FUZZY LOGIC

CONTROL

Sistem Kendali
Kendali = pengukuran,
penggerak
(aktuator) dan
prosesor
Video Pendulum Terbal
ik

Produk yg mendasarkan pada embedded

control

Embedded control di sekitar kita

Inverted Pendulum

Siklus Desain Sistem

Kendali
Uji
eksperi
men pd
sistem
fisik

Pemilih
an
sensor
&
aktuat
or

Pemodel
an
dinamik
&
simulasi
openloop

Hardwarein-the-loop
simulation
& rapid
prototypin
g

Desain
pengen
-dali
Implem
en
tasi
realtime
pengendali

Simula
si
closedloop

CONVENTIONAL CONTROL
Open-loop control is blind to actual
output

Closed-loop control takes account of actual output

and compares this to desired output

Desired
Output

Input

Controller/
Amplifier

Process
Dynamics

Measurement

Output

Digital Control System

Configuration

Fuzzy Control
Fuzzy control provides a formal methodology for
representing, manipulating, and implementing a
humans heuristic knowledge about how to control
a system.

10

Fuzzy Systems

Input

Fuzzifier

Inference
Engine

Defuzzifier

Output

Fuzzy
Knowledge base

11

Fuzzy Control Systems

Input

Fuzzifier

Inference
Engine

Plant

Defuzzifier

Fuzzy
Knowledge base

12

Output

Basic configuration of fuzzy logic control system

KB

Desired
position

E
-

FI

DML

SENSOR

KB: Knowledge Base,

FI: Fuzzification interface,
DML: Decision Making Logic,
DFI: Defuzzification interface

DFI

PLANT

Output

Block diagram for a fuzzy inference system

Rule 1

x is A1

w1

y is B1

(Fuzzy)

Rule 2

(Crisp or
Fuzzy)

x is A2

w2

y is B2

(Fuzzy)
Aggregrator

Rule r

x is Ar

wr

y is Br

(Fuzzy)

(Fuzzy)

Defuzzifier

(Crisp)

Block diagram for a fuzzy inference system

Rule 1

x is A1

w1

y is B1

(Fuzzy)

Rule 2
(Crisp or
Fuzzy)

x is A2

w2

y is B2

(Fuzzy)
Aggregato r

Rule r

x is Ar

w2

y is Br

(Fuzzy)

(Fuzzy)

Defuzzif ie r

(Crisp)

The structure of a FKBC

Valu es

Valu es

FM

DM

Norm alizatio n
FM-F1

Denorm alizatio n
DM-F1
optio nal

Fuzzif ic ation
FM-F2

Inference
Engine
Rulebase

oblig atory
Defuzzif ic ation
DM-F2

m eaning representation

Data base

Com puta tio nal flow

Inform ation flo w
Sym bolic to Meanin g tr ansla tio n

Rule base
sym bolic representation

Three fuzzy partition terms:

N(Negative), Z(Zero), and P(Positive)

A (E )
N

E
-1

-.5

-.4

.4

.5

B (DE )
N

DE
-1

-.4

-.2

.2

.4

C (CI )
N

(CI)
-5

-4

-2

Rule Base Construction

Intuitive Method
Method based Operator (Expert)
Experience
Response Ideal Method

Fuzzy decision rule justification by using step response

DML Rule Prototype with Three Fuzzy Set Term

(negative, zero, positive)

Rule Justification with Three Fuzzy Set Terms

Configuration Procedure and DML Consequence

E
3

DE
-1

-4

-3

3 4

CI
5

-3
Konsekuen dari DML
6
.25
CI

Masukan kendali setelah tahap defuzzifikasi

(Masukan kendali C1: Metoda maksimum rata-rata
Masukan kendali C2: Metoda pusat area)

Konsekuen dari DML

6
.25
CI
C2

C1

Various defuzzification schemes for obtaining a crisp output.

Z
Smallest of Max.
Largest of Max.

Centroid of Area
Bisecter of Area
Mean of Max.

Fuzzy Partition using Seven Terms

NB

-1

NM

NS

ZE PS

PM

PB

+1

Fuzzy Rules with 7 Terms

{NB, NM, NS, ZE, PS, PM, PB}

Tabel Justifikasi kaidah dengan tujuh subset

fuzzy
Nomor Kaidah

DE

CI

Daerah Acuan

14
15
16
17
18
19

PB
PS
NB
NS
PS
NS

NS
NB
PS
PB
NS
PS

PM
NM
NM
PM
ZE
ZE

i(rise-time)
ii(overshoot)
iii
viii
ix
xi

The rule base of a PI-like FKBC in tabular form:

the five groups of rule

e
NB

NM

NS

ZO

PS

PM

PB

NB

NB

NB

NB

NB

NM

NS

ZO

NM

NB

NB

NB

NM

NS

ZO

PS

NS

NB

NB

NM

NS

ZO

PS

PM

ZO

NB

NM

NS

ZO

PS

PM

PB

PS

NM

NS

ZO

PS

PM

PB

PB

PM

NS

ZO

PS

PM

PB

PB

PB

PB

ZO

PS

PM

PB

PB

PB

PB

Group 0
Group 1
Group 2
Group 3
Group 4

The Mamdani fuzzy inference system using min and max for T-norm
And T-conorm operators, respectively.

A1

B1

C1

R1

C '1
X

A2

Y
B2

Z
C2

R2

C '2
X

Y
y

Max

C'
Z

IF X is A1 and Y is B1 THEN Z is C1 : R1
IF X is A2 and Y is B2 THEN Z is C2 : R2

Z COA