Digital Signal Processing Reference
In-Depth Information
Chapter 11
Elements of Fuzzy Logic in Protective
Relays
As mentioned, e.g. in the preceding chapter on decision making, the protective
relay equipped with standard, deterministic signal processing algorithms (mea-
surement with fixed data window, discrimination with fixed thresholds or char-
acteristics) may not perform properly for all conceivable power system operating
conditions. The problems may arise mainly due to:
• transient inaccuracy of measurement with commonly used deterministic algo-
rithms (here especially the measurement overshoot is dangerous for relay
selectivity),
• difficulties in setting of the relay (especially when the regions corresponding to
the classes to be distinguished overlap each other).
In this chapter the main features of fuzzy logic (FL) based systems are pre-
sented that are hoped to improve protection operation for unclear cases, situations
when some data are missing or corrupted or when the measurement unit may
deliver criteria signal estimates obtained with low level of confidence. Below the
fundamentals of the fuzzy sets and logic are outlined, which is followed by the
application hints and selected examples of this technique for power system pro-
tection purposes.
11.1 Fuzzy Sets and Fuzzy Numbers
Most of the concepts describing the problems of real life are not very precise. It is
difficult to assess many things quantitatively with high degree of precision,
especially when the situation is dynamic (well known Heisenberg's uncertainty
principle related to measurement) or influenced by a number of factors, some of
which are weakly defined. To describe certain concepts, e.g. people's height or
distance to destination, the linguistic statements like ''high'', ''short'', ''far away'',
etc. are usually used. The reasons for this are, among others, that
Search WWH ::
Custom Search