See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/229007775

A System for the Generation and Detection of Electrical Disturbances

Article · January 2004

CITATIONS READS
0 2,517

6 authors, including:

Carlos León Jorge Ropero

Universidad de Sevilla Universidad de Sevilla
170 PUBLICATIONS 1,963 CITATIONS 35 PUBLICATIONS 360 CITATIONS

SEE PROFILE SEE PROFILE

Juan-Carlos Montaño José Manuel Elena

Universidad de Sevilla Universidad de Sevilla
75 PUBLICATIONS 1,088 CITATIONS 18 PUBLICATIONS 184 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Carlos León on 15 May 2014.

The user has requested enhancement of the downloaded file.

 A SYSTEM FOR THE GENERATION AND DETECTION
OF ELECTRICAL DISTURBANCES
Iñigo Monedero1, Carlos León1, Jorge Ropero1, José Luis de la Vega2,, Juan C. Montaño2, José Manuel Elena1

1 2
School of Computer Science and Engineering IRNAS
Electronic Technology Department Campus Reina Mercedes
Avda, Reina Mercedes s/n P.O. Box 1052
41012 Seville (Spain) 41080 Seville (Spain)
contact email: [email protected] phone: 34-954624711

Abstract. Power Quality is defined as the study of the damage suddenly. Harmonics, sometimes referred to as
quality of electric power lines. The detection and electrical pollution, are distortions of the normal voltage
classification of the different disturbances which cause waveforms found in ac transmission, which can arise at
power quality problems is a difficult task which requires a virtually any point in a power system. While harmonics
high level of engineering expertise. Thus, neural networks can be as destructive as transients, often the greatest
are usually a good choice for the detection and damage from these distortions lies in the loss of
classification of these disturbances. This paper describes a credibility of the power utilities vis-a-vis their customers.
powerful system, developed by the Institute for Natural The classification and identification of each one of the
Resources and Agrobiology at the Scientific Research disturbances is normally carried out from standards and
Council (CSIC) and the Electronic Technology recommendations depending on where the utilities operate
Department at the University of Seville, for the generation (IEEE in the United States, UNE in Spain, etc).
and detection (by means of neural networks) of electrical
disturbances.

Key Words

Power quality, electrical disturbance, Wavelet transform,

neural network.

1.- Introduction
Power Quality

Power Quality (PQ) is defined as the study of the quality

of electric power lines. PQ has been a topic of
consideration for the last two decades, and has recently
acquired intensified interest due to the wide spread use of
electronical devices in complicated industrial processes Figure 1.- Standard IEEE 1159
and the generalised power quality of commercial electric
power [1]. Thus nowadays, customers demand higher The detection and classification of the different
levels of PQ to ensure the proper and continued operation disturbances which cause power quality problems is a
of such sensitive equipment. difficult task which requires a high level of engineering
expertise [2]. Due to the above mentioned difficulties,
The poor quality of electrical power is normally attributed artificial intelligence tools [3] emerge as an interesting
to power line disturbances such as waveshape faults, alternative in the detection of electrical disturbances. The
overvoltages, capacitor switching transients, harmonic main intelligent tools of interest include expert systems,
distortion and impulse transients. Thus, electromagnetic fuzzy logic and artificial neural networks (ANNs) [4].
transients, which are momentary voltage surges powerful
enough to shatter a generator shaft, can cause catastrophic

442-286 180
Neural Networks on Power Quality
HEADER
For the detection and classification of disturbances, ANNs Name: Impulse-01.txt
can be combined with mathematical analysis such as Ideal cycles: 5
Fourier and Wavelet transforms for the generation of Sampling period (ms): 0.156250
signal features which serve as inputs in the network [5]. END HEADER
Thus, feature extraction by wavelet transforms provides -28.171582
an unique characteristic which can represent every single -12.934081
PQ disturbance at different resolutions using the 2.334579
technique called multi-resolution signal decomposition or 17.597614
multi resolution analysis. In this way, while the detection 32.818256
of the power quality signals has tended to be easy, their 47.959835
classification is still a difficult task in which ANNs play 62.985875
an important role [6][7][8]. 77.860176
92.546906
Pattern recognition in ANNs generally requires 107.010682
preprocessing of data, feature extraction and final 121.216659
classification. One of the most important tasks in the ...
design and development process of an ANN is to generate
an adequate number of training patterns in order to Figure 2.- File from the Electrical Pattern Generator
approximate future inputs. Sometimes an optimal design
of the ANN is found but the limited number of train The type of disturbances includes: impulse, oscillation,
patterns does not give good results. In particular, in PQ a sag, swell, interruption, undervoltage, overvoltage,
great numbers of electrical patterns are necessary due to harmonics and frequency variations. In amplitude
the multiple combinations of different disturbances which disturbances (impulses, sags, swells, interruptions,
can coincide in one or various samples. Another undervoltages and overvoltages), the tool allows us that
additional problem with ANNs applied to PQ is the parameters such as amplitude, start time, final time, rising
impossibility of getting real patterns directly from the and falling slope, be configured. The edition of harmonics
power line due to the irregularity in the apparition of allows the configuration of amplitude and phases as far as
disturbances. forty harmonic levels including the possibility of adding
them an offset.
2.- Electrical Pattern Generator
For the task of training neural networks for the detection
and classification of electrical disturbances we are
developing an electrical pattern generator. The objective
of this generator is to create an unlimited number of
patterns to be used by a classification system.

