CARLO GAVAZZI POWERSOFT, Инструкция по эксплуатации

Страница: 60 / 74

PowerSoft
PowerSoft – Instruction manual
60
Determining the optimal power profile, it is important to understand that
this analysis is meaningless if the consumptions are not constant in the
selected period. In fact, it is useless to find a “typical” trend if it is known
that the consumptions are changing. The procedure described in the
previous paragraph can help in understanding if this analysis can be done,
avoiding any incorrect estimation.
One (or more) day(s) of the week can substantially differ from the other
days; in this case the analysis is to be repeated for each group of
“homogeneous” days. It is possible to tick/untick the days in the
Excluded
day section and carry out the analysis; then the operation is to be repeated
for the other days. Consider that a trend which is regularly higher or lower
than the “standard” trend allows to an over-estimated proposed contractual
power.
Even if the previous factors are considered, other anomalous data can still
be present. They can be due to different reasons: holidays during the
working days of the selected period; production stops; strikes; damages or
anomalies in the energy supply.
The holidays can be excluded from the calculation selecting the relevant
days in the
Holiday section. In the Excluded days box the desired days can be
added to the holiday list. To remove a wrong day, select it and press the
Delete button.
To delete the other anomalous points (due to occasional and short
damages, production stops, etc.), an automatic filter is available. Press the
2 - erase anomalous points button to execute the filtering. A report (erased
points/total points) is available in the bottom status bar.
Now the system is able to provide the optimised contractual powers. The
graph displays the arithmetical mean, the median and the variance (that is
the square of the standard deviation) of the demanded power.
The arithmetical mean of the power is not a reliable value to estimate the
typical consumptions, because any anomalous point can lead to an
inaccurate estimation.