On-line Discrete Wavelet Transform in EMTP

Environment and Applications in Protection

Relaying
N. Perera, A.D. Rajapakse and R.P. Jayasinghe

Abstract--This paper describes the development of an on-line multi-resolution properties of wavelets transform make them
discrete wavelet transform tool for an electromagnetic transient well suited to analyze transient signal superimposed on a
simulation program. Multi-resolution properties of wavelet continuous fundamental.
transform make it ideally suitable for analyzing power system
Due to the wide variety of signals and problems
transient signals which consist of non-periodic high frequency
oscillations superimposed on power frequency signal. New power encountered in power engineering, there are various
system devices such as power quality monitors and protective applications of wavelet transform. These include detection,
relays based on algorithms involving wavelet transformation are and analysis of power quality disturbances and power quality
emerging. Thus, it is highly useful for power system data compression [4], [5], high voltage insulation condition
electromagnetic transient simulation programs to have monitoring [6], fault detection [7], [8], and disturbance
integrated capability for wavelet transformation. This paper also classification [9]. Several power engineering products based
briefly presents several applications of wavelet transformation in
power system protection and power quality monitoring.
on wavelet transformation such as protective relays [10] and
power quality monitors [11] are now emerging.
Keywords: Electromagnetic transient simulation, Wavelet Electromagnetic transient programs (emtp-type programs)
transforms, Applications of wavelet transform in power system, are widely used for power system studies involving power
Power system protection, Power quality. quality issues, protection system operation, etc. An integrated
tool that can be used to perform wavelet transformation of the
I. INTRODUCTION simulated waveforms is a highly useful feature for those
studies investigating wavelet transformation based techniques.
W avelets are mathematical functions that decompose a
signal into different frequency components, and then
study each component with a resolution matched to its scale
Although several wavelet transform programs such as
MATLAB® wavelet toolbox are available, their use generally
[1]. They have added advantages over traditional Fourier requires simulated waveforms to be saved in data files and
methods because wavelet transformation localizes information then perform the analysis external to the emtp-type simulation.
in the time-frequency plane; and capable of trading frequency The novelty of the proposed component is that it can be used
resolution with time resolution and vice-versa. These to perform the wavelet transform as the waveforms are
properties have made Wavelets transformation highly suitable generated by the simulation. Thus it is possible to simulate
for analyzing physical situations where the signal contains systems such as protective relays based on wavelet
discontinuities and sharp spikes [1],[2]. transformation in a closed loop manner.
Waveforms associated with fast electromagnetic transients This paper is organized as following. Section II gives a
are typically non-periodic and contains both high frequency brief introduction to the wavelet transformation for the benefit
oscillations and localized impulses superimposed on power of readers not familiar with wavelet transformation. Then the
frequency and its harmonics. These characteristics present a implementation of wavelet transformation using filter banks is
problem for traditional Fourier analysis because its use explained in Section III. A brief description of the usage and
assumes a periodic signal and because a wide-band signal capabilities of the wavelet tool developed is given in Section
requires denser sampling and longer time periods to maintain IV. Section V presents several application examples, and
good resolution in the low frequencies [3]. On the other hand, conclusions are given in Section VI.

II. WAVELET TRANSFORMATION

This work was supported by Manitoba HVDC Research Centre, Manitoba,
Canada. The wavelet transformations can be Continuous Wavelet
N. Perera and A.D. Rajapakse are with Department of Electrical and Computer Transformation (CWT) or Discrete Wavelet Transformation
Engineering, University of Manitoba, Canada (e-mail:
[email protected], [email protected]). (DWT). If f(t) is a signal with a finite energy, its CWT is
R.P. Jayasinghe is with Manitoba HVDC Research Centre (e-mail: defined as
∞
[email protected]). (1)
CWTΨ f ( a , b) = ∫ f (t )Ψ
*
a ,b ( t ) dt
Presented at the International Conference on Power Systems −∞

Transients (IPST’07) in Lyon, France on June 4-7, 2007 where,

 Ψ ( t −ab )
− 12 III. FILTER BANK IMPLEMENTATION OF DWT
Ψa ,b (t ) = a (2)

