International Journal of Recent Technology and Engineering (IJRTE)

ISSN: 2277-3878, Volume-8 Issue-1S3, June 2019

Power Quality Assessment using Change

Detection and DFT
K. Deepthi

sliding window version of FFT exhibits better results with

Abstract: Uninterrupted examination of Power Quality (PQ) regard to resolution and frequency sensitivity. For
problems in electrical power distribution network has become a non-stationary signals, STFT unable to track the mobility of
significant problem for consumers at different levels. For
the signal accurately due to rigid window width. Wavelet
maintaining accurate performance of electrical and electronic
equipment, assessment of PQ disturbances is required. This paper transform [5]-[7] is used for analyzing non-stationary
introduces a new technique for PQ disturbances detection and signals. Both time and frequency domains information are
classification with minimum computational complexity. The extracted using WT. The accuracy of WT depends on the
presence of disturbance in the sinusoidal signal is identified using wavelet chosen. It deteriorates from the drawbacks of
Voltage slope detection method and the type of disturbance is computational complication and affected by the noise level.
classified using Discrete Fourier Transform (DFT). Results
Using SLT [8], single and multiple disturbances are
proves that the proposed method is optimum and the overall
accuracy is high compared to other methods. identified. But due to the presence of noise in the signals, the
performance of SLT technique is degraded. A combination
Keywords: Classification, Disturbance detection, Discrete of Kalman filter and fuzzy expert system [9] has been
Fourier Transform, Power Quality (PQ), PQ Disturbances, signal proposed for PQ analysis. The solution of KF are based on
model. model of the scheme used and choice of filter parameters.
S-transform [10]-[12] is a varying window of STFT or an
I. INTRODUCTION extension of WT. It depends on the scalable localizing
Power Quality (PQ) has eventually be a big task for Gaussian window and presents the frequency dependent
customers. Widening usage of non linear loads such as resolution. ST has the capacity to observe the disturbance in
power converters, lightning controls, adjustable speed drives the existence of noise also. But the disadvantage is its
and arcing devices contribute to poor PQ [1]. Poor PQ may complicated calculation, because every time the frequency is
cause equipment deterioration and reduce the life of load. To multiplied by gaussian window and then by inverse Fourier
provide good power quality, the power must be free form transform. In [13], Empirical mode decomposition (EMD)
disturbances. Power Quality is defined as the study of with Hilbert transform (HT) for PQ assessment is used. From
powering and grounding sensitive electronic equipment in a non-stationary and non-linear signals, EMD technique
manner suitable for the equipment [2]. Electrical equipments generates Intrinsic mode functions (IMF). The major
such as a motor, a transformer, a desktop computer, or a drawback is mode mixing, where the wanted data will be
refrigerator are inclined to deficiency when exposed to single reduced.
or multiple PQ disturbances. In a power distribution In this paper, a new power quality assessment technique
network, some PQ disturbances like surges, voltage sag or is proposed for identification and classification of PQ
dip,undervoltage, voltage swell, interruption, transients, problems. To detect the disturbance present in the sinusoidal
flicker, harmonics and voltage unbalance result in signal, a new Voltage Slope Detection technique is proposed.
deterioration of quality of electric power. These are mainly This method detects the disturbances within a short period of
caused by faults, switching of heavy loads and non-linear time. Using one of the Phasor estimation technique, Discrete
loads [1]. These PQ disturbances weaken the performance of Fourier Transform (DFT), the features of different PQ
customer equipment. So, PQ improvement has a positive disturbances are obtained and are classified. The advantage
impact on the distribution system for its continuous profits. of the proposed method is its less computational complexity,
IEEE deals with the definition, detection, mitigation and because it detects the disturbance in the signal when there is
classification of PQ events [2]. Digital signal processing change in the signal only. The proposed method is proved
tools like Fourier Transform (FT), Short Time Fourier with signals having varying amplitude, time and frequency to
Transform (STFT), Slantlet Transform (SLT), Wavelet show their effectiveness under noisy conditions. Results
Transform (WT), Hilbert transform, Kalman Filter (KF) and demonstrate the robustness of this method.
Stockwell Transform (ST) are used to abstract frequency and This paper is standardized into four sections. Section II
time information of different PQ disturbances. Fourier describes the representation of PQ signal model. Proposed
transform [3] provides only the frequency resolution and it is algorithm for power quality assessment is discussed in
not suited for non-stationary signals.STFT [4] well known as Section III. In Section IV, MATLAB results of the proposed
technique and comparison results are shown. Finally, in
Section V Conclusion is
Revised Manuscript Received on June 01, 2019. presented.
K.Deepthi, Department of Electrical & Electronic Engineering, Sagi
RamaKrishnam Raju Engineering College, Bhimavaram, Andhra Pradesh,
India (Email:[email protected])

