measured at the same circumstance where the normal current was
measured many the
To simulate times
DCdue to arc
series the fault
different
in theload
AEA,each
a DCtime. There
series wereexperiment
arc fault a total of platform
104 signals.
wasThe
set
normal
up, current
as shown in data
Figuretake
7. the
Theproportion
experimental of 25.96%
platforminconsisted
the database while
of a DC the arc
source, current
switch, data take
voltage and
74.04%. measurement
current There are only two adjustable
circuit, ideal results of and
load, the CNN: 1 and 0, which
arc generators, wherewere1 stands for theaccording
established arc fault
Energies 2020, 11,
condition andx; doi:
0 FOR PEER REVIEW
represents the normal condition. We extracted 5 fault signals and 5 normal 9 of 15
signals
to the UL1699B (Underwriters Laboratories Standard for Safety for Arc-Fault Circuit Interrupters)
randomly
standard
Energies 2020,
from
and
11, is
the database
illustrated
x; doi: FOR PEER
as the test
inREVIEW
Figure 8. database, and the other 94 signals from the training database. 9 of 15
In order for the CNN to learn sufficiently, the time–frequency spectrum of the training data from the
training database had been put into the CNN several times.

4. Results and Discussion

4.1. Experimental Results

To simulate the DC series arc fault in the AEA, a DC series arc fault experiment platform was
set up, as shown in Figure 7. The experimental platform consisted of a DC source, switch, voltage
and current measurement circuit, adjustable load, and arc generators, which were established
according to the UL1699B (Underwriters Laboratories Standard for Safety for Arc-Fault Circuit
Interrupters) standard and is illustrated in Figure 8.

Figure 7. Schematic diagram of the DC series arc fault experimental platform.

Figure
Figure 7. Schematic
Schematic diagram
diagram of
of the
the DC
DC series
series arc
arc fault
fault experimental platform.

To power supply Stationary

Energies 2020, 11, x; doi: FOR PEER REVIEW www.mdpi.com/journal/energies
To power supply electrode
Stationary Moving electrode To load
electrode Moving electrode To load

Insulating Lateral
Insulating
clamps Lateral
adjustment
clamps adjustment
Sliding block
Sliding block
Fixed bace
Fixed bace

Figure 8. The structure diagram of the arc generator.

Figure 8. The structure diagram of the arc generator.
Figure 8. The structure diagram of the arc generator.

InIn
thetheexperiment,
experiment,the
thestationary
stationary electrode
electrode and moving electrode
and moving electrodewere
wereseparated
separated gradually
gradually so so
that
that the arc generator could produce a distinct arc, which was beneficial to analyze the features of of
the arc generator could produce a distinct arc, which was beneficial to analyze the features
thethe
DCDC arc.
arc.The
Themechanism
mechanismofofsimulating
simulating the
the arc in the
arc in the laboratory
laboratoryand
andthe
theactual
actualarc
arcinin aircraft
aircraft is is
similar, which
similar, whichisisa aprocess
processofofdischarge
discharge after
after the air breakdown
the air breakdown[5].[5].The
Theelectrodes
electrodeswere
were polished
polished
before the
before theexperiment
experimentbecause
becausethetheburning
burning arc
arc might lead to
might lead tothe
theelectrodes
electrodesdissolving.
dissolving. The
The normal
normal
Energies 2020, 13, 4190 9 of 14

In the experiment, the stationary electrode and moving electrode were separated gradually so
that the arc generator could produce a distinct arc, which was beneficial to analyze the features of the
DC arc. The mechanism of simulating the arc in the laboratory and the actual arc in aircraft is similar,
which is a process of discharge after the air breakdown [5]. The electrodes were polished before the
experiment because the burning arc might lead to the electrodes dissolving. The normal current data
and arc current data with different loads were collected by voltage and current measurement circuits,
constituting the database.
An oscilloscope with a high sampling rate of 400 MHz (2000 points in 5 µs) was used to record the
experimental data. Waveforms of an arc current and a normal current are shown together in Figure 9.
Figure 10 shows the Fourier frequency spectrum of the arc current.
Figure 9 illustrates the contrast between the arc current and normal current in the time domain.
It can be seen that the normal current fluctuates near 0, which is far less than the arc current, and the
time when the maximum magnitude of the arc current appears is not fixed due to the nonstationary
form of the arc. From Figure 10, we can see the frequency of the arc current is broadband and
random. As a result, it is not an excellent idea to choose the maximum, variance, and mean amplitude
Energies
in 2020, 11, x; doi:
the frequency FOR PEER
domain by REVIEW 10 of
Fourier transform as the criteria to recognize the arc fault, due to the15
Energies 2020, 11, x; doi: FOR PEER REVIEW 10 of 15
arc’s randomness.

