Table of Contents
Advertisement
PowerSoft
The median estimates the typical consumptions with a higher accuracy; this
parameter allows achieving accurate estimation even if anomalous points
are still present, but the number of the analysed samples is to be high
enough.
Variance and standard deviation estimate how the values are distributed
around their mean or median. The higher the variance, the wider the
distribution of the values around the mean/median.
These parameters are very important in determining the optimal
contractual power: the suggested powers are in fact defined as the median
plus the standard deviation (or 2 or 3 times the standard deviation). If the
analysis has been correctly performed (as suggested before) and without
any unpredictable events, setting the Confidence level as Low, the probability
that the actual requested power is lower than the suggested one is 68.3%; if
it is set as Middle the probability is 95.4%, while if it is High the probability
is 99.7%.
61
PowerSoft – Instruction manual
Advertisement
Table of Contents
