Transient Protection of Transmission Line Using Wavelet Transform

TRANSIENT PROTECTION OF TRANSMISSION LINE USING WAVELET TRANSFORM
M Solanki'", Y H Song('), S Potts(2),A Perks")
'"Brunel University, UK i2) Alstom Protection and Control, UK
Abstract: This paper presents a new approach to

developing protection techniques for EHV overhead
Source1 B u A B ~ Source2
B
transmission lines. When a fault occurs, the fault current
and voltage waveforms contain significant high frequency
transient signals. The presented method is based on the
detection of fault generated transient signals using wavelet b
analysis. The simulation study is carried out, using ATP- 190 km
EMTP and wavelet analysis, which indicates that the Figure 1: system configuration
proposed approach has the potential of developing high
speed protection relays with accurate fault detection and the wavelet components are calculated and then employed
classification. to detect and classify the faults.
Introduction Simulation of Faulted Transmission System
The main function of digital relays in power system is to An extensive series of simulation studies have been
reduce the consequences of faults by fault detection, carried out to obtain various fault signals for subsequent
localization and clearance of faults. The main power analysis. I
signal analysis tools, which are currently used in the
digital relay, have proven very useful and efficient in A typical 400 kV power transmission line system is
power system steady state analysis. Among these are (1) shown in figure 1. The simulation of this system, was
Kalman filtering based algorithms, (2) Fourier analysis carried out using ATP-EMTP software and it included
based algorithms, (3) Least squares methods based study of system for different fault types, fault positions
algorithms and (4) FIR filtering based protection. and different fault inception angles.
However in, presence of non-stationary signals, the
performance of these techniques is limited. Wavelet Transform
A more recent solution to the problem is the wavelet The wavelet transform is a recently developed
transform. F Jiang et a1 ( 1 ) states that the wavelet analysis mathematical tool for signal analysis. It has become a
has the capability of providing accurate transient very important tool for research in the field of
information in both time and frequency domain. The mathematics, physics and engineering. It transforms a
current and voltage signals obtained from a transmission time domain signal to time-scale domain. This process of
line when a fault occurs have long duration low frequency transformation is called signal decomposition because a
components and short duration high frequency signal is decomposed into several other signals with
components. Wavelet transforms appear to offer the right different level of resolution. From these decomposed
characteristics to analyze the information contained in signals, it is possible to recover the original time domain
these signals for the purpose of line protection. Wavelet signal perfectly. This reverse process is known as signal
transform has a special feature of variable time-frequency reconstruction.
localization, which is very different from windowed
.Fourier pansform. This feature can be explored as an Given a signal or a function x(t), its Continuous Wavelet
alternative to the methods mentioned above. Transform (CWT) is defined as follows:
This paper presents a technique to detect and classify
faults on EHV overhead transmission line. The wavelet 1 " t-b
transform is applied to decompose the current signals into C W x ( a , b )= J=&p)r(-)dt U (1)
a series of detailed wavelet components, each of which is
a time-domain signal that covers a specific frequency Where a and b are the scaling (dilation) and translation
band. Thus the time and frequency domain features of the (time shift) parameters respectively and y/ is the mother
transient signals are extracted. The spectral energies of wavelet function. Goswami J C & Chan A K (2) states
Developments in Power System Protection, Conference Publication No.479 0 IEE 2001 . 299
that the choice of the mother wavelet function is flexible reconstruction is also performed using the filter bank
provided that it satisfies the admissibility conditions. approach similar to the decomposition but in reverse
There are many mother wavelet functions available and order. Figure 3 shows the reconstruction of c o [ n ] .
they have different properties. The selection of mother Mathematically it can be written as:
wavelet depends on the given application.
k
d, [n]The wavelet transform has a digitally
implementable counter part, the Discrete Wavelet
Transform (DWT). The DWT of x (k) is defined as:
S
d , [n]Where the asterisk denotes a complex conjugate,
the parameters a and b in Eq.( 1) are replaced by digitised
parameters a: and k a r , also k and m are scale and
time-shift parameters. The above transformation is
performed by dilating the mother wavelet by a factor of c1[nl dSn1
two and hence it is also known as dyadic transformation. Figure 2 Wavelet decomposition of a signal
Misti M et al ( 3 ) states that the implementation of the