A. Decomposition and Reconstruction

The function Ψ(t) is the basis function or the mother
DWT can be implemented efficiently as a filter bank as
wavelet, the asterisk denote a complex conjugate, and a (≠0,
shown in Fig. 2 [1],[2]. This implementation is commonly
∈ℜ) is the scale parameter and b (∈ℜ) is the translation
known as Mallat tree algorithm and consists of series of low-
parameter. The mother wavelet function must satisfy several
pass filters and their dual high pass filters. H(n) denotes a high
conditions: it should be short and oscillatory, i.e. it must have
pass filter and L(n) is its dual low pass filter. The filter
zero average and decay quickly at both ends. Several
coefficients are determined by the type of mother wavelet
examples of wavelets are shown in Fig.1.
selected. The circle with downward arrow behind 2 denotes
down sampling by a factor of 2. The outputs d1, d2, d3, etc. are
called the detail wavelet coefficients while the output from the
last low pass filter is referred to as the approximation wavelet
coefficient.
It is possible to obtain the original signal f(t) through
wavelet series reconstruction. The reconstruction can also be
carried out efficiently using a tree algorithm as shown in Fig.
3. The filters H’(n) and L’(n) are the inverse filters of H(n)
and L(n) respectively. In Fig. 3, the circles with upward arrow
behind 2 denotes up sampling by a factor of 2.

Level-1
Fig. 1. Different types of wavelets 2 d1 Level-2
H(n)
2 d2 Level-3
f(n) H(n)
2 d3
In discrete wavelet transformation, mother wavelet is L(n) 2 H(n)
L(n) 2
dilated and translated discretely by selecting L(n) 2
a = aom ; b = nbo aom (3)
Fig. 2. Mallat tree algorithm for wavelet decomposition
where ao (>1) and bo (>0) are fixed real values and m and n
are positive integers. Then the discretized mother wavelet Level-1
d1 2
Level-2 H’(n)
becomes Level-3
d2 2 H’(n)
f(n)
1 t − nbo aom (4) d3 2 H’(n) 2 L’(n)
Ψ (t ) =
m,n Ψ( ) 2 L’(n)
aom aom
2 L’(n)
The corresponding discrete wavelet transformation is given
by Fig. 3. Wavelet reconstruction
DWTΨ f (m, n) = ∑ f (k )Ψ*m,n (k ) (5)
k
B. Multilevel Decomposition and Reconstruction
DWT provides a decomposition of a signal into sub bands The reconstruction algorithm can be used to provide the
with a bandwidth that increases linearly with frequency. In the wavelet coefficients of different scales a finer time resolution.
case of dyadic transform corresponding to ao = 2 and bo = 1, Fig. 4 illustrates this process: decomposition of sampled signal
the result is geometric scaling, i.e. 1, 1/a, 1/a2… and x(n) for three levels and use of up sampling and filtering to
translation by 0, n, 2n… This scaling gives the DWT obtain finer reconstruction wavelet coefficients. The original
logarithmic frequency coverage in contrast to the uniform signal x(n) can now be easily reconstructed simply by adding
frequency coverage of Fourier transformation. the reconstruction wavelet coefficients: the detail coefficients
D1(n), D2(n), D3(n) and approximation coefficient A3(n).
Note that all these coefficients now have the same sampling
rate as the original signal.
H(n) 2 2 H’(n) D1(n)

x(n) H(n) 2 2 H’(n) 2 L’(n) D2(n)

L(n) 2 H(n) 2 2 H’(n) 2 L’(n) 2 L’(n) D3(n)

L(n) 2

L(n) 2 2 L’(n) 2 L’(n) 2 L’(n) A3(n)

Fig. 4. Multilevel wavelet reconstruction

 IV. WAVELET TRANSFORM TOOL IN EMTP ENVIRONMENT types of mother wavelets have been implemented: Harr,
Simulation of devices that use wavelet transform based Daubechies (DB) (order 1, 2, 4, and 8), Symlets (Sym) (order
techniques is essential for research and development as well as 1, 2, 4, and 8), Coeiflets (order 1 and 2) (Harr, Daubechies
for validating the suitability of those new devices for practical order1, and Symlet order 1 are essentially the same). In
applications. Since emtp-type programs are widely used for addition, a user can specify the level of details computed. This
studies concerning power quality issues, protection, and selection requires a change in the dimension of the output.
transients in power systems, the capability of applying wavelet The total number of coefficients calculated is equal to the
transformation in emtp environment itself will provide many number of detail levels plus one (for the approximation
advantages. Although there are several free/commercial component).
software tools available for wavelet analysis, they can be used B. Validation
only for post simulation analyses using the waveforms saved The accuracy of computation is validated by comparing
to data files. This is not obviously a convenient way, with MATLAB® wavelet toolbox, which is one of the most
especially when simulations are needed to carry out widely used tools for wavelet analysis. Fig. 6 shows a sample
repetitively. On the other hand, since the analysis has to be of results which compares the wavelet coefficients of a signal
carried out after completing the simulations, simulation of with transient (top most graph) obtained through MATLAB
devices that use wavelet transform cannot be performed. The and the DWT tool developed in PSCAD/EMTDC.
online wavelet transformation tool developed here enable
80
simulating systems such as protective relays based on wavelet Signal

