Beruflich Dokumente
Kultur Dokumente
Pertemuan ke 13
FUZZY LOGIC
Lotfi A. Zadeh Fuzzy Sets, Information and Control, Vol 8, pp.338-353,1965.
Clearly, the class of all real numbers which are much greater than 1, or the class of beautiful women, or the class of tall men, do not constitute classes or sets in the usual mathematical sense of these terms (Zadeh, 1965).
PROF. ZADEH
Fuzzy theory should not be regarded as a
Fuzzy System
Complexity of System
Where few numerical data exist and where only ambiguous or imprecise information maybe available.
FUZZY SET
Unlike sharp boundary in crisp set, here boundaries seem vague Transition from member to nonmember appears gradual rather than abrupt
Universal Set X always a crisp set. Crisp set assigns value {0,1} to members in X Fuzzy set assigns value [0,1] to members in X These values are called the membership functions m. Membership function of a fuzzy set A is denoted by :
(Depok/0.8) } FUZZY
CONTOH
Grade ( m )
1
Dingin Panas
30 oC
30 oC
CRISP
FUZZY
TINGGI BADAN
m
sedang 1
CRISP
150
155
160
tinggi
m
sedang 1 0,5 0
150 155 160 tinggi
FUZZY
SEASONS
Spring Summer Autumn Winter
1 Membership
0.5
AROUND 4
1
Membership
0.5
0 0
4 Measurements
AGE
1
young very young old
Membership
0.5
0 0 20 40 60 80 100
Age
MEMBERSHIP FUNCTION
1 0.5 0 (a) (d) (g) (j)
1
0.5 0 (b) 1 0.5 0 -100 (e) (h) (k)
0 (c)
100 -100
0 (f)
100 -100
0 (i)
100 -100
0 (l)
100
SET OPERATION
1
A
Membership 0.5
SET OPERATION
AB
AB
LINGUISTIC VARIABLES
Linguistic variable is a variable whose values are words or sentences in a natural or artificial language. Each linguistic variable may be assigned one or more linguistic values, which are in turn connected to a numeric value through the mechanism of membership functions.
LINGUISTIC VARIABLES
Fuzzy linguistic terms often consist of two parts:
1) Fuzzy predicate : expensive, old, rare, dangerous, good, etc. 2) Fuzzy modifier: very, likely, almost impossible, extremely unlikely, etc. The modifier is used to change the meaning of predicate and it can be grouped into the following two classes: a) Fuzzy truth qualifier or fuzzy truth value: quite true, very true, more or less true, mostly false, etc. b) Fuzzy quantifier: many, few, almost, all, usually, etc.
FUZZY PREDICATE
Fuzzy predicate If the set defining the predicates of individual is a fuzzy set, the predicate is called a fuzzy predicate Example z is expensive.
w is young.
The terms expensive and young are fuzzy terms. Therefore the sets expensive(z) and young(w) are fuzzy sets
FUZZY PREDICATE
When a fuzzy predicate x is P is given, we can interpret it in two ways : P(x) is a fuzzy set. The membership degree of x in the set P is defined by the membership function mP(x) mP(x) is the satisfactory degree of x for the property P. Therefore, the truth value of the fuzzy predicate is defined by the membership function : Truth value = mP(x)
FUZZY VARIABLES
Variables whose states are defined by linguistic concepts like low, medium, high. These linguistic concepts are fuzzy sets themselves. Very Low High high
Temperature
FUZZY VARIABLES
Usefulness of fuzzy sets depends on our capability to construct appropriate membership functions for various given concepts in various contexts. Constructing meaningful membership functions is a difficult problem GAs have been employed for this purpose.
EXAMPLE
if speed is interpreted as a linguistic variable, then its term set T (speed) could be T = { slow, moderate, fast, very slow, more or less fast, sligthly slow, ..}.
where each term in T (speed) is characterized by a fuzzy set in a universe of discourse U = [0; 100]. We might interpret slow as a speed below about 40 km/h" moderate as a speed close to 55 km/h" fast as a speed above about 70 km/h"
SPEED
MEMBERSHIP FUNCTION
m
Positive Large
0
Negative Medium
m
NL NM NS ZERO PS PM PL
Fuzzification
Scales and maps input variables to fuzzy sets Approximate reasoning Deduces the control action Convert fuzzy output values to control signals
Inference Mechanism
Defuzzification
FUZZY CONTROLLER
Fuzzy Controller
Rule Base Pre processing Fuzzification Inference engine Defuzzification Post Processing
FUZZY CONTROLLER
action Defuzzification Module Fuzzy Controller
Controlled Process
condition
Fuzzification Module
FUZZIFICATION
Transformation from crisp input to fuzzy input.
1 0.6
input
APPROXIMATE REASONING
y = f(x) Then we can make inferences easily
premise y = f(x)
fact x = x -----------------------------consequence y = f( x )
BASIC INFERENCES
x is A AB -----------x is B x is A x is B --------------x is A B and Mary is very young very young young ------------------------Mary is young pressure is not very high pressure is not very low -----------------------------------pressure is not very high
y is B
INFERENCE
Penarikan kesimpulan dari semua fungsi keanggotaan
yang sudah didefinisikan menggunakan RULE BASE. IF A AND B THEN C RULE 1 : A1 , B1 C1 RULE 2 : A2 , B2 C2 FAKTA : x , Dx ---------------------------------------KESIMPULAN : C
MIN-MAX METHOD
NL A1 B1 PM C1 PL
C1'
NM A2 B2
PS C2
PM
C2'
Dx
DEFUZZIFICATION
Transformation from fuzzy output to crisp ouput.
Defuzzification is a process to get a non-fuzzy value that best represents the possibility distribution of an inferred fuzzy control action. There is no systematic procedure for choosing a good defuzzification strategy. Selection of defuzzification procedure depends on the properties of the application.
EXAMPLE
MEAN OF MAXIMUM
EXAMPLE
HEIGHT METHOD
WEIGHTED AVERAGE
EXAMPLE
FIRST/LAST MAXIMA
EXAMPLE
FUZZY SYSTEM
FUZZY SYSTEM
FUZZY SYSTEM
EXAMPLE
LANGKAH KEDUA
AC CONTROLLER
FUZZY RULES
AC CONTROLER
Air Temperature
Fan Speed
Set stop {30, 0, 0} cold {50, 0, 0} Set slow {50, 30, 10} cool {65, 55, 45} just right {70, 65, 60} Set medium {60, 50, 40} Set fast {90, 70, 50} warm {85, 75, 65} Set blast {, 100, 80} hot {, 90, 80}
LANGKAH KETIGA
RELASI FUZZY
AKTIF RULE
SUHU 65 0F
SUHU 63 0F
TAMBAHKAN OUTPUT
DEFUZZIFIKASI OUTPUT
INVERTED PENDULUM
Goal: maintain stick vertical ! q is measured at time k; Dq is change of q at (k)
q(k), Dq(k)
Action : u(k)
NB
PB
FUZZY RULES
q(k) Dq (k) NB N Z P PB NB Z N Z P Z PB
NB
PB
Here, 7 fuzzy relations are enough to get a working minmax-cog fuzzy controller !