electric power quality events detection and … · hilbert transform. the hilbert transform has...
TRANSCRIPT
VOL. 11, NO. 1, JANUARY 2016 ISSN 1819-6608
ARPN Journal of Engineering and Applied Sciences
©2006-2016 Asian Research Publishing Network (ARPN). All rights reserved.
www.arpnjournals.com
577
ELECTRIC POWER QUALITY EVENTS DETECTION AND
CLASSIFICATION USING HILBERT TRANSFORM
AND MLP NETWORK
P. Kalyana Sundaram and R. Neela Department of Electrical Engineering, Annamalai University, India
E-Mail: [email protected]
ABSTRACT
This paper shows a new technique for the detection and classification of various power quality events using
Hilbert transform. The Hilbert transform has been introduced as a powerful tool for input features exaction such as mean,
peak value and standard deviation from the distorted voltage waveforms using Matlab simulation for disturbance of
various classes in the test system. The MLP based neural network has been chosen as the classifier of several types of
power quality events and the neural network has been trained using 900 number of test data at the rate of 100 samples for
each class of disturbance. The algorithm has been tested with 900 number of test data and the outcomes are recorded.
Keywords: power quality, power quality disturbances, hilbert transform, neural network, MLP based neural network.
Nomenclature � = Real signal ��� = Hilbert transform signal � = Shifting operator −� �� � = Shifting the negative frequency of � � = Envelop signal of � � = Instantaneous phase signal of �
1. INTRODUCTION
Now a day’s power quality has become a main problem in the electric power system. The reason for the
poor quality of electric power is caused by the power line
disturbances such as sag, swell, interruption, harmonics,
sag with harmonics, swell with harmonics, flicker and
notches. The various types of power quality disturbances
were detected and classified using wavelet transform
analysis as illustrated in [1].The time and frequency of
multi resolution wavelet has been presented in [2] to
analyze the electromagnetic power system transients.
Another approach of wavelets to identify the
various power system transient signals such as capacitor
switching, lighting impulse, etc has been discussed in [3].
Comparison of various signal processing techniques such
as Fourier transform, multi resolution analysis and the
continuous wavelet transform forthe analysis of power
quality disturbanceshas been explained in [4]. The
Fourier and wavelet transform based fuzzy expert system
for the detection and classification of PQ disturbances has
been demonstrated in [5]. The wavelet based neural
classifier has been discussed in [6] for the recognition of
PQ disturbance waveform. The wavelet multi resolution
analysishas been used for the online classification of
power quality events along with pattern classifier in [7].
Discrete wavelet transform along with neural network for
PQ disturbance classification has been explained [8]. As
wavelet transforms cannot be applied for the analysis of
non stationary signals, S-transforms were implemented
due to their excellent frequency resolution characteristics.
Application of s-transform for power quality analysis has
been discussed in [9] and a fuzzy logic based pattern
recognition system along with multi resolution S-
transform for power quality event classification has been
discussed in [10].
S-transform based neural network for the
detection and classification of PQ disturbance signal has
been implemented in [11]. A combination of wavelet
transform along with Kalman filter has been presented in
[12] for the detection and analysis of voltage event in
power system. Discrete wavelet transform and wavelet
coefficients based neural network classifier has been
demonstrated for the classification of PQ disturbances in
[13]. The S-transform and modular neural network based
power quality classifier has been presented in [14] and
this combines the frequency resolution characteristics of
S transform with the pattern recognizing ability of a
neural network. A hybrid method of adaptive filtering and
discrete wavelet transform for detection and analysis of
PQ events in [15]. The classification of the various types
of power quality disturbances based on S-transform and
Probabilistic neural network has been discussed in [16].
Wavelet transform support vector machines
based power quality disturbance classifier is presented in
[17] where the analysis takes into account the different
noise conditions also. Hilbert transform has been used for
the detection and classification of power quality events
along with RBF network in [18]. A combination of
kalman filter and fuzzy expert system for the
characterization of power quality disturbances has been
illustrated in [19]. The dual neural network as ADALINE
and FFNN has been implemented for the classification of
single and combined form power quality disturbance in
[20].A Hilbert transform and MLP neural network based
power quality analyzer in which features are extracted
VOL. 11, NO. 1, JANUARY 2016 ISSN 1819-6608
ARPN Journal of Engineering and Applied Sciences
©2006-2016 Asian Research Publishing Network (ARPN). All rights reserved.
www.arpnjournals.com
578
using Hilbert transform and disturbances are identified
and classified using an MLP based neural network is
presented in this paper.
