introduction to softcomputing

Introduction to Softcomputing

Son KuswadiRobotic and Automation Based on Biologically-inspired Technology (RABBIT)Electronic Engineering Polytechnic Institute of SurabayaInstitut Teknologi Sepuluh Nopember

Agenda

AI and Softcomputing From Conventional AI to

Computational Intelligence Neural Networks Fuzzy Set Theory Evolutionary Computation

AI and Softcomputing

AI: predicate logic and symbol manipulation techniques

Use

r In

terf

ace

Inference Engine

Explanation Facility

Knowledge Acquisition

KB: •Fact•rules

GlobalDatabase

KnowledgeEngineer

HumanExpert

Question

Response

Expert Systems

User

AI and Softcomputing

ANNLearning and

adaptation

Fuzzy Set TheoryKnowledge representation

ViaFuzzy if-then RULE

Genetic AlgorithmsSystematic

Random Search

AI and Softcomputing

ANNLearning and

adaptation

Fuzzy Set TheoryKnowledge representation

ViaFuzzy if-then RULE

Genetic AlgorithmsSystematic

Random Search

AISymbolic

Manipulation

AI and Softcomputing

cat

cut

knowledge

Animal? cat

Neural characterrecognition

From Conventional AI to Computational Intelligence Conventional AI:

Focuses on attempt to mimic human intelligent behavior by expressing it in language forms or symbolic rules

Manipulates symbols on the assumption that such behavior can be stored in symbolically structured knowledge bases (physical symbol system hypothesis)

From Conventional AI to Computational Intelligence

Intelligent Systems

Sensing Devices(Vision)

Natural Language Processor

MechanicalDevices

Perceptions

Actions

TaskGenerator

KnowledgeHandler

DataHandler Knowledge

Base

MachineLearning

Inferencing(Reasoning)

Planning

Neural Networks

Neural Networks

f

z-1

z-1

0

1

N

z-1

z-1

1

+

-

e(k+1)

0

^

^

yp(k+1)

yp(k+1)^

u(k)

Parameter Identification - Parallel

Neural Networks

ff

z-1

z-1

0

1

NN

z-1

z-1

1

+

-

e(k+1)

0

^

^

yp(k+1)

yp(k+1)^

u(k)

Parameter Identification – Series Parallel

Neural Networks

Control

ANN

Gp(s)Gc(s)R(s) C(s)+

-

+

+

Plant

Feedforward controller

Feedback controller

ANN-

+

Learning Error

Neural Networks

Control

Ball-position sensor

ControllerCurrent-driven magnetic field

Iron ball

Neural Networks

Neural Networks

Experimental Results

Feedback controlonly

Feedback with fixed gain feedforward control

Feedback with ANNFeedforward controller

Fuzzy Sets Theory

What is fuzzy thinking Experts rely on common sense when they solve the

problems How can we represent expert knowledge that uses

vague and ambiguous terms in a computer Fuzzy logic is not logic that is fuzzy but logic that is

used to describe the fuzziness. Fuzzy logic is the theory of fuzzy sets, set that calibrate the vagueness.

Fuzzy logic is based on the idea that all things admit of degrees. Temperature, height, speed, distance, beauty – all come on a sliding scale.

Jim is tall guy It is really very hot today

Fuzzy Set Theory

Communication of “fuzzy “ idea

This box is too heavy.. Therefore, we

need a lighter one…

Fuzzy Sets Theory

Boolean logic Uses sharp distinctions. It forces us to draw

a line between a members of class and non members.

Fuzzy logic Reflects how people think. It attempt to

model our senses of words, our decision making and our common sense -> more human and intelligent systems

Fuzzy Sets Theory

Prof. Lotfi Zadeh

Fuzzy Sets Theory

Classical Set vs Fuzzy set

No NameHeight

(cm)

Degree of Membership of “tall men”

Crisp Fuzzy

1 Boy 206 1 1

2 Martin 190 1 1

3 Dewanto 175 0 0.8

4 Joko 160 0 0.7

5 Kom 155 0 0.4

Fuzzy Sets Theory

Classical Set vs Fuzzy set

1

0175 Height(cm)

1

0175 Height(cm)

Universe of discourse

Membership value Membership value

Fuzzy Sets Theory

Classical Set vs Fuzzy set

Ax

AxxfXxf AA if,0

if,1)(where},1,0{:)(

Let X be the universe of discourse and its elements be denoted as x.In the classical set theory, crisp set A of X is defined as function fA(x) called the the characteristic function of A

In the fuzzy theory, fuzzy set A of universe of discourse X is defined by function called the membership function of set A)(xA

.inpartlyisif1)(0

;innotisif0)(

;intotallyisif1)(],1,0[:)(

Axx

Axx

AxxwhereXx

A

A

AA

Fuzzy Sets Theory

Membership function

0 2 4 6 8 10

0.2

0.4

0.6

0.8

1

Derajat keanggotaan [0, 1]

elemen semesta pembicaraan

A

0

1

0.5

c-b c-b/2 c c+b/2 c+b

b

Fuzzy Sets Theory

Fuzzy Expert Systems

Kecepatan (KM)

Jarak (JM)

Posisi Pedal Rem (PPR)

Fuzzy Sets Theory

Membership function

Kecepatan (km/jam)

0 20 40 60 80

Sangat Lambat

Lambat

Cukup

Cepat

Cepat Sekali

Jarak (m)

0 1 2 3 4

Sangat Dekat

AgakDekat

Sedang

Agak Jauh

Jauh Sekali

Posisi pedal rem (0)

0 10 20 30 40

Injak Penuh

Injak Agak Penuh

InjakSedang

Injak Sedikit

Injak Sedikit Sekali

KM JM PPR

Fuzzy Sets Theory

Fuzzy Rules

Aturan 1: Bila kecepatan mobil cepat sekali dan jaraknya sangat dekat maka pedal rem diinjak penuhAturan 2:Bila kecepatan mobil cukup dan jaraknya agak dekat maka pedal rem diinjak sedangAturan 3:Bila kecepatan mobil cukup dan jaraknya sangat dekat maka pedal rem diinjak agak penuh

Fuzzy Sets Theory

Fuzzy Expert Systems

Aturan 1:

Kecepatan (km/jam)

0 20 40 60 80

Cepat Sekali

Posisi pedal rem (0)

0 10 20 30 40

Injak Penuh

Jarak (m)

0 1 2 3 4

Sangat Dekat

Fuzzy Sets Theory

Fuzzy Expert Systems

Jarak (m)

0 1 2 3 4

Agak Dekat

Posisi pedal rem (0)

0 10 20 30 40

Injak Sedang

Aturan 2:

Kecepatan (km/jam)

0 20 40 60 80

Cukup

Fuzzy Sets Theory

Fuzzy Expert Systems

Jarak (m)

0 1 2 3 4

Sangat Dekat

0 10 20 30 40

Posisi pedal rem (0)

Injak Agak Penuh

Kecepatan (km/jam)

0 20 40 60 80

Cukup

Aturan 3:

Fuzzy Sets Theory

Fuzzy Expert Systems

MOM : PPR = 200

10x0,2+20x0,4COA : PPR =

0,2+0,4 = 16,670

Posisi pedal rem (0)

0 10 20 30 40

MOMCOA