Power Generation Operation y Control Allen Wood 511

Download as pdf or txt
Download as pdf or txt
You are on page 1of 1

AN INTRODUCTION TO ADVANCED TOPICS IN STATE ESTIMATION

Bus 1

Bus 2

497

Bus 3

FIG. 12.18 Unobservable system showing importance of location of pseudomeasurements.

estimated values from the estimator (assuming that it is run periodically) as


pseudo-measurements. Therefore, if needed, we can provide the state estimator
with a reasonable value to use as a pseudo-measurement at any bus in the system.
The three-bus sample system in Figure 12.18 requires one pseudomeasurement. Measurement MI allows us to estimate the voltage magnitude
and phase angle at bus 2 (bus 1s voltage magnitude is measured and its phase
angle is assumed to be zero). But without knowing the generation output at
the generator unit on bus 2 or the load on bus 3, we cannot tell what voltage
magnitude and phase angle to place on bus 3; hence, the network is unobservable. We can make this three-bus system observable by adding a
pseudo-measurement of the net bus injected MW and MVAR at bus 2 or
bus 3, but not at bus 1. That is, a pseudo-measurement at bus 1 will do no
good at all because it tells nothing about the relationship of the phase angles
between bus 2 and bus 3.
When adding a pseudo-measurement to a network, we simply write the
equation for the pseudo-measurement injected power as a function of bus
voltage magnitudes and phase angles as if it were actually measured. However,
we do not wish to have the estimator treat the pseudo-measurement the same
as a legitimate measurement, since it is often quite inaccurate and is little better
than a guess. To circumvent this difficulty, we assign a large standard deviation
to this measurement. The large standard deviation allows the estimator
algorithm to treat the pseudo-measurement as if it were a measurement from
a very poor-quality metering device.
To demonstrate the use of pseudo-measurements on our six-bus test system,
all measurements were removed from buses 2, 3, 4, 5, and 6 so that bus 1 had
all remaining measurements. This rendered the network unobservable and
required adding pseudo-measurements at buses 2, 3, and 6. In the case, the
pseudo-measurements were just taken from our base-case power flow. The
results are shown in Table 12.8. Notice that the resulting estimates are quite
close to the measured values for bus 1 but that the remaining buses have large
measurement residuals. The net injections at buses 2, 3, and 6 do not closely
match the pseudo-measurements since the pseudo-measurements were weighted
much less than the legitimate measurements.
BLOG FIEE

http://fiee.zoomblog.com

You might also like