2. PROPOSED METHOD
The classification technique proposed has two
stages namely a feature extraction stage and a
classification stage. In the feature extraction stage,
Hilbert transform is used for extracting features such as
amplitude, slope and standard deviation. The
classification stage consists of a MLP based neural
network with four hidden layers. Disturbance waveforms
were generated using Matlab simulink.
2.1 Feature extraction stage using Hilbert transform
The Hilbert Transform is used to generate an
analytical signal obtained by convolving the real signal
with the function as shown below:
��� = � �� (1)
��� = � ∫ � ��−�∞−∞ � (2)
��� = −� �� � � � (3)
The output of the Hilbert transform is 90 phase
shift of the original signal� , a complex signal. It is
defined as
�� = � + ��� (4)
�� = � �� � (5)
The analytical signal has the information about
amplitude as well phase of the signal. It is clear that
� = √� + ��� (6)
� = ��− ��� �� � (7)
2.2 Multi-layer perceptron (MLP) neural network
A multilayer perceptron neural network is a
feed-forward artificial neural network that has an input
layer, output layer and one or more hidden layers. A MLP
based neural network consists of multiple layers of nodes
in which each layer connected to the next one fully in a
directed graph. Except for the input nodes, each node is a
neuron with a nonlinear activation function. MLP based
neural network utilizes a supervised learning technique
called back propagation for training the network. MLP
based neural network architecture diagram is shown as in
the Figure-1. The training parameters of the MLP used in
this work are shown in Table-1.
Figure-1. Architecture of MLP neural network.
Table-1. MLP architecture and training parameters.
Architecture
The number of layers 3
The number of neuron on the layers Input: 13, hidden: 10, output: 9
The initial weights and biases Random
Activation functions Tangent sigmoid
Training parameters
Learning rule Back-propagation
Learning rate 0.75
Mean-squared error 1E−08
VOL. 11, NO. 1, JANUARY 2016 ISSN 1819-6608
ARPN Journal of Engineering and Applied Sciences
©2006-2016 Asian Research Publishing Network (ARPN). All rights reserved.
www.arpnjournals.com
579
Back propagation learning algorithm
BP has two phases:
Forward pass phase: Calculate the ‘functional signal’ and feed forward propagation of input pattern signals through network
Backward pass phase: Calculate the ‘error signal’ and propagates the error backwards through network
starting at output units (the difference between actual
and desired output values).
The back-propagation network has an input
layer, an output layer, and atleast one hidden layer. There
is no limit on the number of hidden layers but typically
there is just one or two. But in some case a minimum
requirement of four layers (three hidden layers plus an
output layer) are used to solve complex problems. Each
layer is fully connected to the succeeding layer.
Recall is the process of setting input data into a
trained network and receiving the answer. Back-
propagation is not used during recall, but only when the
network is learning a training set.
3. CLASSIFICATION STAGE
In this stage, Hilbert Transform extracted input
features such as standard deviation, peak value, variances.
The extracted input features are applied to the multi-layer
perception based neural network in order to classify the
disturbances. MLP networks are very useful for the
classification of those input signals.
3.1 Flowchart of the proposed method
The flowchart for the Classification of Power
Quality disturbances is shown in below. It has three
different blocks.
Block-(a)-Features extraction such as amplitude, slope
andstandard deviation
Block-(b) -Detection and classification of the power
quality disturbances
Figure-2. Flowchart for the classification of power quality disturbances.
VOL. 11, NO. 1, JANUARY 2016 ISSN 1819-6608
ARPN Journal of Engineering and Applied Sciences
©2006-2016 Asian Research Publishing Network (ARPN). All rights reserved.
www.arpnjournals.com
580
4. SIMULATION AND TEST RESULTS
Training and Test data were generated using
Matlab simulink for various classes of disturbances .The
signals closer to real situation can be simulated and
different signals belonging to same class can be generated
with ease so that the generalization ability of neural
network based classifier could be improved. The single
line diagram for the test system and the Matlab
simulation block diagram are shown in Figure-3 and
Figure-4.
Figure-3. Single line diagram of test system model.
Figure-4. Matlab simulation block diagram for the test system model.