Figure 9. Current waveform in normal and arc fault conditions.

Figure9.9.Current
Figure Currentwaveform
waveformininnormal
normaland
andarc
arcfault
fault conditions.
conditions.

Figure 10.
Figure 10. Fourier frequency spectrum of
of the
the arc
arc current.
current.
Figure 10. Fourier frequency spectrum of the arc current.
4.2.
4.2. Assessment
Assessment of
of Detection
Detection Accuracy
Accuracy
4.2. Assessment of Detection Accuracy
Each
Each output
output is
is recorded
recorded when
when the
the data
data in
in the
the training
training database
database are
are inputted
inputted to
to the
the CNN,
CNN, and
and aa
scatter Each output
diagram of is recorded
the results when
was the data
plotted, as in the in
shown training
Figuredatabase
11. are inputted to the CNN, and a
scatter diagram of the results was plotted, as shown in Figure 11.
scatter diagram
As seen of the11,
in Figure results wasall
almost plotted,
of the as shown
results in Figure
gather 11. the value of 0.5 at the beginning.
around
With the amount of training data increasing, the fault data and normal data were separated gradually
from each other and the fault data got close to 1, while the normal data clustered together around 0.
The green line and the yellow line show the developing tendency of the results, and that the accuracy
ratio of the judgement will be improved as more training data are inputted into the CNN.
In order to separate the normal and fault data, the threshold of the result was set to 0.5. When the
output result is greater than 0.5, the input data is considered fault data, and the others thus considered
Energies 2020, 13, 4190 10 of 14
Figure 9. Current waveform in normal and arc fault conditions.

normal data. In this paper, the True Positive Rate (TPR) and False Positive Rate (FPR) were introduced
to assess the accuracy of the method.

TP
TPR = (6)
TP + TN
FP
FPR = (7)
FP + FN
where TP is the number of normal signals correctly identified; TN is the number of normal signals
incorrectly identified; FP is the number of fault signals correctly identified; and FN is the number of
fault signals incorrectly identified. Figure
Table10.2 Fourier frequency
shows the spectrum
assessment of thefor
results arcCNN.
current.
“Epoch” refers to
the number of times the whole database had been repeatedly trained. Although each training uses the
4.2. Assessment
same database, different of Detection
training Accuracy
times has different effects on the model weight updating because the
loss function isEach
different when using different
output is recorded when the training
data times
in the [39]. The database
training larger theare
epoch is, the to
inputted closer the and a
the CNN,
output ofscatter
the CNN is to the
diagram ideal
of the value,
results wasand the smaller
plotted, the in
as shown value of the
Figure 11. loss function is. However,
epoch cannot increase all the time. When it exceeds a certain value, the CNN will be overfitting.

Training
Figure 11.Figure 11. results of results
Training the CNN.of the CNN.

Table 2. CNN assessment results.

As seen in Figure 11, almost all of the results gather around the value of 0.5 at the beginning.
WithEpoch
the amount
TP ofTNtraining
FP data
FN increasing,
TP + TN the fault
FP + FN data
TPRand normal
FPR data were separated
gradually
1 from8each other
19 and
71 the6fault data got close to 1, while the normal
29.63% 92.21%data clustered together
around20. The 27
green line
0 and
77 the yellow
0 line show the developing
100% tendency
100% of the results, and that
the accuracy
3 ratio
27 of 0the judgement
77 0 will be27 improved 77 as more training
100% data are inputted into the
100%
CNN. 4 27 0 77 0 100% 100%
5 27 0 77 0 100% 100%
6 Overfitting