c, En1 dl [nl
discrete wavelet transform using a filter bank is
computationally very efficient. Refer. to figure 2. First,
the original discrete signal co[n]is decomposed into two
components, c, [n] and d ,[n] by a low-pass filter h[n]
and a high-pass filter g [ n ] , respectively. In fact, these
two filters connect the so-called 'quadrature mirror filters
(QMF) which is familiar to engineers. So the fast
transform (or multiresolution analysis) is an orthonormal
decomposition of the signal. c, [n], called the CO [nl
approximation of the signal, contains the lower frequency Figure 3 Wavelet reconstruction of a signal
components of the signal c o [ n ]while
, d , [ n ], called the
detail of the signal, contains the high frequency of the Wavelet Analysis of Fault Current Signal
c o [ n ] . Next, the approximation cl[n] is again
When a ground fault occurs, the measured current
decomposed into a new approximation c2[n]and a detail waveforms contain significant transient components.
d , [ n ] in at larger scale. Then c 2 [ n ]can be decomposed Figure 4 shows a simulated a-g fault current waveform.
to third scale, and this process can be repeated many The a-g fault occurs at time 0.025 seconds. For the
discrete wavelet analysis of the signal there are many
times. Mathematically cl[n] and d , [ n ] are defined as
wavelets available with different properties. More
follows: information can be found in Misti M et al (3).
For the protection relaying purposes the wavelet should

have properties like availability of discrete transform,
compact support, FIR filter implementation, fast algorithm
(4) and orthogonal or biorthogonal analysis. For real time
relaying purposes the speed of algorithm and easy
The implementation of signal reconstruction is fairly computation are very important. Since the wavelets,
straightforward. Strang G and Nguyen T (4) states that Mexican hat, Morlet and Meyer do not have discrete
transform and fast algorithm implementation, so they were
the original signal c,[n] can be reconstructed perfectly
not considered. The wavelets designed by Daubechies
by using the approximations and details of it. Signal have the properties required so they were investigated
300
using Matlab s wavelet toolbox. More information can asymmetric and Symmetry is a very desirable property in
be found in Daubechies (5). signal processing applications because it leads to linear
phase response. The current signal was analyzed using
Current (A) the three level discrete wavelet transform, which was
implemented in Matlab with the Sym2 wavelet and the
6000 1 1 results are shown in figures 5 , 6 , 7 and 8.
Coeff value
Time(s)
Figure 4 The a-g fault current waveform -2OOL I
0 0.01 0.02 0.03 0.04 0.05
Time(s)
Figure 5 Detail 1 ( d, [ n ]) for fault current signal
Coeff value
Wavelet flops Wavelet flops -200

Db 1 7387 Biorl. 1 8035 0 0.01 0.02 0.03 0.04 0.05
Db2 13561 Biorl.3 20833 Time(s)
I Db3 I 20293 I Biorl.5 I 36133 I Figure 6 Detail 2 ( d, [n]) for a-g fault current signal
Db4 28339 Bior2.2 21 103
Db5 36277 B ior2.4 36547 Coeff value
Db6 46123 Bior2.6 55105
Db7 5595 1 Bior2.8 76336
Db8 67669 Bior3. 1 14479 4000 @
Db9 784 15 Bior3.3 29059
DblO 92014 Bior3.5 46069
Coif5
Haar
167011
7225
13561
Bior6.8
sym5
Sym6
78442
36277
46123
-2000
-4000
1I
sym2
sym3 20293 sym7 5595 1 0 0.01 0.02 0.03 0.04 0.05
sym4 28339 Sym8 67669 Time(s)
Figure 7 Detail 3 ( d , [n]) for a-g fault current signal
301
Coeff value
1;’O
Data Sampling
I
Calculation of
wavelet coeffs I
j Yes
0 0.01 0.02 0.03 0.04 0.05 Fault classification