transformation in a closed loop manner. 0

-40
A. DWT Tool 50
Approximation A6
In the present study, an online wavelet transformation tool 0
was developed in PSCAD/EMTDC® software. However, this -50
can be implemented in any other emtp-type program. The 20
Detail D1 Δt
block diagram in Fig. 5 shows the processing steps involved. 0
-20
Input signal 20
Detail D2
0
Anti-aliasing filter -20
50
Detail D3
0

Sampling -50
20
Detail D4
0
Data buffer -20
20
Detail D5
0
Wavelet -20
decomposition 50
Detail D6
Filter coefficients 0
database
-50
0.06 0.08 0.1 0.12 0.14 0.16 0.18
Wavelet reconstruction Time (s)

Fig. 6. Comparison of the wavelet coefficients calculated using MATLAB

wavelet tool box (thin lines), and the PSCAD DWT tool (thick lines) for a
(Approximation) (Details) sample waveform. The mother wavelet used is Sym8. Note that the online
Reconstruction wavelet coefficients calculation in PSCAD results in a time lag (Δt).

Fig. 5. Processing steps involved in the DWT tool TABLE I

DATA WINDOW SIZES (NUMBER OF SIGNAL SAMPLES) REQUIRED FOR ONLINE
DISCRETE WAVELET TRANSFORM.
As applications may need signal sampling intervals that are
different from the simulation time step, a provision is allowed Level Haar Db2/sym2 Db4/sym4 Db8/sym8 Coif1 Coif2
for re-sampling the input signal at a frequency selected by the 1 2 6 14 30 10 22
user. An anti-aliasing filter is also provided for filtering out 2 4 14 34 74 24 54
high frequency noise. The cutoff frequency of the anti-aliasing 3 8 30 74 162 52 118
4 16 62 154 338 108 246
filter is automatically determined based on the sampling
5 32 126 314 690 220 502
frequency selected. A provision is provided to disable the anti- 6 64 254 634 1394 444 1114
aliasing filtering if required. Sampled data is placed in a buffer
before the decomposition and reconstruction is performed. For the case shown in Fig. 6, Sym8 mother wavelet was
User can specify the type of mother wavelet. Currently nine used. Shown in thicker lines are the results from the
PSCAD/EMTDC online DWT tool (transformation was For an external fault such as F2, wavelet coefficients of the
performed as the signal being generated from the simulation). transients currents measured on the non faulted branches are
MATLAB results were obtained at the end of simulation by opposite in sign compared to those measured on the faulted
using a recorded waveform. Except for the time delay, both branch. This sign difference can be clearly observed in WTCs
curves were found to be identical. Time delay, which cannot shown in Fig. 9, and can be used to determine the direction of
be avoided in online calculation, is dependent on the data fault.
window size (number of signal samples) required to perform If such relays can be installed at strategic locations on a
the calculation. Table-1 gives the data window sizes required power network and communication is provided between the
for different types of mother wavelets at different detail levels neighboring relays, faulted segments can be determined very
to perform online DWT. fast: all the information required is obtained within half a
cycle. Application of such a scheme for a 230 kV, 12 bus
V. APPLICATIONS OF WAVELET TRANSFORM TOOL transmission network is shown in Fig. 10.

A. Rapid Isolation of Faults Measurements at X Measurements at Y Measurements at Z

Phase currents[kA]

Phase currents[kA]

Phase currents[kA]
1 1 1
Wavelet transform can be used to quickly identify the
0 0 0
direction of fault currents using initial transients in the
-1 -1 -1
currents (or voltages) due to the fault [12]. In order to explain
-2 -2 -2
the principle, consider a relay installed at a busbar 0.5 0.505 0.51 0.5 0.505 0.51 0.5 0.505 0.51
Time [sec] Time [sec] Time [sec]

Aerial mode-1 currents[kA]

Aerial mode-1 currents[kA]