Energies 2020, 11, x; doi: FOR PEER REVIEW www.mdpi.com/journal/energies

From Table 2, when the epoch = 1, the TPR has a low value of 29.63%, and the value of the FPR
(92.21%) is also a little bit low for arc fault detection in aircraft. When the epoch = 6, the CNN is
overfitting, so the case of an epoch ≥ 6 is not considered. When the epoch = 2, 3, 4, or 5, the TRP and
FPR are both 100%. However, it can be seen from Figure 11 that the fault data were separated from the
normal data when the network was trained the second round, i.e., signals from No. 105 to 208 were
used; however, the scatter of the data is still relatively large, which means a large identification error.
Considering that the amount of experimental data is limited, a certain margin should be reserved in a
practical application. As a result, the value of the training time from 3 to 5 is more appropriate for
the epoch.
Energies 2020, 13, 4190 11 of 14

4.3. Verification and Comparison of the TFA Plus CNN Method

Data in test database were used to judge the detection accuracy of the algorithm. As per the flow
chart shown in Figure 6, the test data are input into the CNN that has been trained and the results
are displayed in Table 3. The fault data and the normal data in the test database have been separated,
evidently. The minimum of the fault data results is 0.9988, while the maximum of the normal data
results is 0.1356. Some fault data results are 1, which is the ideal output, while the results of the normal
data are close to 0. Thus, TFA + CNN is an effective method to detect the DC arc.

Table 3. Test results of the two measures.

Signal Number Ideal Results Results of DWT + CNN

1 1 1
2 1 1
3 1 0.9988
4 1 1
5 1 1
6 0 0.1311
7 0 0.1322
8 0 0.1315
9 0 0.1347
10 0 0.1356

Not only can DWT produce a time–frequency spectrum, but Short-Time Fourier Transform (STFT)
is another suitable method for TFA.
A comparison between STFT and DWT was performed, and the results are shown in Table 4 [40].
At the same level of sampling frequency, the DWT has a lower computational complexity and spends
less time than the STFT in the DC arc detection field. From this angle, DWT is more appropriate for
DC arc fault detection in an aircraft.

Table 4. Comparison between Short-Time Fourier Transform (STFT) and Discrete Wavelet
Transform (DWT).

Feature STFT DWT

Sampling frequency Medium/High Medium/High
Computational complexity/effort Low/Medium Medium/High
Detection time 0.5 s Less than 0.2 s

When the recognition accuracy of the DWT + CNN is configured, the method STFT + CNN was
conducted with the same training database and test database, and the results are shown in Table 5.

Table 5. Test results of the two measures.

Signal Number Ideal Results Results of STFT + CNN

1 1 1
2 1 1
3 1 0.9982
4 1 0.9997
5 1 1
6 0 0.1270
7 0 0.2654
8 0 0.1873
9 0 0.1249
10 0 0.1588
Energies 2020, 13, 4190 12 of 14

From Table 5, the minimum value of the fault data is 0.9982 while the maximum value of the
normal data is 0.2654; that is, the values of the fault data are close to 1 while those of the normal data
are near 0. It can be concluded that STFT + CNN also can discriminate between the fault data and
normal data. Nevertheless, the results of the STFT + CNN method fluctuate in a wider range than
that of the DWT+CNN method. For example, the values under normal conditions are in the range of
0.2654 to 0.1249 for the STFT+CNN method, while the results of the DWT + CNN method vary from
0.1311 to 0.1356 in Table 3. It seems that the convergence of the DWT + CNN is much better than that
of the STFT + CNN method.
Data in the training database and test database were refreshed by the data in the database, and two
algorithms were detected many times. The results show that DWT + CNN is more reliable than STFT
+ CNN to detect DC arc faults. Therefore, DWT + CNN is more appropriate to monitor the DC arc
occurring in the PDS of an AEA.