Nine classes of different PQ disturbances,
namely pure sine (normal), sag, swell, outage, harmonics,
sag with harmonic, swell with harmonic, notch and
flicker were considered. Total size of the training data set
is 3*900, where 3 represents the number of features
extracted for each type of disturbance and 900 represents
the total number of samples at the rate of 100 samples for
each one of the 9 disturbances. The PQ disturbance
signals generated using the Matlab simulink and the
training performance of neural network with 100 epochs
are shown in figures 5(a) to 14(c) for various classes of
disturbance.
The following case studies are presented to
highlight the suitability of the application of the proposed
method.
Pure sine wave
It is a voltage signal of amplitude at the
frequency 50 Hzand its waveform is as shown in the
Figure-5(a). The mean, peak value and the standard
deviation outputs of the signal are shown in the Figures
5(b) and 5(e).
v + - vm3 v
+ - vm2 v + - vm1
Discrete
Ts = 5e-005 s. powergui
A B C
A B C Z-Source
A B C
A B C Z-Feeder
v + -
Voltage Measurement2
v + -
Voltage Measurement1
v + -
Voltage Measurement
A B C
VS
v + - V3
v + - V2
v + - V1
V-Phase A v +
- V-M
V-I
V
A B C
+ -
Universal Bridge
z 1
Unit Delay signal2
To Workspace1 signal
To Workspace
Timer2
Timer1
Timer
com A B C
a b c
Three-Phase Breaker1
Scope1
Scope
S3 S2 S1
Vabc Iabc A
B C a b c
PCC
Load
A B C A B C
L-G fault
A B C Heavy load
m a
k Diode2
m a
k Diode1
m a
k Diode
com A B C
a b c
CB2
com A B C
a b c
CB1
A B C A B C
3-phase SC fault
A B C Load1
A B C Load
VOL. 11, NO. 1, JANUARY 2016 ISSN 1819-6608
ARPN Journal of Engineering and Applied Sciences
©2006-2016 Asian Research Publishing Network (ARPN). All rights reserved.
www.arpnjournals.com
581
Figure-5(a).
Figure-5(b).
Figure-5(c).
Figure-5(d).
Figure-5(e).
Figure-5. Pure voltage signal.
a, Waveform, b) Envelop detection using Hilbert
Transform, c) Mean, d) Peak value, e) Standard deviation
Voltage sag
The voltage sag (or) voltage dips cause the
decrease of system voltage. The duration of the sag
disturbance is 0.4 to 0.8 cycles in 1 min. The voltage dip
is generated by the occurrence of a single line to ground
fault and its waveform is shown in the Figure-6(a). The
mean, peak value and standard deviation outputs are
shown in the Figures 6(b) to 6(e).
Figure-6(a).
Figure-6(b).
Figure-6(c).
Figure-6(d).
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000-6
-4
-2
0
2
4
6x 10
5 Testing Signal
VoltageM
agnitude
TimeCycles
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 110003.9384
3.9384
3.9384
3.9384
3.9384
3.9384x 10
7 Envelop Detection:Hilbert Transform
VoltageM
agnitude
TimeCycles
1 2 3 4 5 6 7 8 9 103.9384
3.9384
3.9384
3.9384
3.9384
3.9384x 10
7 Mean
1 2 3 4 5 6 7 8 9 103.9384
3.9384
3.9384
3.9384
3.9384
3.9384x 10
7 Peak
1 2 3 4 5 6 7 8 9 103
3.5
4
4.5
5
5.5Standard Deviation
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000-6
-4
-2
0
2
4
6x 10
5 Testing Signal
Vo
lta
geM
ag
nitu
de
TimeCycles
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 100003.1904
3.1906
3.1908
3.191
3.1912
3.1914
3.1916x 10
7 Envelop Detection:Hilbert Transform
VoltageM
agnitude
TimeCycles
1 2 3 4 5 6 7 8 9 103.1907
3.1907
3.1907
3.1907
3.1907
3.1907x 10
7 Mean
1 2 3 4 5 6 7 8 9 103.1915
3.1915
3.1915
3.1915
3.1915
3.1915x 10
7 Peak
VOL. 11, NO. 1, JANUARY 2016 ISSN 1819-6608
ARPN Journal of Engineering and Applied Sciences
©2006-2016 Asian Research Publishing Network (ARPN). All rights reserved.
www.arpnjournals.com
582
Figure-6(e)
Figure-6. Voltage sag.
a, Waveform, b) Envelop detection using Hilbert
Transform, c) Mean, d) Peak value, e) Standard deviation
Voltage swell
Voltage swell causes the rise of system voltage.