Published By:
Blue Eyes Intelligence Engineering
Retrieval Number: A10190681S319/19©BEIESP 100 & Sciences Publication
 Power Quality Assessment using Change Detection and DFT

II. SIGNAL MODEL The proposed method has two stages:

In this section, modelling of simulated power quality
A. Disturbance Detection
disturbances(PQD) are described. Various PQD are
generated in MATLAB software using parametric equation Voltage slope detection is a new method to detect the
model or mathematical model. In this paper, using disturbance present in the signal. It is a process that measures
parametric equation models eleven types of PQD are created how the characteristics of a signal have changed between two
as shown in Table I. Advantages of using parametric or more time periods. In this method, the difference of
equations are the parameters of the signal can be varied voltage between two values of sample is considered for the
easily in a controlled manner and are used for real time detection of disturbance. The pure value of the sum of the
classification. PQD like undervoltage, overvoltage, voltage slopes of the voltage signal for a window of oneentire cycle
sag or dip, spike, voltage swell, interruption, notch and isallotted as disturbance index. Under normal conditions, the
transients are created by applying step functions for duration disturbance index value is zero. At the initiation of
between 𝑡1 and 𝑡2 . Harmonic signals have greater than three disturbance, the disturbance index raises to a bigger value.
frequency elements and are uneven multiples of primary This value of disturbance index is tested with a prearranged
frequency element. Amplitudes of harmonic signal given by threshold parameter.
parameters 𝛼1 , 𝛼3 , 𝛼5 and 𝛼7 are varied from 0.05 to 0.15 Mathematically this can be written as,
and square of sum of all these elements not to go beyond 𝑙(𝑖) = (𝑣(𝑖) − 𝑣(𝑖 − 1))/∆𝑡 (1)
unity magnitude [14]-[17]. Transients have rapid changes 𝑆(𝑖) = ∑𝑖𝑗=𝑖−𝑁+1 𝑙(𝑗) (2)
which occur for a small period of time is controlled by
parameter τ. The parameters 𝜶 represents the magnitude of At the instant i, the slope of the signal is 𝑙(𝑖), sampling
different PQD and 𝑡2 − 𝑡1 represents the duration of interval is given by ∆𝑡, 𝑆(𝑖) is the value of disturbance index
disturbances as presented in Table I. By varying parameters using the proposed method, N is number of samples per cycle
𝜶, 𝑡1 and 𝑡2 , a large number of PQ disturbance signals can be and 𝑣(𝑖) is the present voltage signal sample value.
generated. The signals produced by parametric equation The disturbance is identified,
model can be easily used in the detection and classification if |𝑆(𝑖)| > 𝑇1 (3)
of PQD.
Where 𝑇1 is the threshold value.
III. PROPOSED ALGORITHM
TABLE I
SIGNAL MODELAND THEIR PARAMETERS [16]
PQ Disturbance Signal Model Parameters

Pure sine 𝑣(𝑡) = 𝐴 sin(𝑤𝑡) A = 1(pu), f=50 Hz

Undervoltage 𝑣(𝑡) = 𝐴 (1 − 𝛼𝑢 (𝑢(𝑡 − 𝑡1 ) − 𝑢(𝑡 − 𝑡2 ))) sin(𝑤𝑡) 0.1≤𝛼𝑢 ≤0.9, T≤𝑡2 − 𝑡1 ≤ 9𝑇

Overvoltage 𝑣(𝑡) = 𝐴 (1 + 𝛼𝑜 (𝑢(𝑡 − 𝑡1 ) − 𝑢(𝑡 − 𝑡2 ))) sin(𝑤𝑡) 0.1≤𝛼𝑜 ≤0.8, 𝑇 ≤ 𝑡2 − 𝑡1 ≤ 9𝑇