5. Conclusions and Prospects

This paper proposes a DC arc fault detection method for an SSPC in a PDS. The proposed method
is based on current time–frequency analysis and CNN. The dataspace was composed of normal current
data and arc current data. The experimental results illustrate the method proposed is effective to detect
the DC arc fault and the accuracy is close to 100%; the detection time is also fast, close to 1 s. However,
there is the need to train the CNN 3–5 times by using the training database.
The highlight of this paper is that a CNN was introduced into arc detection. A CNN could
determine whether the DC arc fault occurs in the PDS by recognizing the time–frequency spectrum of
the circuit current. Besides, DWT is a more appropriate way to produce a time–frequency spectrum
than the STFT in the field of aircraft DC arc fault diagnosis.
It is recommended that further research be undertaken in the following areas: First, it is necessary
to study the minimum sampling frequency that reflects the arc characteristics, since a high sampling
rate was used in the experiment. It ensures that the cost of the sampling device is reduced as much as
possible with the same accuracy. However, it is promising to transplant this method to a low sampling
detection since the arc signal is broadband and a low-frequency band in the range of 10 to 100 Hz
is still representative [40]. Secondly, the method proposed in this paper can be tried to identify the
AC arc fault. It may detect all kinds of arc failures in the PDS simultaneously by using the proposed
method here.

Author Contributions: Writing-review & editing, T.L.; Writing-original draft, Z.J.; Data curation, L.W.; Resources,
Y.M. All authors have read and agreed to the published version of the manuscript.
Funding: This work is funded by the National Key Research and Development Program of China
(No. 2018YFB1500800) and the Fault Location Program supported by Tangshan power supply company
(No. E18L00560).
Acknowledgments: Thank Mu Longhua and his research group at Tongji University for providing the technical
support in the experiments.
Conflicts of Interest: The authors declare no conflict of interest. The funders had no role in the design of the
study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, and in the decision to
publish the results.

References
1. Brusso, B.C. History of Aircraft Wiring Arc-Fault Protection [History]. IEEE Ind. Appl. Mag. 2017, 23, 6–11.
[CrossRef]
2. Li, X.; Pan, C.; Luo, D.; Sun, Y. Series DC Arc Simulation of Photovoltaic System Based on Habedank Model.
Energies 2020, 13, 1416. [CrossRef]
3. Yaramasu, A.; Cao, Y.; Liu, G.; Wu, B. Aircraft electric system intermittent arc fault detection and location.
IEEE Trans. Aerosp. Electron. Syst. 2015, 51, 40–51. [CrossRef]
Energies 2020, 13, 4190 13 of 14