The duration of the swell disturbance is 0.4 to 0.2 cycles
in 1 min. The voltage swell is generated by disconnecting
the heavy load for 10 cycles and its waveform is shown in
the Figure-7(a). The mean, peak value and standard
deviation outputs of the Hilbert transform are shown in
the Figures 7(b) to 7(e).
Figure-7(a).
Figure-7(b).
Figure-7(c).
Figure-7(d).
Figure-7(e).
Figure-7. Voltage swell.
a, Waveform, b) Envelop detection using Hilbert
Transform, c) Mean, d) Peak value, e) Standard deviation
Voltage surge
The voltage surge causes the sudden rise of the
voltage in the short duration of time. The duration of the
surge disturbance is 0.2 to 0.22 cycles in 1 min. The
voltage surge is generated by disconnecting the heavy
load for one quarter cycle and its waveform is shown in
the Figure-8(a). Figures 8(b) to 8(e) represents the feature
extraction outputs of Hilbert transform.
Figure-8(a).
Figure-8(b).
1 2 3 4 5 6 7 8 9 102166.5
2167
2167.5
2168
2168.5
2169Standard Deviation
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000-6
-4
-2
0
2
4
6x 10
5 Testing Signal
VoltageM
agnitude
TimeCycles
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 100002.8908
2.891
2.8912
2.8914
2.8916
2.8918x 10
7 Envelop Detection:Hilbert Transform
VoltageM
agnitude
TimeCycles
1 2 3 4 5 6 7 8 9 102.8911
2.8911
2.8911
2.8911
2.8911
2.8911x 10
7 Mean
1 2 3 4 5 6 7 8 9 102.8917
2.8917
2.8917
2.8917
2.8917
2.8917x 10
7 Peak
1 2 3 4 5 6 7 8 9 102011
2011.5
2012
2012.5
2013
2013.5Standard Deviation
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000-6
-4
-2
0
2
4
6x 10
5 Testing Signal
VoltageM
agnitude
TimeCycles
0 2000 4000 6000 8000 10000 120002.019
2.0192
2.0194
2.0196
2.0198
2.02
2.0202
2.0204x 10
7 Envelop Detection:Hilbert Transform
VoltageM
agnitude
TimeCycles
VOL. 11, NO. 1, JANUARY 2016 ISSN 1819-6608
ARPN Journal of Engineering and Applied Sciences
©2006-2016 Asian Research Publishing Network (ARPN). All rights reserved.
www.arpnjournals.com
583
Figure-8(c).
Figure-8(d).
Figure-8(e).
Figure-8. Voltage surge.
a. Waveform, b) Envelop detection using Hilbert
Transform, c) Mean, d) Peak value, e) Standard deviation
Voltage interruption
The Voltage interruption causes the loss of the
voltage on the system. The duration of the voltage
interruption is 0.4 to 0.8 cycles. The voltage surge is
generated by three phase short circuit fault and its
waveform is shown in the Figure-9(a). The mean, peak
value and standard deviation outputs of the Hilbert
transform are shown in the Figures 9(b) and 9(e).
Figure-9(a).
Figure-9(b).
Figure-9(c).
Figure-9(d).
Figure-9(e).
Figure-9. Voltage surge.
a. Waveform, b) Envelop detection using Hilbert
Transform, c) Mean, d) Peak value, e) Standard deviation
Harmonics
Harmonics are generated by the connection of
non linear load to the system. The distortion of the
voltage waveform is shown in the Figure-10(a). Hilbert
transform extract the three types of features. They are
shown in the Figures 10 (b) to 10(e).