Interruption 𝑣(𝑡) = 𝐴 (1 − 𝛼𝑖 (𝑢(𝑡 − 𝑡1 ) − 𝑢(𝑡 − 𝑡2 ))) sin(𝑤𝑡) 0.9≤𝛼𝑖 ≤1.0, 𝑇 ≤ 𝑡2 − 𝑡1 ≤ 9𝑇

Flicker 𝑣(𝑡) = 𝐴(1 + 𝛼𝑓 sin(𝛽𝑤𝑡))sin(𝑤𝑡) 0.1≤𝛼𝑓 ≤0.2, 5 ≤ 𝛽 ≤ 20 𝐻𝑧

Voltage sag or dip 𝑣(𝑡) = 𝐴 (1 − 𝛼𝑑 (𝑢(𝑡 − 𝑡1 ) − 𝑢(𝑡 − 𝑡2 ))) sin(𝑤𝑡) 0.1≤𝛼𝑑 ≤0.9, 𝑇 ≤ 𝑡2 − 𝑡1 ≤ 9𝑇

Voltage swell 𝑣(𝑡) = 𝐴 (1 + 𝛼𝑠 (𝑢(𝑡 − 𝑡1 ) − 𝑢(𝑡 − 𝑡2 ))) sin(𝑤𝑡) 0.1≤𝛼𝑠 ≤0.8, 𝑇 ≤ 𝑡2 − 𝑡1 ≤ 9𝑇

Oscillatory transients 𝑣(𝑡) = 𝐴(sin(𝑤𝑡) + 𝛼𝑡 𝑒 −

𝑡−𝑡1
𝜏 𝑠𝑖𝑛𝑤𝑛 (𝑡 − 𝑡1 ){𝑢(𝑡2 ) − 𝑢(𝑡1 )}) 0.1≤𝛼𝑡 ≤0.8,0.5𝑇 ≤ 𝑡2 − 𝑡1 ≤ 3𝑇

Harmonics 𝑣(𝑡) = 𝐴(𝛼1 sin(𝑤𝑡) + 𝛼3 sin(3𝑤𝑡) + 𝛼5 sin(5𝑤𝑡) + 𝛼7 sin(7𝑤𝑡)) 0.05≤𝛼3 , 𝛼5 , 𝛼7 ≤ 0.15, ∑ 𝛼𝑖2 = 1

Notch 𝑣(𝑡) = 𝐴(sin(wt) − 𝑠𝑖𝑔𝑛(sin(𝑤𝑡)) × 0.1 ≤ 𝐾 ≤ 0.4,0 ≤ 𝑡1 , 𝑡2 ≤ 0.5𝑇,

9
0.01𝑇 ≤ 𝑡2 − 𝑡1 ≤ 0.05𝑇
[∑ 𝐾 × {𝑢(𝑡 − (𝑡1 + 0.02𝑛)) − 𝑢(𝑡 − (𝑡2 + 0.02𝑛))}])
𝑛=0

Spike 𝑣(𝑡) = 𝐴(sin(𝑤𝑡) + 𝑠𝑖𝑔𝑛(sin(𝑤𝑡)) × 0.1 ≤ 𝐾 ≤ 0.4,0≤ 𝑡1 ,𝑡2 ≤ 0.5𝑇,

9
0.01𝑇 ≤ 𝑡2 − 𝑡1 ≤ 0.05𝑇
[∑ 𝐾 × {𝑢(𝑡 − (𝑡1 + 0.02𝑛)) − 𝑢(𝑡 − (𝑡2 + 0.02𝑛))}])
𝑛=0

Published By:
Blue Eyes Intelligence Engineering
Retrieval Number: A10190681S319/19©BEIESP 101 & Sciences Publication
 International Journal of Recent Technology and Engineering (IJRTE)
ISSN: 2277-3878, Volume-8 Issue-1S3, June 2019