4. He, C.; Mu, L.; Wang, Y. The Detection of Parallel Arc Fault in Photovoltaic Systems Based on a Mixed
Criterion. IEEE J. Photovolt. 2017, 7, 1717–1724. [CrossRef]
5. Uriarte, F.M.; Hotz, T.J.; Kwasinski, A.; Gattozzi, A.L.; Herbst, J.D.; Estes, H.B.; Hebner, R.E. A DC Arc Model
for Series Faults in Low Voltage Microgrids. IEEE Trans. Smart Grid 2012, 3, 2063–2070. [CrossRef]
6. Li, W.; He, K.; Liu, W.; Zhang, X.; Dong, Y. A fast arc fault detection method for AC solid state power
controllers in MEA. Chin. J. Aeronaut. 2018, 31, 1119–1129. [CrossRef]
7. Izquierdo, D.; Barrado, A.; Fernandez, C.; Sanz, M.; Lázaro, A. SSPC Active Control Strategy by Optimal
Trajectory of the Current for Onboard System Applications. IEEE Trans. Ind. Electron. 2013, 60, 5195–5205.
[CrossRef]
8. Yu, B.; Wang, L.; Ahmed, D. Drain–Source Voltage Clamp Circuit for Online Accurate ON-State Resistance
Measurement of SiC MOSFETs in DC Solid-State Power Controller. IEEE J. Emerg. Sel. Top. Power Electron.
2020, 8, 331–342. [CrossRef]
9. Izquierdo, D.; Barrado, A.; Raga, C.; Sanz, M.; Zumel, P.; Lázaro, A. Protection devices for aircraft electrical
power distribution systems: A survey. IEEE Trans. Aerosp. Electron. Syst. 2011, 47, 1538–1550. [CrossRef]
10. Miao, W.; Liu, X.; Lam, K.H.; Pong, P.W.T. Arc-Faults Detection in PV Systems by Measuring Pink Noise
With Magnetic Sensors. IEEE Trans. Magn. 2019, 55, 1–6. [CrossRef]
11. Kay, J.A.; Hussain, G.A.; Lehtonen, M.; Kumpulainen, L. New pre-emptive arc fault detection techniques in
medium voltage switchgear and motor controls. IEEE Trans. Ind. Appl. 2015, 52, 1–12.
12. Naidu, M.; Schoepf, T.; Gopalakrishnan, S. Arc fault detection scheme for 42-V automotive DC networks
using current shunt. IEEE Trans. Power Electron. 2006, 21, 633–639. [CrossRef]
13. Yao, X.; Herrera, L.; Ji, S.; Zou, K.; Wang, J. Characteristic Study and Time-Domain Discrete- Wavelet-Transform
Based Hybrid Detection of Series DC Arc Faults. IEEE Trans. Power Electron. 2014, 29, 3103–3115. [CrossRef]
14. Orton, H.E. Method for Diagnosing Degradation in Aircraft Wiring. U.S. Patent 6,909,977, 21 June 2005.
15. Zhao, C.; Li, K.; Zhang, Y.; Wang, Y. Arc fault detection based on electromagnetic radiation. Adv. Technol.
Electr. Eng. Energy 2017, 36, 70–74.
16. Yang, Y.; Peng, Z.; Zhang, W.-M.; Meng, G. Parameterised time-frequency analysis methods and their
engineering applications: A review of recent advances. Mech. Syst. Signal Process. 2019, 119, 182–221.
[CrossRef]
17. Le, V.; Yao, X.; Miller, C.; Tsao, B.-H.; Hung, T.-B. Series DC Arc Fault Detection Based on Ensemble Machine
Learning. IEEE Trans. Power Electron. 2020, 35, 7826–7839. [CrossRef]
18. Chen, S.; Xiong, J.; Li, X. Series Arc Fault Identification for Photovoltaic System Based on Time-Domain and
Time-Frequency-Domain Analysis. IEEE J. Photovolt. 2017, 7, 1105–1114. [CrossRef]
19. Artale, G.; Cataliotti, A.; Cosentino, V.; Di Cara, D.; Nuccio, S.; Tine, G. Arc Fault Detection Method Based on
CZT Low-Frequency Harmonic Current Analysis. IEEE Trans. Instrum. Meas. 2017, 66, 888–896. [CrossRef]
20. Mu, L.; Wang, Y.; Jiang, W.; Zhang, F. Study on characteristics and detection method of DC arc fault for
photovoltaic system. Proc. CSEE 2016, 36, 5236–5244.
21. Ahmadi, M.; Samet, H.; Ghanbari, T. Series Arc Fault Detection in Photovoltaic Systems Based on
Signal-to-Noise Ratio Characteristics Using Cross-Correlation Function. IEEE Trans. Ind. Inform. 2019, 16,
3198–3209. [CrossRef]
22. Georgijevic, N.; Jankovic, M.; Srdic, S.; Radakovic, Z. The Detection of Series Arc Fault in Photovoltaic
Systems Based on the Arc Current Entropy. IEEE Trans. Power Electron. 2016, 31, 5917–5930. [CrossRef]
23. Jovanovic, S.; Chahid, A.; Lezama, J.; Schweitzer, P. Shunt active power filter-based approach for arc fault
detection. Electr. Power Syst. Res. 2016, 141, 11–21. [CrossRef]
24. Ahmadi, M.; Samet, H.; Ghanbari, T. A new method for detecting series arc fault in photovoltaic systems
based on the blind source separation. IEEE Trans. Ind. Electron. 2019, 67, 5041–5049. [CrossRef]
25. Islam, A.; Birtwhistle, D.; Saha, T.K.; Islam, M.S. Interruption of Low-Voltage DC Arc in Air under Axial
Magnetic Field. IEEE Trans. Power Deliv. 2020, 35, 977–986. [CrossRef]
26. Zhang, X.; Liu, Z.; Wang, J.; Wang, J. Time–frequency analysis for bearing fault diagnosis using multiple
Q-factor Gabor wavelets. ISA Trans. 2018, 87, 225–234. [CrossRef]
27. Jiao, Z.; Li, T.; Wang, L.; Mu, L.; Khalyasmaa, A. DC series arc-fault detection of photovoltaic system based
on convolutional neural network. Adv. Technol. Electr. Eng. Energy 2019, 38, 29–34.
Energies 2020, 13, 4190 14 of 14

