Sie sind auf Seite 1von 168

F uzzy Scheduling

W olfgang Slany

Christian Doppler Laboratory for Expert Systems

E TU Wien A Vienna Austria Europe

Phone Fax

URL http www dbai tuwien ac at sta slany html

E Mail wsi vexpert dbai tuwien ac at

CD Technical Report

Expertensysteme Labor
Expertensysteme
Labor

Christian Doppler Laboratory

for Expert Systems

Technische Universit at Wien

Institut fur Informationssysteme

Abteilung fur Datenbanken und Expertensysteme

Technische Universit at Wien Institut fur Informationssysteme Abteilung fur Datenbanken und Expertensysteme

D ISSERTATION

ausgef uhrt zum Zwecke der Erlangung des akademischen Grades

eines Doktors der technischen Wissenschaften

eingereicht an der Technischen Universit at Wien

Technisch Naturwissenschaftliche Fakult at

von

Dipl Ing Wolfgang Slany

Mariannengasse A Wien

Matrikelnummer

geboren am November in Wien

Wien im Juni

T o my wife Kyoko

and to my family

for love and support

Deutsc he Kurzfassung der Dissertation

F uzzy Logic hat sich bereits in vielen praktischen Anwendungen weltweit bewahrt

Typische Anwendungsgebiete in der Industrie sind die Uberwachung und Steuerung

von einfachen technischen Prozessen oder die Vorhersage von schwer erfa baren tech

nischen Gr o en Auch im Bereich der Planung verspricht der Einsatz von Fuzzy Logic

Verbesserungen Um die m oglichen Synergien mit dem Bereich Wissensbasiertes

Scheduling zu untersuchen wurden in dieser Arbeit Methoden der Fuzzy Logic mit

jenen aus dem Bereich zeitlicher Planung im Produktionsbereich kombiniert Die

Erwartung war da Fuzzy Logic einerseits dabei helfen kann unscharf formuliertes

Expertenwissen einfach zu modellieren andererseits dabei ungenaue Daten so zu

verarbeiten da die vorhandene Ungenauigkeit besser ausgenutzt werden kann Das

Ziel war die Qualit at der erzeugten Pl ane zu erhohen bei den Produktionskosten zu

sparen den Durchsatz zu verbessern die Auslastung der Maschinen Arbeitskr afte

und sonstigen Ressourcen zu optimieren und schlie lich in Notf allen schnell eine

gute Alternative zum Originalplan parat zu haben

Konkret wurden allgemeine Werkzeuge zur Modellierung und Verarbeitung von

unscharfen Regeln Einschr ankungen und unsicheren Daten Me werten Zeiten

erstellt Als beispielhaftes Anwendungsgebiet dient der Stahlerzeugungsproze

Die von mir entworfenen Programme FLIP Fuzzy Logic Inference Proces

sor in C ConFLIP Fuzzy Constraints aufbauend auf FLIP und Dy

naFLIP Dynamische Constraints Generierung aufbauend auf ConFLIP zur

direkten Verwendung im Planungsprogramm D ejaV u geeignet erlauben es diese

Fuzzy Einschr ankungen mit graphischer Unterstutzung InterFLIP zu erstellen

zu ver andern zu verarbeiten und zur Erstellung von Pl anen zu benutzen In der

vorliegenden Arbeit wird auch ein von mir entwickeltes Verfahren erl autert mit

dessen Hilfe die beschreibenden Parameter zuverl assig optimiert werden k onnen

Mit Hilfe dieses Verfahrens kann auf leicht verst andlich Weise uberpr uft werden ob

die Wissensbasis aller Fuzzy Einschr ankungen eine vernunftige Entscheidungsbasis

bezuglic h einer Menge von fruheren Referenzentscheidungen darstellt Die Unscharfe

