Beruflich Dokumente
Kultur Dokumente
1. Introduction
The frequency of a power network is an important
operational parameter for the safety, stability and
efficiency of the power system. Reliable frequency
measurement is prerequisite for effective power control,
load shedding, load restoration and system protection.
Therefore, there is a need for fast and accurate estimation
of the frequency of the power network using voltage
waveforms which may be corrupted by noise and
harmonics components.
Several digital methods for the frequency measuring have
been proposed in the past few decades. The use of the
zero crossing detection and calculation of the number of
cycles that occur in a predetermined time interval [1] is a
simple and well-known methodology. The discrete
Fourier transformation, least error squares and Kalman
filter are known signal processing techniques, used for
the frequency measurement [2-6]. As shown in [7], the
bilinear form approach seems to be a very efficient
method for both small frequency deviation and offnominal frequency estimation. An adaptive algorithm for
x ( t ) = C cos( t + ) + R ( t ) .
(1)
frequency, f - fundamental harmonic frequency, phase of the fundamental harmonic, R(t) - part of the
signal which contain higher harmonics and zero mean
white noise.
T
T
T
2
2
2
cos(2 a a )
cos(3 a a ) . . .
CO = cos( a a )
m
m
m
m
m
m
m
T
T
2m
X xn cos( a a n) j xn sin( a a ) =
m n =1
m
m
n =1
= A + jB
...
(2)
(4)
SAM = [x1
x2
x3
. . . xm ]
Ta
1
), where: Ta=1/fa - assumed
=
m mf a
(3)
x(t),
A(t)
[%]
100
50
-50
-100
-150
0.02
(5)
2
A(t ) = x n cos(n) .
m
n =1
T
2
2
cos((m 1) a a )
,
m
m
m
0.04
0.06
time [s]
0.08
4) Because of
fa =
f =
m
1
.
= fa
m ATs
mA
(6)
M A = [m A1
m A2
m A3
. . . m AM ]
(9)
100
50
[%]
Ap
An
A(t)
K
Ts
-50
-100
0.0215
0.0315
time [s]
An
P
and
K + P =1
(7)
An
A p An
m As =
m A1 + m A2 + ... + m AM
,
M
and K =
Ap
A p An
(8)
(10)
m
m As
(11)
350
300
250
200
150
100
50
0
0 .1
0 .2
0 .3
0 .4
0 .5
tim e [s ]
[% ]
150
100
A (t)
50
x ( t),
e s tim a te d f re q u e n c y [H z ]
400
-5 0
-1 0 0
-1 5 0
0 .1 4
0 .1 6
(12)
0 .1 8
t im e [ s ]
Fig. 4. Test signal x(t) corrupted by noise and his cosine component A(t).
0 .2
51
50.8
50.8
50.6
50.6
51
50.4
50.2
50
49.8
49.6
50.4
50.2
50
49.8
49.6
49.4
49.4
49.2
49.2
49
0.5
1.5
2.5
3.5
4.5
49
0.5
1.5
2.5
time [s]
a)
3.5
4.5
tim e [s]
b)
Fig. 5. Estimates frequency of signal shown in Fig.4; a) non averaged estimated frequency; b) averaged estimated
frequency over M=5 periods of the signal.
C. Dynamic Test
k v
f r e
t i m
i r a
s
e
. 2
. 4
. 6
. 8
f r e
i r a n a
. 2
. 4
. 6
. 8
0
0
t i m
time
m time
e [s][ s[s]
]
a)
v r e
. 1
. 3
. 5
. 7
. 9
v r e
m
e
time
. 1
. 3
. 5
. 7
. 9
[ s ]
[s]
b)
Fig. 6. Frequency estimation; a) when the input signal frequency was step changed and b) when the input signal
magnitude was step changed.
v o lt a g e
1 0
Voltage [V]
- 5
-1 0
e s t im a t e d
fre q
a)
0 .0 2
0 .0 4
t im e
0 .0 6
[s ]
0 .0 8
a)
5 0
4 9 .9 5
b)
4 9 .9
4 9 .8 5
4 9 .8
0 .1
0 .3
t im e
0 .5
[s ]
0 .7
b)
Fig. 7. a) Steady state network voltage waves and b) Frequency estimation.
[ V
voltage [V]
- 1
- 2
- 3
- 4
f
a
n
i r a
t i m
s
e
estimated
estimatedfrequency
frequency [Hz]
[Hz]
v r e
a)
5
b)
0
e
1
[ s
time [s]
r e
m
e
[ s
time
[s]
Fig. 8. a) Transient of magnitude of generator voltage and b) Frequency estimation during the voltage
magnitude and frequency transient.
5. Conclusion
In this paper a new approach to zero crossing technique
for frequency estimation is presented. This approach
gives a very simple and robust algorithm with acceptable
accuracy. The Fourier algorithm is used as digital filter in
order to extract cosine and sine parts of the fundamental
frequency component, and the zero crossing technique is
applied on cosine and sine parts of the signal. Therefore,
the derived algorithm requires modest computer
resources for implementation. To verify presented
algorithm, computer simulated and real-life conditions
tests are provided. All tests have shown that presented
algorithn is accurate, wide-range and robust, with good
frequency tracking performance. The results of tests
confirmed that the algorithm can be a very useful tool in
the power system protection applications.
References
[1] A. Phadke, J. Thorp, and M. Adamiak, A new
measurement technique for tracking voltage phasors,
local systems frequency, and rate of change of
frequency, IEEE Trans. Power Applicat. Syst., vol.
PAS-102, pp. 10251038, May 1983.
[2] M. S. Sachdev and M. Nagpal, A recursive least
error squares algorithm for power system relaying
and measurement applications, IEEE Trans.Power
Delivery, vol. 6, pp. 10081015, July 1991.
[3] M. uri, . urii, An Algorithm for OffNominal Frequency Measurements in Electric
Power Systems, Electronics, vol. 7, No 1, pp. 1114, Sept. 2003.