The Electrical Pattern Generator make it possible to

configure parameters such as the duration of the sample, Figure 3.- Ideal signal (one-phase representation)
the frequency of the signal and the number of samples in
an ideal cycle (50Hz or 60Hz) and to add one or more Models of disturbances
disturbances. From the selected parameters, the generator
creates a text file with the voltage values of the sample. 1.- Harmonic distortion
The structure of the file consists of a header with the file
information (name, number of sample cycles and Harmonic distortion is defined as the phenomenon in
sampling period) and a data column corresponding to the which diverse sinusoidal signals with diverse frequencies
voltages of each of the samples. One file example from which are multiples of the fundamental frequency are
the Electrical Pattern Generator can be seen in Figure 2. superposed on the ideal signal (Figure 4).

181
 40  40
X ( t )= A + ∑ Ai sin 2πif c t + ϕ i + B + ∑ B j sin( 2πjf m t + ϕ j
)
i =1  
j =1

(3)
which shows the harmonic content and frequency
modulation.

The addition of another kind of disturbance was carried

out from the previous expression X(t). Therefore, the
Figure 4.- Harmonic distortion previous expression would be the result of an ideal
electrical signal or a frequency or/and harmonic disturbed
The following mathematical model was implemented in signal.
the generator:
3.- Overvoltages, swells, undervoltages and sags.
N
C (t ) = A + ∑ Ai sin(2π f i t + ϕ i ) (1) In this kind of disturbances the amplitude of the signal
i =1 rises (overvoltages or swells) or falls (undervoltages and
sags) a certain value along a time interval.
• A: DC term (V).
• Ai: Amplitude of the ith harmonic of signal (V).
• fi: Frequency of the ith harmonic of signal (Hz)
• φi: Phase of the ith harmonic (Rad)
• i: Harmonic order (i= 1,.., N).

In our harmonic model C(t) is considered as consisting of

a fundamental and 39 harmonic components.

2.- Frequency deviation Figure 6.- Sag

Frequency deviation is a signal disturbance added to the
harmonic distortion. The model consists of the frequency In the development of the disturbance generator, a
modulation of the signal C(t) by means of the carrier trapezoidal model for the amplitude evolution (lineal
signal M(t), which is named modulating signal. slope) was considered. The model makes it possible to
approximate the amplitude disturbances most frequently
encountered in power systems. Figure 7 shows a graphical
of the model used for overvoltages or swells (inverse
trapeze for undervoltages and sags).

Figure 5.- Frequency deviation

The mathematical expression of this signal is:

40
M (t )= B + ∑B
j =1
j sin( 2 π jf m t + ϕ j ) (2)

• B: DC term (V).
• Bj: Amplitude of the jth harmonic of signal (V)
• fm: Fundamental frequency (Hz)
• φi: Phase of the ith harmonic (Rad) Figure 7.- Overvoltage and swell model
Considering (1) and (2) the resultant signal would end up • p: initial sample of the trapeze
as: • pip: slope of the initial ramp

182
• pfp: slope of the final ramp 3.- Electrical disturbances classifier based on
• n1: number of samples of the initial ramp Neural Networks.
• n2: number of samples of the final ramp
• t: number of samples when the climb is reached
• select: total number of samples In addition to the generation system, we have developed a
first version of a system for the detection and
classification of electrical disturbances. The system is a
detector of power line disturbances based on artificial
intelligence techniques (in particular, a first version based
on ANNs). We have used the previously described
Electrical Pattern Generator in order to carry out the ANN
training. At the moment, we have generated around one
thousand patterns of electrical signals which include all
the above-mentioned disturbances.
Figure 8.- Overvoltage
The detection system uses Wavelet transform of the
4.- Transients acquired signal for the generation of signal features
[5][6][7][8]. The aim of feature extraction by Wavelet
The electrical pattern generator models transients as a transforms was to provide an unique characteristic which
damped sine through a superposed exponential function, can represent every single PQ disturbance.
which is added to X(t) at a certain point.
For programming tasks we have used the MATLAB tool
which has some powerful toolboxes (specialized
functions) of signal, Wavelets and Neural Networks.

The input vectors of the ANN are generated carrying out a

number of operations on the Wavelet transform. It is
known the Wavelet transform detects better the slow
variations in the last levels and fast variations in first
levels. So, our solution is based on the concept of that the
Figure 9.- Transient amplitude disturbances would be better detected in the
first levels of Wavelet transform while the frequency ones
The implemented mathematical model obeys to: would be better detected in the last levels. Therefore, we
have used two parallel neural networks. One of them is
T ( t ) = e − at Ar sin( 2πf r t + ϕ r ) (4) used to detect the amplitude disturbances and the other
one the frequency disturbances.
• a: transitory exponent.
• Ar: amplitude of the ripple (V)
In particular, we have used the following values as input
• fr: frequency of the ripple (Hz)
vector of the amplitude neural network: the VRMS of the
• φr: initial phase of the ripple (Rad)
signal, the integral, the maximum and the VRMS of the
detail coefficients of 1,2 and 3 level.
5.- Noise

The generator makes it possible to add Additive White

Gaussian Noise (AWGN) (Figure 10) in order to simulate
more realistic signals of the power line.