Einschr ankungen erlauben es auf pr azise Art und Weise anzugeben bis zu welchem

Wert Einschr ankungen verletzt werden durfen und gleichzeitig exakt festzuhalten

welche Werte als wie wunsc henswert einzustufen sein sollen Beispielsweise mu

unter anderem sichergestellt werden da bei zwei hintereinander zu produzieren

den Stahlsorten die Uberschneidung der Analysenintervalle fur Kupfer sein

soll damit sie ohne zusatzlic he Ma nahmen hergestellt werden k onnen Was nun

wenn diese Ungleichung nicht absolut gilt und ein Wert von auch noch akzep

tiert werden k onnte insbesondere dann wenn sich alle anderen Werte in sicheren

Bereichen be nden und sich dadurch eine wesentlich bessere Produktionsreihenfolge

erg abe Was wenn wir angeben wollen da ein Wert von als sicherer

einzustufen sein soll als ein Wert von nur Durch Fuzzy Regeln k onnen

dem Planungs Programm solche akzeptablen Verletzungen von Einschr ankungen auf

DEUTSCHE KURZFASSUNG DER DISSERTATION

einfache Art und Weise mitgeteilt werden Ebenso wird dadurch festgelegt welche

Verletzung einer Einschr ankung als wie gut oder wie schlecht einzustufen sein soll

Weiters f allt es mit Fuzzy Logic sehr leicht anzugeben welche Art von Kompromis

sen erlaubt sein soll Es ist auch sehr einfach die zu beruc ksichtigenden Kriterien

verschieden stark zu gewichten um ihren unterschiedlichen Bedeutungen Rechnung

zu tragen

Weitere Aspekte die sich sehr leicht mit Fuzzy Logic modellieren lassen

betre en die Verarbeitung von M oglichkeitsverteilungen von Werten So k onnen

Gr o en deren genaue Werte zum Planungszeitpunkt noch nicht festliegen z B die

Gie geschwindigkeit w ahrend der Verarbeitung eines bestimmten Auftrags auf der

Stranggu anlage mittels M oglichkeitsverteilungen modelliert und dadurch trotzdem

fur die Planerstellung beruc ksichtigt werden Auch die Vorhersage von m oglichen

Werten bestimmter wichtiger Parameter mittels zum Teil nicht v ollig bekannten

Daten wird mittels Fuzzy Logic leichter handhabbar Man stelle sich vor die Ver

wendungsdauer fur einen bestimmten Teil einer Anlage betrage im Normalfall

Minuten kann aber abh angig von verschieden Parametern die zum Teil erst zur

Produktionszeit festgelegt werden auch nur oder aber bis zu Minuten be

tragen Nun ist es meistens fur den Menschen zu muhsam die vielen zum Teil

nur ungenau bekannten Ein u faktoren zu beruc ksichtigen um einen jeweils neuen

Sch atzwert fur diesen Parameter zu berechnen Der Computer hat es da mit Hilfe

der Fuzzy Logic wesentlich leichter sogar eine ganze M oglichkeitsverteilung fur den

Parameter zu sch atzen und bei weiteren Entscheidungen zu beruc ksichtigen Da

her wurde in dieser Arbeit auch das Wissen und das notwendige Modell fur eine

solche Aufgabe zur Ermittlung der M oglichkeitsverteilungen der voraussichtlichen

Lebensdauer eines Gie rohres als Teil einer Stranggu anlage erl autert

Die theoretische Komplexit at von zeitlichen Planungsproblemen wird

zwar durch die Wissensrepr asentation mittels Fuzzy Einschr ankungen und

M oglichkeitsverteilungen nicht reduziert da der Suchraum im allgemeinen eher

um einige Kompromi l osungen vergr ossert wird Andererseits erlaubt die gradu

elle Erfullung der Einschr ankungen den Einsatz und die e ziente Steuerung von

Such Heuristiken die sich in der Praxis bereits extrem bew ahrt haben In der

vorliegenden Arbeit wurden erstmals mehrere solche auf vollst andigen aber sub

optimalen Losungen operierende Such Heuristiken mit den Fuzzy Methoden zur

Wissensrepr asentation kombiniert und damit wesentlich bessere Ergebnisse erzielt

als mit traditionellen Such Algorithmen ohne Fuzzy Wissensrepr asentation

Die vorgestellten Methoden sind auch zur Losung zahlreicher anderer Entschei

dungsprobleme aus der Realit at sehr gut geeignet Trotzdem konzentriert sich diese