IV. RESULTS
The proposed power quality assessment method
presented is designed to detect and classify the eleven PQ
disturbances such as undervoltage, overvoltage, voltage sag
or dip, voltage swell, flicker, interruption, harmonics,
oscillatory transients, notch and spike. The fundamental
frequency of each signal is 50 Hz and the sampling
frequency is 3200Hz, i.e., 64 samples for each cycle [2].
Fig. 1. Voltage slope detection example.
Considering the overall duration of the signals as 0.2 seconds
A suitable threshold value is essential for disturbance
except for long duration voltage variations. The input signal
detection because of noise, sudden load changes as faults and
is voltage signal, the disturbance is detected by voltage slope
other uncertainties. Fig.1 shows the example of disturbance
detection method and then processed for classification using
detection using voltage slope detection method.
phasor estimation technique DFT by calculating their
B. Phasor Estimation Technique magnitude.
Discrete Fourier transform is a phasor estimation Fig. 3. shows the amplitude versus time of normal
technique used for reconstructing discrete time rule voltage signal with maximum amplitude of 1pu.Considering
information into its equivalent discrete frequency rule the duration greater than 1 min for disturbances like
information. It is widely used in digital communication, undervoltage and overvoltage are shown in Figs. 3-4, where
signal processing, linear filtering, spectral analysis and the maximum amplitude obtained is 0.5 pu for undervoltage
wireless communication. Using DFT, the amplitude and and 1.5 pu for overvoltage [1]. Figs. 6(a)-13(a), clearly
phase angle of distinctive signal frequency components are describes the amplitude versus time of different PQ
determined. disturbances generated by using signal models as shown.
for the given series, the M-pointDFT is shown as
−𝑗2𝜋𝑚𝑖
⁄𝑀
𝑥(𝑖) = ∑𝑁−1
𝑚=0 𝑥(𝑚) 𝑒 (4)
Where i=0,1,2,………..M-1
Here, the discrete time rule signal is expressed as 𝑥(𝑚)asm
be the discrete time rule index and the frequency rule
components is taken as 𝑥(𝑖) where i is the normalized
frequency rule index. Benefits of using DFT are high speed
operation, high efficiency and low computational
complication. Fig. 3. Pure sine
The flowchart for the detection and classification of power
quality disturbances is shown in Fig. 2.

Fig. 4. Undervoltage.

Fig. 5. Overvoltage.

(a)

Fig. 2. Flowchart of the proposed algorithm

In the present work, DFT is used for the
classification of eleven PQ disturbance signals. By observing
the magnitude and duration in the plots, the type of
disturbance is classified as per definitions[1].

Published By:
Blue Eyes Intelligence Engineering
Retrieval Number: A10190681S319/19©BEIESP 102 & Sciences Publication
 Power Quality Assessment using Change Detection and DFT

(b) (a)

Fig. 6.(a) Interruption. (b) Detection occurs at time, t=0.05 sec.

(a)
(b)

Fig. 10.(a) Oscillatory transients. (b) Detection occurs at time, t=0.05 sec.

(b)

(a)

Fig. 7.(a) Flicker. (b) Detection occurs at time, t=0.05 sec.

(a) (b)

Fig. 11. (a) Harmonics. (b) Detection occurs at time, t=0.05 sec.

(b)
(a)

Fig. 8. (a) voltage sag. (b) Detection occurs at time, t=0.05 sec.

(a)
(b)

Fig. 12. (a)Notch. (b) Detection occurs at time, t=0.05 sec.

(b)

Fig. 9.(a) Voltage swell. (b) Detection occurs at time, t=0.05 sec.

Published By:
Blue Eyes Intelligence Engineering
Retrieval Number: A10190681S319/19©BEIESP 103 & Sciences Publication
 International Journal of Recent Technology and Engineering (IJRTE)
ISSN: 2277-3878, Volume-8 Issue-1S3, June 2019

classified using DFT by observing their magnitude. Figs.

6(b)-13(b) show the results of disturbance detection method,
by observing the index value and the time at which
disturbance is identified. When there is no disturbance, the
(a) disturbance index value is zero. By observing the figures, the
disturbance occurs at time t=0.05sec, for signals interruption,
flicker, voltage sag, voltage swell, oscillatory transients,
harmonics, notch and spike.
A total of 275 signals with 25 signals of each class are
generated using signal models by varying parameters such as
starting time, end time, amplitude and frequency. In the case
of signals with amplitude less than unity, the overall
accuracy is reduced to 96% due to disturbances flicker and
(b) spike.This is due to rapid changes in voltage signal, these
flicker and spike signals may be misclassified. Otherwise,
the proposed method works well for these disturbance
signals also.
Fig. 13.(a) Spike. (b) Detection occurs at time, t=0.05 sec.