Output
Capa
layer
Hidden de salida
Input
Capa Capas
layers
de layer
entrada ocultas

Figure 11.- Multilayer perceptron

Figure 10.- A signal with AWGN

183
The kind of ANN is a multilayer perceptron with 3 hidden Thus, we have developed an electrical pattern generator
layers with 20, 14 and 8 neurons respectively. We have which is capable of generating common disturbances
reached this structure after testing a great number of which can be found in a power line with the aim of
different numbers of neurons and layers. Although we making the training of neural networks easier. Besides,
have tested other kinds of structures like LVQ and RBF at we have developed a first version of a classifier system
the moment the best results have been reached with the based on ANNs.
perceptron.

In frequency we have used the following values for the

input vector of the network: the VRMS of the signal, the
integral, the maximum and the VRMS of the detail
coefficients of 6,7 and 8 level. The structure of the ANN
of frequency is a multilayer perceptron with 3 hidden
layer with 4,3 and 2 neurons respectively.

In the training of the networks we used 80% of the

generated signals as training patterns and 20% as test
patterns. In the first network we reached the best results
with a training of 180 epochs getting the right
classification of the 93.82% of the test signals. Besides,
most of the incorrectly detected disturbances were in the
threshold with other kind of disturbance. The kind of
different amplitude disturbances as well as the percent of
correctly detected disturbances in each one is shown in
table 1.
Figure 12.- Evolution of training of the frequency ANN
Table 1.- Results in amplitude disturbances
Our final objective is to implant this detector as a real-
Test Number Correctly
Disturbance time system capable of classifying disturbances on line. In
signals of errors detected %
Sag 51 2 96.08 order to get a fast system, we are currently working in
Ideal 56 6 89.29 reducing and compiling the processing of the signal
Swell 25 2 92.00 previous to ANN. We are testing other possibilities of
Undervoltage 21 0 100 processing of the signal and other structures of the ANN
Overvoltage 38 4 89.47 in order to get better the results.
Frequency deviation 52 1 98.08
Total 243 15 93.83 References
[1] M. McGranaghan, B. Roettger, Economic Evaluation
On the other hand, in the frequency network were of Power Quality, IEEE Power Engineering Review, Feb
correctly classified the 97.53% with a training of 261 2002.
epochs. With the help of the generator of disturbances, we [2] A.Hussain, M.H. Sukairi, A. Mohamed, R. Mohamed,
are currently working in the generation of new patterns Automatic Detection Of Power Quality Disturbances and
for the training of the ANN. Even when the current results Identification of Transients Signals”, International
are very good, we expect to get better them with the Symposium on Signal Processing and its Applications,
increase of the number of patterns. Kuala Lumpur, Malaysia, 13-16 August 2001.
[3] R. Adapa, Power Quality Analysis Software, IEEE
4.- Conclusions Power Engineering Review, February 2002, 0272-
1724/02.
Today it is known that neural networks are a good choice [4] W.R. Anis Ibrahim and M.M. Morcos, Artificial
for detecting and classifying electrical power Intelligence and Advanced Mathematical Tools for Power
disturbances. In the huge literature [4][5][6][7] about Quality Applications: A Survey, 668 IEEE Transactions
detection of electrical disturbances which we can find, on Power Delivery, Vol. 17, April 2002
often the problem lies in generating a sufficient number of [5] G. Zheng, M.X. Shi, D. Liu, J. Yao, Z.M. Mao, Power
training patterns to get that neural network obtains good Quality Disturbance Classification Based on Rule-Based
results in future inputs. With the help of the generator is and Wavelet-Multi-Resolution Decomposition,
possible to carry out a training of the ANN sufficiently Proceedings of the First International Conference on
entire in order to get reliable results. Machine Learning and Cybernetics, Beijing, 4-5
November 2002

184
 [6] R. Daniels, Power quality monitoring using neural Acknowledgements
networks, Proc.1st Int. Forum Applications Neural
Networks Power Syst., 1991, pp. 195–197. The work described in this paper has been supported by
[7] C. Xiangxun, Wavelet-based Measurement and the Spanish Ministry of Science and Technology (MCYT:
Classification of Power Quality Disturbances, Power Ministerio de Ciencia y Tecnología) through project
Engineering Society Winter Meeting, 2002, vol. 2, 2002. reference number DPI2002-04420-C03-03.
[8] D. Borrás, M. Castilla, N. Moreno and J.C. Montaño,
Wavelet and neural structure: a new tool for diagnostic of
power system disturbances, IEEE Trans. on Industry
Applications, Vol. 37, No. 1, pp. 184-190, 2001.

185

View publication stats