Arbeit auf das Problem der zeitlichen Planung unter Unsicherheit da die Forschung

und Entwicklung auf diesem Gebiet in den letzten Jahren einen bedeutenden Auf

schwung erlebt hat Im Anhang ndet sich eine ausfuhrlic he Bibliographie zum Teil

mit Kommentaren versehen zum Thema der Dissertation

Abstract

Real w orld scheduling is decision making under vague constraints of di erent impor

tance often using uncertain data where compromises between antagonistic criteria

are allowed The author explains in theory and by detailed examples a new combi

nation of fuzzy set based constraints and iterative improvement repair based heuris

tics that help to model these scheduling problems He simpli es the mathematics

needed for a method of eliciting the criteria s importances from human experts He

introduces a new consistency test for con guration changes This test also helps

to evaluate the sensitivity to con guration changes He describes the implementa

tion of these concepts in his fuzzy logic inference processor library FLIP in his

fuzzy constraint library ConFLIP in his dynamic constraint generation library

DynaFLIP and in his heuristic repair library D ej aVu All these libraries are

implemented in a layered framework enhanced by his common user interface Inter

FLIP The benchmark application to compare his fuzzy constraint iterative im

provement repair heuristic with constructive methods based on classic constraints is

a scheduling system for a continuous caster unit in a steel plant In addition an ear

lier fuzzy scheduling system that was applied to another steel plant as well as a fuzzy

expert system that predicts maintenance intervals for the continuous caster unit are

described This thesis also discusses research issues and challenges as well as previous

work done in the eld of fuzzy scheduling and related areas and provides an exhaus

tive and partly annotated bibliography concerning its sub ject An online version of

the thesis is located at URL ftp mira dbai tuwien ac at pub slany thesis ps Z

Keywords

Fuzzy scheduling fuzzy constraint satisfaction problems fuzzy multiple criteria op

timization fuzzy qualitative modeling fuzzy decision making trade o s compro

mising importance scale priority of constraints repair based heuristic iterative

improvement versus constructive algorithm tabu list fuzzy logic for production

control and CIM scheduling in steelmaking fuzzy resource allocation fuzzy plan

ning and design non fuzzy uncertainty management in scheduling fuzzy knowl

edge representation possibility distribution possibilistic modeling knowledge based

scheduling fuzzy constraint relaxation real world scheduling maintenance interval

prediction fuzzy expert system continuous caster scheduling fuzzy linear program

ming computational complexity benchmarking

Kurzfassung abstract in German

 

Abstract

 

Keywords

Table of Contents

List of Tables

List of Figures

Introduction

 

Fuzzy logic in arti cial intelligence

 

What is scheduling

Why fuzzy scheduling

What s new in this thesis

Organization of the thesis

Research issues and challenges in fuzzy scheduling

 

Motivation and complexity issues

 

Types of imprecision in scheduling

Fuzzy schedule construction

Research challenges in fuzzy scheduling

CONTENTS

 

Uncertainty management in production process scheduling

Introduction

Description of the steelmaking process

Constraints in steel production scheduling

Heuristics used by the experts

Schedule construction and repair

Example

Evaluation of schedules

Constructing a preliminary schedule

Improving the schedule by repair

Comparison to related systems

Conclusion

 

Fuzzy expert system to predict

 

maintenance intervals in a continuous caster

 

Introduction

Fuzzy expert system

Conclusion

Fuzzy multiple criteria representation

 

Introduction

Fuzzy multiple criteria representation

Fuzzy constraint satisfaction problems

Fuzzy constraints

Aggregating several fuzzy constraints

Fuzzy constraints of di erent importance

How to nd the importance of constraints

A consistency test for con guration changes

Decision function and con ict identi cation with DynaFLIP

Implementation issues and results with ConFLIP

Conclusion

 

CONTENTS

 

Fuzzy multiple criteria optimization

 

Introduction

Fuzzy constraint satisfaction problems revisited

Fuzzy repair

An application example

 
 