1 2 3 4 5 6 7 8 9 102.0191
2.0191
2.0191
2.0191
2.0191
2.0191x 10
7 Mean
1 2 3 4 5 6 7 8 9 102.0202
2.0202
2.0202
2.0202
2.0202
2.0202x 10
7 Peak
1 2 3 4 5 6 7 8 9 101141
1141.5
1142
1142.5
1143
1143.5Standard Deviation
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000-6
-4
-2
0
2
4
6x 10
5 Testing Signal
VoltageM
agnitude
TimeCycles
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 100003.13
3.1302
3.1304
3.1306
3.1308
3.131
3.1312
3.1314
3.1316
3.1318x 10
7 Envelop Detection:Hilbert Transform
VoltageM
agnitude
TimeCycles
1 2 3 4 5 6 7 8 9 103.1304
3.1304
3.1304
3.1304
3.1304
3.1304x 10
7 Mean
1 2 3 4 5 6 7 8 9 103.1317
3.1317
3.1317
3.1317
3.1317
3.1317x 10
7 Peak
1 2 3 4 5 6 7 8 9 102387
2387.5
2388
2388.5
2389
2389.5Standard Deviation
VOL. 11, NO. 1, JANUARY 2016 ISSN 1819-6608
ARPN Journal of Engineering and Applied Sciences
©2006-2016 Asian Research Publishing Network (ARPN). All rights reserved.
www.arpnjournals.com
584
Figure-10(a).
Figure-10(b).
Figure-10(c).
Figure-10(d).
Figure-10(e).
Figure-10. Harmonics.
a, Waveform, b) Envelop detection using Hilbert
Transform, c) Mean, d) Peak value, e) Standard deviation
Sag with harmonics
This type of disturbance is caused by the
presence of a nonlinear load and a voltage dip in the
system for duration of 0.5 to 0.8 cycles. The waveform
contains harmonic distortion with sag event as shown in
the Figure-11(a). The mean, peak value and standard
deviation outputs are shown in the Figures 11(b) to 11(e).
Figure-11(a).
Figure-11(b).
Figure-11(c).
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000-4
-3
-2
-1
0
1
2
3
4x 10
5 Testing Signal
VoltageM
agnitude
TimeCycles
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 100002.72
2.7201
2.7202
2.7203
2.7204
2.7205
2.7206
2.7207
2.7208x 10
7 Envelop Detection:Hilbert Transform
VoltageM
agnitude
TimeCycles
1 2 3 4 5 6 7 8 9 102.7202
2.7202
2.7202
2.7202
2.7202
2.7202x 10
7 Mean
1 2 3 4 5 6 7 8 9 102.7209
2.7209
2.7209
2.7209
2.7209
2.7209x 10
7 Peak
1 2 3 4 5 6 7 8 9 10917.5
918
918.5
919
919.5
920Standard Deviation
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000-4
-3
-2
-1
0
1
2
3
4x 10
5 Testing Signal
VoltageM
agnitude
TimeCycles
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 100002.2738
2.274
2.2742
2.2744
2.2746
2.2748x 10
7 Envelop Detection:Hilbert Transform
VoltageM
agnitude
TimeCycles
1 2 3 4 5 6 7 8 9 102.2741
2.2741
2.2741
2.2741
2.2741
2.2741x 10
7 Mean
VOL. 11, NO. 1, JANUARY 2016 ISSN 1819-6608
ARPN Journal of Engineering and Applied Sciences
©2006-2016 Asian Research Publishing Network (ARPN). All rights reserved.
www.arpnjournals.com
585
Figure-11(d).
Figure-11(e).
Figure-11. Sag with harmonics.
a, Waveform, b) Envelop detection using Hilbert
Transform, c) Mean, d) Peak value, e) Standard deviation
Swell with harmonics
This type of disturbance is caused by the
presence of nonlinear load and a voltage swell in the
system for the duration of 0.4 to 0.8 cycles. The
waveform contains harmonic distortion with swell event
as shown in the Figure-12(a). Figures 12(b) to 12(e)
represents the feature extraction outputs of Hilbert
transform.
Figure-12(a).
Figure-12(b).
Figure-12(c).
Figure-12(d).
Figure-12(e).
Figure-12. Swell with harmonics.
a, Waveform, b) Envelop detection using Hilbert
Transform, c) Mean, d) Peak value, e) Standard deviation
Flicker
This type of disturbance type is caused by the
continuous and rapid variation of the system load.
Theflicker disturbance is generated by disconnecting the
heavy load in the continuous nature at regular interval
and the waveform of the flicker is shown in the Figure-
13(a). The mean, peak value and standard deviation
outputs of the Hilbert transform are shown in the Figures
13(b) and 13(e).