Aerial mode-1 currents[kA]

interconnecting three lines segments as shown in Fig.7.
0 0 0

-2 -2 -2

-4 -4 -4

0.5 0.505 0.51 0.5 0.505 0.51 0.5 0.505 0.51

Time [sec] Time [sec] Time [sec]
Aerial mode-1 WTC [kA]

Aerial mode-1 WTC [kA]

Aerial mode-1 WTC [kA]

0.2 0.2 0.2

0 0 0

-0.2 -0.2 -0.2

-0.4 -0.4 -0.4

0.5 0.505 0.51 0.5 0.505 0.51 0.5 0.505 0.51
Time [sec] Time [sec] Time [sec]

Fig. 7. Wavelet based relay at a busbar interconnecting three lines

Fig.8. Phase currents, corresponding modal signals and their WTCs for an
Three sets of CTs measure the currents on each branch at internal fault between phases A and B.
X, Y, and Z. A fault in the region between the CTs, such as
fault F1, is an internal fault whereas a fault outside the CTs, Measurements at X Measurements at Y Measurements at Z
Phase currents[kA]

Phase currents[kA]

Phase currents[kA]
such as fault F2, is an external fault. 1 1
1
The measured three-phase currents are transformed to
0 0
modal domain using the constant Clark’s transformation 0

matrix in (6) before applying wavelet transform. -1 -1 -1

0.5 0.505 0.51 0.5 0.505 0.51 0.5 0.505 0.51
⎡ I1 ⎤ ⎡1 1 1 ⎤ ⎡I a ⎤ Time [sec] Time [sec] Time [sec]
Aerial mode currents[kA]

Aerial mode currents[kA]

Aerial mode currents[kA]

⎢ I ⎥ = 1 ⎢2 − 1 − 1 ⎥⎥ ⎢⎢ I b ⎥⎥
4 4 4
(6)
⎢ 2⎥ 3⎢ 2 2 2
⎢⎣ I 3 ⎥⎦ ⎢⎣0 3 − 3 ⎥⎦ ⎢⎣ I c ⎥⎦ 0 0 0
In (6), Ia, Ib, and Ic are the phase currents and I1, I2, and I3 -2 -2 -2
0.5 0.505 0.51 0.5 0.505 0.51 0.5 0.505 0.51
are the modal components. Only the components I2, and I3 , Time [sec] Time [sec] Time [sec]
Aerial mode-1 WTC [kA]

Aerial mode-1 WTC [kA]

Aerial mode-1 WTC [kA]

which are known as aerial mode components, are used for the 0.5 0.5 0.5
fault locations. By using these two components all types of
ground and ungrounded faults can be handled. Thus the use of 0 0 0

modal transformed quantities gives some computational

-0.5 -0.5 -0.5
advantage, in addition to providing decoupled signals. In the 0.5 0.505 0.51 0.5 0.505 0.51 0.5 0.505 0.51
following analysis, Level-1 wavelet transform coefficients Time [sec] Time [sec] Time [sec]

(WTCs) obtained with DB4 mother wavelet were used. Fig. 9. Phase currents, corresponding modal signals and their WTCs for an
For an internal fault such as F1, WTCs of the currents external fault between phases A and B.
measured at X, Y and Z will all have the same sign as shown
in Fig. 8.
 G1

Infinite bus 9

[EXternal ]

X
A
1 R1
B
X

X
C D

7
[External ]
Communication link

X
E F

X
2 R2
G H

X
X
10

G2

Y
X
[External ]
X

Z
R6 T
6 U

X
X
[External]
X

A’ C’ R5 5
X

B’
X

X
X
8 V
12 Fig. 12. Two faults F1 and F2 occur on a segment connected to a relay agent
[External ] G4
and their lattice diagrams
I
X

J
X

N P
X

3 O X
K
L M
R3 R4 4 In order to explain the concept, consider two ungrounded
X

X
X
X

R
[External ] Q S
faults; fault F1 close to the near end of Segment-A and fault F2
11
F close to the remote end of the same segment. Fig. 12 also
G3
Communication link shows the lattice diagram for the resulting traveling waves;
dotted lines correspond to F1 and the dark lines correspond to
Fig. 10. Wavelet based protection scheme. Arrows indicate fault current
directions identified by each relay
F2. In order to estimate the fault distance, it is necessary to
find the time between the arrival of successive traveling wave
WTCs at R3
5 fronts originating or reflected from the fault. For example, for
A erial m ode-1 W TCs

