Page 312 - Electrical Engineering Dictionary

linguistic models. A typical example is con- Seealsoconvexfuzzyset, fuzzysingleton.
trol of dynamic systems based on dynamic
fuzzy models, which consist of a number
of local linear models smoothly connected fuzzy observer a device to estimate the
through a set of nonlinear membership func- states of a dynamic system, involving fuzzy
tions. logic concepts. A typical example is an ob-
server constructed from a number of local
fuzzy modeling combination of available observers through fuzzy membership func-
mathematical description of the system dy- tions.
namics with its linguistic description in terms
of IF-THEN rules. In the early stages of fuzzy operator logical operator used on
fuzzy logic control, fuzzy modeling meant fuzzy sets for fuzzy reasoning. Examples are
just a linguistic description in terms of IF- the complement (NOT), union (OR), and the
THEN rules of the dynamics of the plant intersection (AND).
and the control objective. Typical exam-
ples of fuzzy models in control application
fuzzy optimal control optimal control in-
includes Mamdani model, Takagi–Sugeno–
volving fuzzy logic concepts or fuzzy control
Kang model, and fuzzy dynamic model.
to achieve an optimal control performance in-
dex.
fuzzyneuralcontrol acontrolsystemthat
incorporates fuzzy logic and fuzzy inference
fuzzy optimal filter optimal filter involv-
rules together with artificial neural networks.
ing fuzzy logic concepts or fuzzy filter to
achieve an optimal filter performance index.
fuzzy neural network artificial neural
network for processing fuzzy quantities or
fuzzy OR See triangular co-norm.
variables with some or all of the following
features: inputs are fuzzy quantities; outputs
are fuzzy quantities; weights are fuzzy quan- fuzzy output feedback fuzzy control
tities; or the neurons perform their functions based on feedback of a plant output. This
using fuzzy arithmetic. is closely related to fuzzy dynamic models.

fuzzy neuron a McCulloch–Pitts neuron fuzzy parameter estimation a method
with excitatory and inhibitory inputs repre- that uses fuzzy interpolation and fuzzy ex-
sented as degrees between 0 and 1; output is trapolationtoestimatefuzzygradesinafuzzy
a degree to which it is fired. search domain based on a few cluster center-
grade pairs. An application of this method is
fuzzy nonlinear control nonlinear con- to estimate mining deposits.
trol involving fuzzy logic concepts or fuzzy
control with application to nonlinear sys- fuzzy partition partition of a plant oper-
tems. ating space based on fuzzy logic concepts. A
typical example is a partition of a state space
fuzzy number a convex fuzzy set of the by overlapping subspaces which are charac-
real line such that terized by a set of fuzzy membership func-
1. it exists exactly one point of the real line tions.
with membership 1 to the fuzzy set;
2. its membership function is piecewise fuzzy pattern matching a pattern match-
continuous. ing technique that applies fuzzy logic to deal
In fuzzy set theory, crisp (nonfuzzy) num- with ambiguous or fuzzy features of noisy
bers are modeled as fuzzy singletons. point or line patterns.

c
2000 by CRC Press LLC