1 2 3 4 5 6 7 8 9 102.275
2.275
2.275
2.275
2.275
2.275x 10
7 Peak
1 2 3 4 5 6 7 8 9 101617.5
1618
1618.5
1619
1619.5
1620Standard Deviation
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000-4
-3
-2
-1
0
1
2
3
4x 10
5 Testing Signal
VoltageM
agnitude
TimeCycles
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 100002.0875
2.088
2.0885x 10
7 Envelop Detection:Hilbert Transform
VoltageM
agnitude
TimeCycles
1 2 3 4 5 6 7 8 9 102.0877
2.0877
2.0877
2.0877
2.0877
2.0877x 10
7 Mean
1 2 3 4 5 6 7 8 9 102.0885
2.0885
2.0885
2.0885
2.0885
2.0885x 10
7 Peak
1 2 3 4 5 6 7 8 9 101401.5
1402
1402.5
1403
1403.5
1404Standard Deviation
VOL. 11, NO. 1, JANUARY 2016 ISSN 1819-6608
ARPN Journal of Engineering and Applied Sciences
©2006-2016 Asian Research Publishing Network (ARPN). All rights reserved.
www.arpnjournals.com
586
Figure-13(a).
Figure-13(b).
Figure-13(c).
Figure-13(d).
Figure-13(e)
Figure-13. Flicker.
a, Waveform, b) Envelop detection using Hilbert
Transform, c) Mean, d) Peak value, e) Standard deviation
Notch
This is a disturbance of the nominal power
voltage waveform lasting for less than half a cycle. The
disturbance is initially of opposite polarity and hence it is
to be subtracted from the waveform as shown in the
Figure-14(a). Figures 14(b) to 14(e) represents the feature
extraction outputs of Hilbert transform.
Figure 14(a).
Figure-14(b).
Figure-14(c).
Figure-14(d).
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000-6
-4
-2
0
2
4
6x 10
5 Testing Signal
Volta
geM
agnitu
de
TimeCycles
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 100003.0442
3.0444
3.0446
3.0448
3.045
3.0452
3.0454
3.0456x 10
7 Envelop Detection:Hilbert Transform
VoltageM
agnitude
TimeCycles
1 2 3 4 5 6 7 8 9 103.0446
3.0446
3.0446
3.0446
3.0446
3.0446x 10
7 Mean
1 2 3 4 5 6 7 8 9 103.0455
3.0455
3.0455
3.0455
3.0455
3.0455x 10
7 Peak
1 2 3 4 5 6 7 8 9 101762.5
1763
1763.5
1764
1764.5
1765Standard Deviation
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000-6
-4
-2
0
2
4
6x 10
5 Testing Signal
VoltageM
agnitude
TimeCycles
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 100003.7839
3.7839
3.7839
3.784
3.7841
3.7841x 10
7 Envelop Detection:Hilbert Transform
VoltageM
agnitude
TimeCycles
1 2 3 4 5 6 7 8 9 103.784
3.784
3.784
3.784
3.784
3.784x 10
7 Mean
1 2 3 4 5 6 7 8 9 103.7841
3.7841
3.7841
3.7841
3.7841
3.7841x 10
7 Peak
VOL. 11, NO. 1, JANUARY 2016 ISSN 1819-6608
ARPN Journal of Engineering and Applied Sciences
©2006-2016 Asian Research Publishing Network (ARPN). All rights reserved.
www.arpnjournals.com
587
Figure-14(e).
Figure-14. Voltage notch.
a. Waveform, b) Envelop detection using Hilbert
Transform, c) Mean, d) Peak value, e) Standard deviation
Table-3. Classification accuracy.
S. No. Power quality
disturbances
Input
Features
Percentage
of accuracy
1 Voltage Sag 100 100
2 Voltage Swell 100 100
3 Voltage
interruption 100 99
4 Voltage surge 100 98
5 Harmonics 100 99
6 Sag with
Harmonics 100 100
7 Swell with
Harmonics 100 100
8 Flicker 100 98
9 Notch 100 99
Overall accuracy 99.22
Figure-15. Bar diagram for the percentage of accuracy of the proposed method.
4. CONCLUSIONS
The paper has presented a new method for
detection and classification of power quality disturbances
in the electric power systems. The method combines the
Hilbert transform and MLP based neural network. The
disturbance waveforms were generated through Matlab
simlink and features such as mean, peak value and
standard deviation were extracted through Hilbert
transform and an MLP based neural network has been
applied for classifying the disturbances. The method
enables the accurate classification of all nine types of PQ
disturbances and it is well suitable for real world
applications were the classifier is applied over the data
captured in field. MLP neural network has the wide range
of skills that it can be trained for any input combination
and its application is particularly suitable for
classification of disturbances of varying nature.