I
2.5 fault F1, t11 and t12 are the arrival times of the transient
K
0
originating from the fault and its first reflection from the fault,
J
L
respectively. Knowing the traveling time (t12 - t11), and
-2.5
M assuming the propagation speed to be that of light, the
-5
20.0644 20.0648 20.0652 20.0656 20.066 distance to the fault can be estimated. The time interval (t12 -
Time (s)
WTCs at R4 t11) can be estimated by using the WTCs of the currents as
20
shown in Fig. 13a.
A erial m ode-1 W TCs

10 N
S Fault F1 Fault F2
-3
R Measurements at X x 10 Measurements at X
0
0.01 5
-10 P
Q O
WTC

WTC

0 0
-20
20.0644 20.0648 20.0652 20.0656 20.066
Time (s)
Fig. 11. Wavelet coefficients observed by R3 and R4 -0.01 -5
0.2 0.2001 0.2002 0.2002 0.2003 0.2004 0.2005
Time [sec] Time [sec]
Fig. 11 shows the wavelet coefficients observed at bus-3 Measurements at Y -3
x 10 Measurements at Y
0.01 5
and bus-4. WTC of branch M has a sign opposite to those of
the other branches connected to bus-3. Similarly, WTC of
WTC

WTC

0 0
branch Q has a sign opposite to those of the other branches
connected to bus-4. From the above information, R3 -0.01 -5
0.2 0.2001 0.2002 0.2002 0.2003 0.2004 0.2005
determines that the fault is in the direction of bus-4, while R4 Time [sec] Time [sec]
determines that the fault is in the direction of bus-3. If Measurements at Z -3
x 10 Measurements at Z
0.01 5
communication is provided, R3 and R4 can jointly determine
that the fault is in the line M-Q connecting them. In Fig. 10,
WTC

WTC

0 0
fault type (internal/external) and fault direction as identified
by each relay are also indicated.
-0.01 -5
0.2 0.2001 0.2002 0.2002 0.2003 0.2004 0.2005
B. Fault Location using Traveling Waves Time [sec] Time [sec]
t11 t12 t21 t22 t23
Distance to a fault can be estimated using traveling waves
(a) (b)
originating from the fault [13]. Consider a fault locater Fig. 13. Transients observed at three measurement points for fault F1 and F2.
installed at a busbar connecting three line segments A, B and
C in Fig. 12.
 The situation is more complicated for a fault such as F2, VII. REFERENCES
due to the arrival of reflections from other points before the [1] A. Graps, (2003), Introduction to Wavelets, Istituto di Fisica dello
arrival of wave front reflected from the fault. For example, t21 Sapazio Interplanetario, Rome, Italy, [online]. Available:
and t23 are the relevant transient arrival times required to http://www.amara.com.
[2] C. H. Lee, Y. J. Wang, and W.L. Huang, “A Literature Survey of
estimate the distance to the fault. However, the transient Wavelets in Power Engineering Applications”, in Proc. of Science
arriving at time t22 (where t22 < t23) should be ignored in order Council, ROC (A), Vol. 24, No.4, pp 249-258, 2000.
to correctly estimate the fault distance. This can be achieved [3] D.C. Robertson, O. I. Camps, J.S. Mayer, and W.B. Gish , “Wavelets
and Electromagnetic Power System Transients”, IEEE Trans. on Power
with the help of fault direction finding method described in
Delivery, Vol. 11, No. 2, April 1996.
the previous section. As seen in Fig. 13b, for transients [4] S. Santoso, E.J. Powers, W.M. Grady, “Power Quality Disturbance Data
arriving from line Segment A, the WTC of the current Compression using Wavelet Transform Methods”, IEEE Trans. on
measured at point X has a sign opposite to the WTCs of that Power Delivery, Vol. 12, No. 3, Jul. 1997
[5] H.T. Yang and C.C. Liao, “A De-Noising Scheme for Enhancing
measured at the other two points (Y and Z). This permits
Wavelet Based Power Quality Monitoring System”, IEEE Transactions
distinguishing the transient coming from the fault from the on Power Delivery, Vol. 16, No. 3, Jul. 2003.
other transients. [6] Z. Hao, T.R. Blackburn, B.T. Phung, M.S. Naderi, “A novel wavelet de-
noising method for on-site PD measurements on HV cables”, in Proc. of
C. Power Quality Disturbance Detection 2005 International Symposium on Electrical Insulating Materials,
Wavelets can also be used to detect the power quality Kitakyushu, Japan, June 5-9, 2005,
[7] C.H. Kim, H. Kim, Y.H. Ko, S.H Byun, R.K. Aggarwal and A.T. Johns,
events such as voltage sags and swells, transients, harmonics, "A Novel Fault-Detection Technique of High-Impedance Arcing Faults
etc. Detection of such disturbances is necessary for in Transmission Lines Using the Wavelet Transform," IEEE Trans.
applications such as disturbance recorders: for example, Power Del., vol. 17, no. 4, pp. 921-929, Oct. 2002.
[8] W. Chen, O.P. Malik, X. Yin, D. Chen, and Z. Zhang, “ Study of
change in wavelet coefficients can be used to trigger
Waveltet-based Ultra High Speed Directional Transmission Line
waveform recording. Fig. 14 shows an example voltage Protection”, IEEE Transactions on Power Delivery, Vol. 18, No. 4, Oct.
waveform with a swell and sag, and the corresponding 2003.
wavelet coefficients. WTC shown in Fig. 14 is the level-1 [9] L.S. Safavian, W. Kinsner, H. Turanli, “Aquantitative comparison of
different mother wavelets for characterizing transients in power
detail coefficient obtained with DB8 mother wavelet. systems”, in Proc. of IEEE Canadian Conf. on Electrical and Computer
Engineering, pp. 1342 – 1346, May 2006
200 [10] J. Stoupis, M. Maharsi, R. Nuqui, S. Kunsman, and R. Das, “Ground
100
Alert: Reliable Detection of High Impedance Faults Caused by Downed
Voltage (kV)

