Characteristics of Measurement of Criterion Values and Adaptive - Digital Signal Processing in Power System Protection and Control - page 194

Fig. 9.15 Sine/cosine filters'
frequency responses with
marked frequencies and their
harmonics when fundamental
frequency changes

h ,

C h S

1

- harm. h

1

- har m. h'

0.8

2

0.6

0.4

2h 3h

4h' 6h' 7h'

0.2

1h

1h ' 2h ' 3h ' 5h '

0

0 25 50 75 100 125 150 175 200

f [Hz]

u(n)

f

Coarse frequency
measurement

N

U, I
P, Q
R, X, Z

ORTHOGONAL
FILTERS

MEASUREMENT
ALGORITHMS

i(n)

Fig. 9.16 Block scheme of the frequency-adaptive measurement procedure

One can avoid the above effect using adaptive solutions, where the filters are
being matched to actual frequency of the signals, as well as applying correction of
coefficient C according to actual, measured frequency [ 3 , 4 , 6 , 10 ]. The solutions
can be outlined with the block scheme depicted in Fig. 9.16 . Important part of the
scheme is a coarse frequency measurement, which need not be very accurate if
applied algorithms of criterion values measurement are insensitive to small fre-
quency deviations.
Modification of the filters is easy. One period FIR sine or cosine filter with the
window N 0 has its basic angular frequency equal to 2p = N 0 : To get adaptive
solution one should use filter window as close to the period of fundamental
components of signals as possible. To obtain that one must measure actual fre-
quency of signals and control whether the frequency deviation is greater than
assumed.