Scheduling a steelmaking plant with D ej aVu

 

Conclusion

 

Epilogue

 

 

General conclusions

Open problems and future perspectives

Annotations to the Bibliography

 

Bibliography

 

Acknowledgements

 

Curriculum Vitae

 

 

Personal data

Education

Work experience

Scienti c activities and teaching experience

List of publications

Characteristics of jobs for furnace eaf

Characteristics of jobs for furnace eaf

Fuzzy inference to compute chemical compatibility

Compatibility matrix for heat sequences on furnace eaf

Classi cation of jobs

Intermediate schedules

Algorithm to construct an initial schedule

More intermediate schedules

Algorithm to repair a schedule

 

Comparing aggregation operators

Rankings of solutions with weighted constraints

Relative importance attributes

 

More rankings of solutions with weighted constraints

Consistency test for con guration changes

Aggregates in the steelmaking plant

 

Overlapping of alloying cycles

Membership functions of soft constraints

Satisfaction taking into account priority

Outline of constraint tree constructed by DynaFLIP

InterFLIP session in XView

Million queens statistics with repair versus constructive approach

Repair based heuristic versus constructive approach

Chapter

The question of whether a computer can think is no more interesting

than the question whether a submarine can swim

Edsger W Dijkstra

A few years ago one of the authors happened to be dining in a Beverly Hills

restaurant near Mel Brooks The waitress appeared and listed the evening specials

for him One appetizer she said was yellowtail grilled on one side and raw on the

other Hey what is this It s either sushi or it isn t he cried

Daniel McNeill and Paul Freiberger Fuzzy Logic

I was fully cognizant that I was doing something that would spark controversy

Lot A Zadeh

This Chapter introduces fuzzy logic as a part of arti cial intelligence in general and

motivates the choice of the thesis s sub ject fuzzy scheduling Furthermore it gives

a condensed overview about the structure of the thesis

Fuzzy logic in arti cial intelligence

In Alan Turing wrote a paper marking the begin of a new era the era of

the intelligent machine which raised questions that still remain unanswered today

CHAPTER INTRODUCTION

This era was heavily in uenced by the appearance of the computer a machine that

allowed humans to automate their way of thinking

However human thinking is not exact If you had to park your car pr ecisely in

one place you would have extreme di culties To allow computers to really mimic

the way humans think the theories of fuzzy sets and fuzzy logic were created

They should be viewed as formal mathematical theories for the representation of

uncertainty which is essential for the management of real world systems as it mimics

the crucial ability of the human mind to summarize data and focus on decision

relevant information

Marvin Minsky one of the founding fathers of arti cial intelligence once de ned

the latter as

the science of making machines do things that would require intel li

gence if done by men

Similarly Lot A Zadeh who in wrote the founding paper on fuzzy set the

ory once described the aim of this theory as being

the construction of smarter machines

Zadeh recently coined the term MIQ machine intelligence quotient to refer to

this particular aspect of the growing number of intelligent consumer products and

industrial systems

Proponents of the so called strong arti cial intelligence believe that eventu

ally these machines will be as intelligent as we human beings are now Thinking

positively about technology everything that is conceivable to be solved by arti

cial means will eventually be realized if it is interesting enough Of course some

intellectual processes have been shown to be emergent properties such as conscious

ness The concept of emergent properties of complex systems was rst observed by

von Bertalan y in the s in his study of complex biological systems He

noticed that complex assemblies of entities organized in particular ways can reveal

unique properties not possessed by the individual entities alone Emergent proper

ties cease to exist if the whole is broken into components or if the components are

organized in a di erent way Additionally emergent properties cannot be under

stood by the study of isolated components Similar to the notion of a critical mass

in physics an emergent property will suddenly pop up when a su cient amount

of mass has been accumulated Contrary to reductionistic approaches these ap

proaches normally assume a holistic view of the world i e something complex can

be more than simply the accumulation or sum of its parts Of course as with

the atomic bomb which was in a certain sense the rst arti cial application of the

physical e ect described above the ethical aspects have to be carefully considered