These figures clearly illustrates the magnitude and time of its

occurrence. As per definitions[1], the type of disturbance is
TABLE II
RESULTS OBTAINED USING PROPOSED METHOD, WITHOUT NOISE
PQ disturbances Accuracy(%)

Starting time End time Amplitude Frequency Magnitude

Pure sine 100 100 100 100 100

Undervoltage 100 100 100 100 100
Overvoltage 100 100 100 100 100
Interruption 100 100 100 100 100
Flicker 100 100 96 100 100
Voltage sag 100 100 100 100 100
Voltage swell 100 100 100 100 100
Oscillatory transients 100 100 100 100 100
Harmonics 100 100 100 100 100
Notch 100 100 100 100 100
Spike 100 100 96 100 100
Overall accuracy 100 100 99.27 100 100
Overall accuracy=99.85

TABLE III TABLE IV

RESULTS OBTAINED USING PROPOSED METHOD, WITH NOISE THE OVERALL ACCURACYOF THE PROPOSED METHOD
PQ Accuracy(%) PQ No. of Disturbance Classification
Disturbances Disturbances cases detection accuracy(%)
SNR 30dB SNR 40 dB SNR 50 dB accuracy(%)
Pure sine 100 100 100 Pure sine 25 100 100
Undervoltage 100 100 100 Undervoltage 25 100 100
Overvoltage 100 100 100 Overvoltage 25 100 100
Interruption 100 100 100 Interruption 25 100 100
Flicker 100 100 100 Flicker 25 100 96
Voltage sag 100 100 100
Voltage sag 25 100 100
Voltage swell 100 100 100 Voltage swell 25 100 100
Oscillatory 100 100 100
Oscillatory 25 100 100
transients
transients
Harmonics 100 100 100
Harmonics 25 100 100
Notch 100 100 100
Notch 25 100 100
Spike 100 100 100 Spike 25 100 96
Overall accuracy=100 Total=275 Overall accuracy=99.63

Published By:
Blue Eyes Intelligence Engineering
Retrieval Number: A10190681S319/19©BEIESP 104 & Sciences Publication
 Power Quality Assessment using Change Detection and DFT