Conductors”, ABB Review, No.1, 2004.

0
[11] R.P.Bingham, D. Kreiss and S. Santoso, Advances in Data Reduction
-100 Techniques for Power Quality Instrumentation, Dranetz Technologies
-200 Inc. [online]. Available: http://www.dranetz-bmi.com/pdf/
[12] N. Perera, A.D. Rajapakse, “Agent-Based Protection Scheme for
0.020
swell sag Distribution Networks with Distributed Generators”, Power Engineering
0.010 Society General Meeting, 2006, pp 1-6 IEEE 18-22, June 2006.
Detail-1 WTC

0.000 [13] C.Y. Evrenosoglu and A. Abur, "Travelling Wave Based Fault Location
for Teed Circuits," IEEE Trans. Power Del., vol..20, no. 2, pp. 1115-
-0.010
1121, Apr.2005.
-0.020
Time ... 0.200 0.250 0.300 0.350 0.400
VIII. BIOGRAPHIES
Fig. 14. Detection of power quality events using wavelets: a voltage swell and Nuwan Perera received the B.Sc. (Eng.) degree from University of
a voltage sag Moratuwa, Moratuwa, Sri Lanka, in 2003. Currently he is pursuing the MSc.
degree in the Department of Electrical and Computer Engineering at
VI. CONCLUSIONS University of Manitoba, Winnipeg, MB, Canada. His research interests are in
power system protection.
A tool for performing online discrete wavelet
transformation in an emtp-type program (PSCAD/EMTDC) Athula D. Rajapakse received the B.Sc. (Eng.) degree from the University of
Moratuwa, Moratuwa, Sri Lanka, in 1990, the M.Eng. degree from the Asian
was developed. The online DWT tool handles nine different
Institute of Technology, Bangkok, Thailand, in 1993, and the Ph.D. degree
mother wavelet types, and incorporates an integrated anti- from the University of Tokyo, Tokyo, Japan, in 1998. Currently, he is an
aliasing filter and a sampler. Accuracy of calculations was Assistant Professor at the University of Manitoba, Winnipeg, MB, Canada.
extensively validated against MATLAB® wavelet toolbox. His research interests include power system protection, transient simulation of
power and power-electronic systems, distributed and renewable energy
Several applications of DWT in power system protection and systems.
power quality monitoring were presented. The new online
DWT tool will stimulate the development of new wavelet R.P. Jayasinghe obtained his B.Sc. (Eng) degree from the University of
applications in the field of power systems. Moratuwa, Sri Lanka in 1987 and Ph.D. degree from the University of
Manitoba in 1997. He is currently with the Manitoba HVDC Research Centre.
As a Development Engineer he plays a major role in the current developments
in PSCAD/EMTDC simulation program. He is a Registered Professional
Engineer in the Province of Manitoba. Dr. Jayasinghe also serves as an
adjunct professor at the University of Manitoba.