One has to be aware that any technology can be used for good or for evil However

CHAPTER INTRODUCTION

not the technology in itself is good or bad but instead the humans that use it are

so since technology has so far been only a tool for human beings In the case of

intelligence this might be not true anymore since advanced intelligence may entail

new ethical needs but these new forms of intelligence have not yet reached a level

where ethical aspects become prevalent

It is important to note that the term fuzzy logic is used in two distinct senses

In its narrower sense fuzzy logic is only one branch of fuzzy set theory Fuzzy set

theory was invented by Zadeh to be able to better represent such everyday notions

as the set of tall persons Of course this set is de ned vaguely and persons will

more or less be a member of it i e member to a certain degree Fuzzy logic in this

narrow sense deals in a natural way with the representation and inference from such

vaguely formulated or uncertain knowledge similarly to classical logic which deals

with crisp knowledge where statements can only be either true or false well almost

at least if you do not count the ndings of Kurt G odel In recent years however it

has become increasingly common to employ the term fuzzy logic in a much broader

sense making the di erence between the notions of fuzzy set theory and fuzzy logic

vanish To avoid confusion we follow in this Section the trend to use fuzzy logic in

its general sense In all other Chapters the term fuzzy logic is used in its narrower

sense

James Bezdek editor in chief of the IEEE transactions on fuzzy systems de ned

fuzzy logic in a delightful essay to be one part of computational intelligence al

together with such research areas as neural networks evolutionary computation and

genetic algorithms Bezdek contrasts the ABC s on intelligence arti cial biological

and computational In the strictest sense computational intelligence depends on

numerical data supplied by manufacturers and does not rely on knowledge Ar

ti cial intelligence on the other hand uses what Bezdek calls knowledge tidbits

Heuristically constructed arti cial intelligence such as an expert system is an exam

ple Practicing knowledge engineers and neural smiths know the distinction is at

times not precise Expert extraction of feature data for training a layered perceptron

certainly falls in the area of arti cial intelligence Using these features to train the

layered perceptron is primarily computational Fuzzy inference engines crafted by

experts fall into the de nition of arti cial intelligence Algorithmic tuning of the

engine with raw data however is computational intelligence

Even though the boundary between computational intelligence and arti cial in

telligence is not distinct we can making certain assumptions monitor the volume

of research activity in each Indeed the separate identities of computational intelli

gence and arti cial intelligence are con rmed by inspection of the recent volume of

publishing and patent activity

However the term Computational Intelligence itself is not undisputed since it

had already been widely used to mean arti cial intelligence before it was rede ned

by Bezdek see for example the journal Computational Intelligence published since

CHAPTER INTRODUCTION

the conference Computational Intelligence taking place annually since

and numerous other publications and organizations using the term in this traditional

sense

In both cases arti cial intelligence as well as fuzzy logic one tries in some

sense to imitate life in its problem solving capability The ways how to achieve

this goal are di erent in many respects but there are also many common points

where the two elds overlap Robert Marks counted entries on fuzzy logic

in the INSPEC data base from to containing citations from over

selected journals books conference proceedings and technical reports of

them were cross categorized in the expert system category and with neural

networks Based on various bean countings Marks concludes that the overlapping

areas cover depending on the way to count from to

It should not be left untold that there has been a lot of scienti c antagonism

between fuzzy logic and arti cial intelligence and accordingly skeptics on both sides

exist and treat the other side with reservation if not with open hostility There are

many reasons for this e g some critics of fuzzy logic credit the word fuzzy for being

too controversial and misleading in itself others maintain that anything that can be

done with fuzzy logic and fuzzy set theory can be done equally well with classical

logic and probability theory and still others insist on denying fuzzy logic the

status of a logic itself Of course these claims were refuted or see

discussions in the archives of the news groups mentioned on page Fuzzy logic in

its narrow sense is simply a logic of fuzziness not a logic which itself is fuzzy Just

as the laws of probability are not random so the laws of fuzziness are not vague

On the other hand critics of arti cial intelligence have observed that the some

times over ambitious predictions made in the past did not come true Some even go

as far as to deny that there has been even one successful expert system implemented