REFERENCES
[1] Surya Santoso, Edward J. Powers and W. Mack
Grady. 1994. Electric power quality disturbance
1 2 3 4 5 6 7 8 9 10221
221.5
222
222.5
223
223.5Standard Deviation
97%
98%
98%
99%
99%
100%
100%
101%
Percentage of accuracy
Sine wave
Sag
Swell
Interruption
Surges
Harmonics
Sag with harmonics
Swell with harmonics
Flicker
Notch
VOL. 11, NO. 1, JANUARY 2016 ISSN 1819-6608
ARPN Journal of Engineering and Applied Sciences
©2006-2016 Asian Research Publishing Network (ARPN). All rights reserved.
www.arpnjournals.com
588
detection using wavelet transforms analysis. IEEE
Transaction on power delivery.
[2] David C. Robertson, Octavia I. Camps, Jeffrey S.
Mayer, William B. Gish. 1996. Wavelet and
electromagnetic power system transients. IEEE
Transaction on power delivery.
[3] G. T. Heydt, A. W. Galli. 1997. Transient power
quality problems analyzed using wavelets. IEEE
Transaction on power delivery.
[4] O. Poisson, P. Rioual and M. Meunier. 1999. New
signal processing tools applied to power quality
analysis. IEEE Transaction on power delivery.
[5] Mladen Kezunovic, Fellow. 2000. Advanced
assessment of the power quality events. IEEE
Transaction on power delivery.
[6] Surya Santoso, Edward J. Powers, and W. Mack
Grady. 2000. Power quality disturbance waveform
recognition using wavelet-based neural classifier-part
1: theoretical foundation. IEEE Transaction on power
delivery.
[7] A.M. Gaouda, S.H. Kanoun, M.M.A. Salama. 2001.
On-line disturbance classification using nearest
neighbor rule. Electric power systems research.
[8] Dolores Borrás, M. Castilla, Narciso Moreno, and J.
C. Montaño. 2001. Wavelet and neural structure: a
new tool for diagnostic of power system
disturbances. IEEE Transaction on industry
applications.
[9] P.K. Dash, B.K. Panigrahi, G. Panda. 2003. Power
quality analysis using S-transform. IEEE Trans.
Power Delivery. 18(2): 406-411.
[10] M.V. Chilukuri, P.K. Dash, Multiresolution S. 2004.
Transform-based fuzzy recognition system for power
quality events, IEEE Trans. Power Deliv. 19(1): 323-
330.
[11] I. W. C. Lee, P. K. Dash. 2003. S-transform-based
intelligent system for classification of power quality
disturbance signals. IEEE Transaction on industry
electronics.
[12] Enrique Pérez, Julio Barros. 2006. Voltage event
detection and characterization methods: A
comparative study. IEEE Transaction on power
delivery.
[13] A. K. Pradhan, A. Routray, A. Behera. 2006. Power
quality disturbance classification employing Modular
Wavelet Network. IEEE Transactions.
[14] C.N. Bhende, S. Mishra, B.K. Panigrahi. 2008.
Detection and classification of power quality
disturbances using S-transform and modular neural
network. Electric power systems research.
[15] U. D. Dwivedi, S. C. Srivastava. 2007. Analysis of
transient disturbances in distribution systems: A
Hybrid Approach. IEEE Transactions.
[16] S. Mishra, C. N. Bhende and B. K. Panigrahi. 2008.
Detection and classification of power quality
disturbances using S-Transform and Probabilistic
Neural Network. IEEE Trans. Power Delivery. Vol.
23.
[17] Sami Ekici. 2009. Classification of power system
disturbances using support vector machines. Expert
Systems with applications.
[18] T. Jayasree, D. Devaraj, R.Sukanesh. 2010. Power
quality disturbance classification using Hilbert
transform and RBF networks. Neurocomputing.
[19] Abdelazeem A. Abdelsalam, Azza A. Eldesouky,
Abdelhay A. Sallam. 2012. Characterization of
power quality disturbance using hybrid technique of
linear kalman filter and fuzzy expert system, Electric
power systems research.
[20] Martin Valtierra-Rodriguez, Rene de Jesus Romero-
Troncoso, Roque Alfredo Osornio-Rios and Arturo
Garcia-Perez. 2014. Detection and classification of
single and combined power quality disturbances
using neural networks. IEEE Transaction on industry
electronics.