Wavelet-Multiresolution Signal Decomposition”, IEEE Transactions on

TABLE V Power Delivery, Vol. 14(4), pp. 1469-1176(1999).
COMPARISON RESULTS OF PROPOSED METHOD WITH OTHER 8. G. Panda, P. K. Dash, A. K. Pradhan, and S. K. Meher “Data
METHODS Compression of Power Quality Events Using the Slantlet Transform”,
References No. of PQ Overall IEEE Transactions on Power Delivery, Vol. 17, No. 2, April 2002.
Disturbances accuracy(%) 9. A. Abdelazeem, A. AbdelsalamAzza, “Classification of power system
disturbances using linear Kalman filter and fuzzy-expert system”
ST+FUZZY [18] 07 98 Electrical Power and Energy Systems 43 (2012) 688-695.
10. P. K. Dash, B. K. Panigrahi, G. Panda, “Power quality analysis using
ST+PNN [19] 11 97.4 S-transform”, IEEE Trans. Power Deliv. 2003, 188, 406-411.
11. Raj Kumar, Bhim Singh, Ambrish Chandra, and Kamal Al-Haddad
ST+PNN [12] 10 94.7
“Recognition of Power Quality Disturbances Using S-Transform-Based
ST+FUZZY [20] 13 92.7 ANN Classifier and Rule-Based Decision Tree”, IEEE Trans. Ind.
Electron., vol. 50, no. 2, pp. 1249-1258, March/April 2015.
Proposed method 11 99.63 12. I. W. C. Lee and P. K. Dash, “S-transform-based intelligent system for
classification of power quality disturbance signals”, IEEE Trans. Ind.
Electron., vol.50, no. 4, pp. 800-805, Aug. 2003.
The overall accuracy is 99.85% for all eleven types of 13. S. Shukla, S. Mishra, B. Singh, “Empirical-Mode Decomposition with
disturbance signals without noise as presented in Table II. To Hilbert Transform for Power-Quality Assessment”, IEEE Transactions
know the performance of the proposed method under noisy on Power Delivery, vol. 24, no. 4, pp. 2159-2165, Oct.2009.
14. Nermeen Talat, W. R. Ibrahim, George L. Kusic, “New Technique for
condition, the signals are combined with white Gaussian Categorization of Power Quality Disturbances”, Power Quality and
noise of signal to noise ratio (SNR) varying from 30 to 50 Supply Realibility Conference, 2008, IEEE, pp. 11-16.
dB. Table III demonstrates the results obtained for signals at 15. B. K. Panigrahi, V. R. Pandi, “Optimal feature selection for
classification of power quality disturbances using wavelet packet based
different SNR using the proposed method and the overall fuzzy k-nearest neighbor algorithm”, IET Gener. Transm. Distrib.,
accuracy is 100%. For a complete set of 275 signals, the 2009, Vol. 3, Issue. 3, pp. 296-306.
individual and overall accuracy results are shown in Table 16. R. H. G. Tan and V. K. Ramachandramurthy, “Numerical model
framework of power quality events,” Eur. J. Sci. Res., vol. 43, no. 1, pp.
IV. In this, the overall accuracy of disturbance detection is 30-47, Jun. 2010.
100% and the overall classification accuracy is 99.27%. 17. W. P. Khaing, T. Zin and H. M. Tun, “Detection and Localization of
Table V show the proposed method accuracy comparison Electrical Power Quality Disturbances”, International Journal of
Engineering Trends and Technology, Vol. 11, no.4, May. 2014.
with other methods[18]-[20]. 18. M. V. Chilukuri and P. K. Dash, “Multiresolution S-transform-based
The proposed method is able to detect and classify the fuzzy recognition system for power quality events,” IEEE Trans. Power
11 types of PQ disturbances with an accuracy of 99.63% Del., vol. 19, no.1, pp. 323-330, Jan. 2004.
19. S.Mishra, C. N. Bhende and B. K. Panigrahi, “Detection and
which is higher than other methods. The proposed method classification of power quality disturbances using S-transform and
results obtained show that this method is suitable for probabilistic neural network,” IEEE Trans. Power Del., vol. 23, no.1
disturbance detection and the classification of PQ signals. pp. 280-287, Jan. 2008.
20. M. Biswal and P. K. Dash, “Detection and characterization of multiple
power quality disturbances with a fast S-transform and decision tree
V. CONCLUSION based classifier,” J. Digit. Signal Process., vol. 23, no.4, pp. 1071-1083,
Jul. 2013.
A new power quality assessment technique based on
voltage slope detection and Discrete Fourier Transform has AUTHORS PROFILE
been developed for PQ disturbance signals detection and
classification. It is observed that this method correctly K.Deepthi is a student of Masters of
Technology in Power systems and
detects and classifies the eleven types of PQ disturbances Automation, JNTUK University, Andhra
with higher accuracy, in the presence of noise environment Pradesh, India. She has received the Bachelor
also. The overall accuracy obtained is 99.63% which is of Technology degree in the department of
Electrical and Electronics Engineering from
higher than other methods. The advantage of this method is JNTUK University, Andhra Pradesh in the
less computational complexity. The obtained results show year 2017. Her research interest includes
the best performance and high accuracy of the proposed Power Quality, Power System Optimization
techniques, Flexible AC transmission systems, Digital Signal Processing
method. and AI techniques.

REFERENCES
1. R. C. Dugan, M.MC Granaghan, H. Beaty “Electrical power system
quality”, McGraw-Hill companies, 1996.
2. IEEE Recommended Practice for monitoring Electric Power Quality,
IEEE Std. 1159, 2009.
3. Emanuel Fuchs, BetiTrajanoska, Sarah Orhouzee, and Herwig Renner,
“Comparison Of Wavelet and Fourier Analysis in Power Quality”,
IEEE certified Journal, 2012.
4. Y. H. Gu and M. H. J. Bollen, :Time-frequency and time-scale domain
analysis of voltage disturbances”, IEEE Trans. Power Del., vol. 15, no.
4, pp. 1279-1284, oct. 2000.
5. F. Jurado, N. Acero, B.Ogayar, “Application of signal processing tools
for power quality analysis:, Canadian Conference on Electrical and
Computer Engineering, IEEE CCECE 2002. Vol. 1, 12-15 May 2002.
6. S.Santoso, E. J. Powers, W. M. Grady, and P. Hofmann, “Power quality
assessment via wavelet transform analysis”, IEEE Trans. Power Del.,
vol. 11, no. 2, pp. 924-930, Apr. 1996.
7. A. M. Gaouda, M. M. A. Salama, M. R. Sultan, and A. Y. Chikhani,
“Power Quality Detection and Classification Using

Published By:
Blue Eyes Intelligence Engineering
Retrieval Number: A10190681S319/19©BEIESP 105 & Sciences Publication