Anda di halaman 1dari 35

FUNDAMENTALS

OF

Fuzzy Logic:
XI. Fuzzy Inference
Linguistic Variables

 •Numerical variables:
 –e.g. temperature=30°C, speed=60mph
 –They normally have crispand exactvalues
 •In our daily life, we tend to use non-exact terms (variables)
 –e.g. When the room is hot, turn the air conditioning unit
high
 –What is the exact value for “hot”? 25-30°C?
 –How about “very hot”? 30-40°C?
Linguistic Variables (cont’)

 •Linguistic variablesare introduced to provide a


formal framework to capture these non-exact
concepts
 •If a variable can take words in natural languages
as its values, it is called a linguistic variable,
where the words are characterized by fuzzy sets
defined in the universe of discourse
Linguistic Variables (cont’)
 A linguistic variable can be more formally characterized by a
quintuple (x, T(x), U, G, M), where:
 –x is the name of variable, e.g. the speed of a Honda car
 –T(x) is the set of linguistic values that xcan take, e.g. T(x)
={slow, medium, fast}
 –U is the actual physical domain in which the linguistic variable
xtakes its quantitative (crisp) values; e.g. U=[0,100mph]
 –G is syntactic rule for generating the names of the values of x
 –M is semantic rule for associating each value of xwith its
meaning
Linguistic Variables (cont’)

•Linguistic variables allow the vague descriptions in natural


languages to be formulated in precise mathematical terms

•It is the first step to incorporate human knowledge into


engineering systems in a systematic and efficient manner
Linguistic Hedges

 In our daily life, we often use words to


enhance the basic meaning, e.gveryfast,
more or lesssmall In order to capture these
words, linguistic hedgeis introduced.
 The linguistic hedgeis an operator for
modifying the meaning of its operator.
 ‘very’, ‘very very’, ‘more or less’and
‘roughly’are some of the frequently used
linguistic hedges.
Linguistic Hedges (cont’)
 •‘very’–let Abe a fuzzy set in U, the ‘very’Ais
defined as a fuzzy set in U with the membership
function:

 •Similarly, the membership function of ‘very


very’Acan be defined as
Linguistic Hedges (cont’)

 •The membership function of ‘more or less’Acan be


defined as:

 The membership function of ‘roughly’Acan be defined as:


Linguistic Hedges (cont’)
Linguistic Hedges (cont’)
Fuzzy If-Then Rules
•Qualitative knowledge can be expressed in
terms of fuzzy IF-THEN rules•A fuzzy IF-THEN
rule is a conditional statement expressed as:

•IF <antecedent> THEN <consequence>•Both


antecedent and consequence are fuzzy
propositions

•Example:–If temperature is low then heating


is high
Fuzzy Propositions
 •Atomic fuzzy proposition
 –A single statement with usually a single linguistic variable
anda single linguistic value
 –e.g. temperature is low,
 –temperature is high,
 –heating is high,
 –heating is low…etc
 •Compound fuzzy proposition
 –It is a composition of atomic fuzzy propositions using the
connectives “and”, “or”, and “not”which represent fuzzy
intersection, fuzzy union, and fuzzy complement respectively
 –e.g. Temperature is low or wind is strong,
 –electricity is expensive and gas is cheap,
 –car is not fast and acceleration is low
Fuzzy Propositions (cont’)

 “and”connective –fuzzy intersections

 –Let xand ybe linguistic variables in U and V,


and Aand Bbe fuzzy sets in U and V
respectively, then the compound fuzzy
proposition:

 xis A and y is B
Fuzzy Propositions (cont’)
or”connective –fuzzy unions e.g. x is A or y
•“

is B

function•“not”connective –fuzzy
complements e.g. x is not A
Interpretations of Fuzzy If-Then rules
Interpretations of Fuzzy If-Then rules (cont’)
Interpretations of Fuzzy If-Then rules (cont’)
Fuzzy Rule base
Fuzzy Systems
Fuzzy Systems (con
Fuzzy Systems (cont’)
Fuzzy Inference
Fuzzy Inference (cont’)
Fuzzy Inference Engine
Fuzzy Inference Engine (cont’)
Fuzzy Inference Engine (cont’)
Fuzzy Inference Engine (cont’)
Fuzzy Inference Engine (cont’)
Fuzzifiers
Fuzzifiers(cont’)
Fuzzifiers(cont’)
Defuzzifiers
Defuzzifiers(cont’)
Defuzzifiers(cont’)
Defuzzifiers(cont’)

Anda mungkin juga menyukai