that really became used Others believe that the aim to create arti cial intelligence

is useless and impossible on philosophical grounds However such views are likely

to become muted with the passage of time and a better understanding of the ba

sic ideas underlying the theories of both arti cial intelligence and fuzzy logic We

observe nevertheless that nurtured by the current success of fuzzy logic in the real

world dangerously unrealistic predictions and claims appear again Bart Kosko

a respected scholar in the eld and author of a best selling textbook on Neural

Networks and Fuzzy Systems for instance predicts for the next few decades

fuzzy logic based natural language understanding machines that write interesting

novels and screenplays in a selected style such as Hemingway s or even sex robots

with a humanlike repertoire of behavior Some researchers suggest however

that as attempts are made to make fuzzy systems larger they will encounter sim

But Cheeseman also rejects nonmonotonic reasoning default logic and

Dempster Shafer s the

ory arguing that probabilities are better suited to model the world and the other methods are

at most harmless if not outright wrong For an outline of his paper see the annotation to his

paper

CHAPTER INTRODUCTION

ilar di culties as conventional reasoning methodologies Fuzzy logic is certainly

not a philosopher s stone solving all problems that confront us today But it has

a considerable potential for practical applications The management of uncertainty

will be of growing importance This uncertainty can have various reasons ranging

from uncertainty due to the lack of knowledge or evidence due to an abundance

of complexity and information to uncertainty due to the fast and unpredictable

development of scienti c political social and other structures nowadays

The applications of fuzzy technologies fall mainly into two categories fuzzy con

trol applications which are often rather simple but very e cient fuzzy rule based

systems such as autofocusing systems in cameras washing machines automobile

transmissions subway control or even handwriting recognition In these applica

tions fuzzy logic is used as a powerful knowledge representation technique that

allows to hide unessential details and to handle uncertain data However their ef

ciency depends also heavily on the use of sensors and e ectors thus their success

should really be explained by the interaction of these various parts The second cat

egory consists of those much more complex systems that aim at supporting or even

replacing a human expert Such applications are exempli ed by medical diagno

sis systems securities funds and portfolio selection systems tra c control systems

fuzzy expert systems and fuzzy scheduling systems In this second category there

are still many problems that remain to be addressed and there is an equally press

ing need for a better understanding of how to deal with knowledge based systems

in which knowledge is both uncertain and imprecise

Areas where fuzzy logic and arti cial intelligence meet in current research in

clude fuzzy expert systems e g for medical diagnosis or intelligent tutoring sys

tems theoretical investigations e g combinations of fuzzy logic with modal logics

and other forms of defeasible reasoning i e based on questionable knowledge ma

chine learning e g combinations of fuzzy logic with neural networks genetic algo

rithms associative memories symbolic learning methods such as case based reason

ing robotics involving motion control and planning capabilities e g when ying

a fully automated helicopter or driving a car on a freeway pattern matching e g

face recognition or constraint satisfaction problem solving methods applied for

example in manufacturing process scheduling as in this thesis or in bridge design

Let us take a closer look at fuzzy expert systems as the archetypical spin o

coming from the combination of techniques from fuzzy logic and arti cial intelli

gence Classical expert systems are computer programs that emulate the reasoning

of human experts or perform in an expert manner in a domain for which no human

expert exists This could be due to a dangerous working environment or simply

because of a domain that is to large for one human being These expert systems

typically reason with uncertain and imprecise information using various methods

besides fuzzy logic to handle them There are many sources of imprecision and

uncertainty The knowledge that the expert systems embody is often not exact

in the same way as a human s knowledge is imperfect The facts or user supplied

CHAPTER INTRODUCTION

information are also often uncertain

An expert system is typically made up of at least three parts an inference

engine a knowledge base and a working memory The inference engine uses the

domain knowledge together with acquired information about a problem to provide

an expert solution The knowledge base contains the expert domain knowledge for

use in problem solving very often in form of explicit facts and IF THEN rules

A fuzzy expert system is an expert system that uses a collection of fuzzy mem

bership functions and rules to reason about data