28. Qi, P.; Jovanovic, S.; Lezama, J.; Schweitzer, P. Discrete wavelet transform optimal parameters estimation for
arc fault detection in low-voltage residential power networks. Electr. Power Syst. Res. 2017, 143, 130–139.
[CrossRef]
29. Xia, K.; He, S.; Tan, Y.; Jiang, Q.; Xu, J.; Yu, W. Wavelet packet and support vector machine analysis of
series DC ARC fault detection in photovoltaic system. IEEJ Trans. Electr. Electron. Eng. 2019, 14, 192–200.
[CrossRef]
30. Telford, R.D.; Galloway, S.; Stephen, B.; Elders, I. Diagnosis of Series DC Arc Faults—A Machine Learning
Approach. IEEE Trans. Ind. Inform. 2017, 13, 1598–1609. [CrossRef]
31. Siegel, J.E.; Pratt, S.; Sun, Y.; Sarma, S.E. Real-time Deep Neural Networks for internet-enabled arc-fault
detection. Eng. Appl. Artif. Intell. 2018, 74, 35–42. [CrossRef]
32. Li, H.; Huang, J.; Ji, S. Bearing Fault Diagnosis with a Feature Fusion Method Based on an Ensemble
Convolutional Neural Network and Deep Neural Network. Sensors 2019, 19, 2034. [CrossRef] [PubMed]
33. Lu, S.; Sirojan, T.; Phung, B.T.; Zhang, D.; Ambikairajah, E. DA-DCGAN: An Effective Methodology for DC
Series Arc Fault Diagnosis in Photovoltaic Systems. IEEE Access 2019, 7, 45831–45840. [CrossRef]
34. Bong, K.; Choi, S.; Kim, C.; Yoo, H.-J. Low-Power Convolutional Neural Network Processor for a
Face-Recognition System. IEEE Micro 2017, 37, 30–38. [CrossRef]
35. Dolz, J.; Ayed, I.B.; Yuan, J.; Gopinath, K.; Lombaert, H.; Desrosiers, C. HyperDense-Net: A hyper-densely
connected CNN for multi-modal image semantic segmentation. IEEE Trans. Med. Imaging 2018, 38, 1116–1126.
[CrossRef]
36. Zhang, C.; Yue, X.; Wang, R.; Li, N.; Ding, Y. Study on Traffic Sign Recognition by Optimized Lenet-5
Algorithm. Int. J. Pattern Recognit. Artif. Intell. 2020, 34, 158–165. [CrossRef]
37. Krizhevsky, A.; Sutskever, I.; Hinton, G.E. ImageNet classification with deep convolutional neural networks.
Proc. Adv. Neural Inf. Process. Syst. 2012, 1, 1097–1105. [CrossRef]
38. Luo, J.-H.; Zhang, H.; Zhou, H.-Y.; Xie, C.-W.; Wu, J.; Lin, W. ThiNet: Pruning CNN Filters for a Thinner Net.
IEEE Trans. Pattern Anal. Mach. Intell. 2018, 41, 2525–2538. [CrossRef]
39. Lu, S.; Phung, B.; Zhang, D. A comprehensive review on DC arc faults and their diagnosis methods in
photovoltaic systems. Renew. Sustain. Energy Rev. 2018, 89, 88–98. [CrossRef]
40. Johnson, J.; Pahl, B.; Luebke, C.; Pier, T.; Miller, T.; Strauch, J.; Kuszmaul, S.; Bower, W. Photovoltaic DC arc
fault detector testing at Sandia National Laboratories. In Proceedings of the 2011 37th IEEE Photovoltaic
Specialists Conference, Seattle, WA, USA, 19–24 June 2011; pp. 3614–3619.

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).