Time(s)
Figure 8 Approximation3 ( C, [n]) for a-g fault current Figure 9 Flow-chart for wavelet algorithm
From figures 5, 6, 7 and 8 it can be seen that the fault EHV transmission line. After detailed analysis of the
generates big spikes in the coefficient values (Coeff properties of different mother wavelets, in particular their
value), which can be used to detect and classify the faults. computation speed and symmetry, mother wavelet Sym2
The coefficient values before the fault occurs are near tumed out to be the best candidate for the given
zero but after the fault occurs the coefficient values jump application. The details and approximations up to three
to a value. The actual value is dependent on the fault levels obtained from the wavelet decomposition of the
condition. The coefficient values represent the spectral current signals are used to detect and classify the faults.
energy of the signal. Abur A and Magnago (6) states that Numerical studies clearly show promising prospects for
the energy of switching transients is negligible compared the application of wavelet analysis in power system
to the energy of the signals produced by a fault. protection. Work is on-going for fkrther developing the
approach into protection algorithms.
From the extensive analysis of many different signals, it
was found that the coefficient values for the fault signals Acknowledgement
were much higher for the fault signals than the other
transient signals without fault. Also the coefficient values The authors would like to thank Alstom Protection and
for the faulty phase is much higher than the other phases Control & EPSRC (UK) for financial support.
without fault. From these results a threshold can be set to
detect the faults and the details and approximations can be References:
compared to identify the ground faults. The proposed
algorithm for detecting and classifying the ground fault is 1. F Jiang, Bo Z Q, Weller G, Redfem M A and Chen Z,
shown in Figure 9. The fault signal from current 2000, “ Application of Wavelet in transient protection
transformers are sampled ,first, then three level discrete - review” UPEC 2000 CD-ROM
wavelet is carried out, which generates details and 2. Goswami J C., and Chan A K., 1999,
approximation of the signal at three different levels. Then “ Fundamentals of wavelets”, John Wiley & Sons Inc.
Coefficient values are compared with threshold values. If 3. Misti M, Misti Y, Oppenheim G and Poggi J M 1996
they exceed the threshold, the trip signal is sent. The fault “ Wavelet toolbox user guide”, The Mathworks Inc.
classification can be carried out by comparing the details 4. Strang G and Nguyen T., 1997
and approximation coefficients of signals of all three “Wavelets and Filter Banks”, Wellesly-Cambridge
phases. Press.
5. Daubechies I, 1992
Conclusions “ Ten Lectures on Wavelets”, Siam, Philadelphia PA
6. Abur A and Magnago F H., 2000, “Use of time delays

A wavelet based approach is presented in this paper, between modal components in wavelet based fault
which can be used for detecting and classifying faults on location”Electrica1 Power & Energy Systems, 397-
403
302

Transient Protection of Transmission Line Using Wavelet Transform

Uploaded by

Copyright:

Available Formats

Transient Protection of Transmission Line Using Wavelet Transform

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Transient Protection of Transmission Line Using Wavelet Transform

Uploaded by

Copyright:

Available Formats

TRANSIENT PROTECTION OF TRANSMISSION LINE USING WAVELET TRANSFORM

M Solanki'", Y H Song('), S Potts(2),A Perks")

'"Brunel University, UK i2) Alstom Protection and Control, UK

Abstract: This paper presents a new approach to

Misti M et al ( 3 ) states that the implementation of the

For the protection relaying purposes the wavelet should

Wavelet flops Wavelet flops -200

0 0.01 0.02 0.03 0.04 0.05 Fault classification

6. Abur A and Magnago F H., 2000, “Use of time delays

You might also like