Assumption University
Bangkok, Thailand
November 1996
GENETIC-BASED
SELF-TUNING CONTROLLER
by
Pataya Dangprrasert
A doctoral dissertation submitted in partial
fulfillment of the requirements for the degree of Doctor of Philosophy
in Computer and Engineering Management.
- Examination Committee :
- Prof. Dr. Srisakdi Charmonman
- Assoc.Prof. Somchai Tayarnyong
- Air Marshal Dr.Chulit Meesajjee
- Dr.Suphamit Chittayasothorn
- Dr.Pornchai Phutlek
- Dr.Prapon Phasukyud
- Dr.Aran Namphol
- Dr.Vichit Avatchanakorn
- Name :
- Pataya Dangprasert
- Nationality :
- Thai
- Previous Degree :
- Master of Sciences in CIS
- Assumption University, Thailand
- Bachelor of Sciences (Pharmacy)
- Oregon State University,USA
- Bachelor of Sciences (Gen.Sciences)
- Kasetsart University, Thailand
Approval Page
RESEARCH TITLE : GENETIC-BASED SELE-TUNING CONTROLLER
CANDIDATE NAME : Pataya Dangprasert
ADVISOR NAME : Dr.Vichai Avatchanakorn
ACADEMIC YEAR : 1996
The Graduate School of Assumption University had
approved this dissertation as a partial fulfillment of the requirement
for the degree of Doctor of Philosophy in Computer and Engineering Managemant
- Abstract
- Recently,the replacement of intelligent control
in place of conventional control has increased due to the latter often
failing to detect changes in adaptive environment. Among various intellgence
control concepts, self-tuning regulator (STR) seems to be the most
practical to implement. However, the success of STR has been laid
on recursive off-line methods which are local search schemes based
on a gradient-following technique.
- Genetic algorithms (GAs) are search techiques
based on natural genetic evolution.By simulating evolutionary process,
GAs can bring the power of population genetics to provide autonomous
search for artificial systems. They have been used as general-purpose
optimization methods that can avoid local optima.
- This research brings the adaptive search power
of GAs and combines it with the concept in controller design provided
by STR to propose a new intelligent controller, the genetic-based
self-tuning controler. In STR controller design approach, the controller
is separated into estimator loop and regulator loop. Genetic-based
self-tuning controller uses GAs as parameter identifying scheme in
estimator loop and uses the single-input single-output model to represent
the process dynamics of a controlled system for designing the regulator.
- Results of computer simulations by using load frequency
control (LFC) of the power system as a domain problem, show efficient
and robust control power of genetic-based self-tuning controller as
an on-line parameter identification without any prior input-output
knowledge. It is also capable of controlling a system with non-linearity
characteristics.
- Acknowledgements
- I express my sincere appreciation to my advisor,
Dr. Vichit Avatchanakorn,for his continuous advice, guidance,and support.
His profound knowledge and kindness will always be an inspiration.
- Thanks are also due to Prof. Dr.Srisakdi Charmonman
for founding the Ph.D.Programs at ABAC. He was the first one who cncouraged
me to apply to this program and he further provided valuable advice
throughout my study.
- Many thanks are extended to Associate Professors
Somchai Tayarnyong, Air Marshall Dr. Chulit Meesajjee, Dr. Suphamit
Chittayasothorn, Dr.Prapon Phasukyud,Dr.Aran Namphol, and Dr.Pornchai
Phutlek for their careful reading of this manuscript,and to Dr.Boonmark
Sirinaovakul for serving on my qualifying examination committee. Their
constructive and valuable comments have made this work better in various
aspects.
- I am grateful to Rev. Dr.Prathip M.Komolmas,
(AU president) for his constant encouragement.
- The friendship and encouragement of my fellow colleagues,
Ms. Pornthip Archadej, and Ms.Somsong Chirapahanakul have been important
forces in my study. Special thanks to Dr.Kamales Santivejkul for his
kind words of advice,and my friend, Ms. Nalinee Nandhabiwat, for her
time in second language tutorials. Thanks are also to Mr.Uttrawooth
Narknisorn for his support.
- Table of Contents
- List of Figures
- list of Table
- Nomenclatures
- Chapter 1 : Introduction
- Chapter 2 : Genetic Algorithms
-
- 2.1 Introduction
- 2.2 Simple Genetic Algorithms
- 2.3 Genetic Algorithms Applidations
- 2.4 Conclusions
- Chapter 3 : Design of a Genetic-Based Self-Tuning
Controller
-
- 3.1 Introduction
- 3.2 Controller Design
- 3.3 Self-Tuning Regulator
- 3.4 Genetic-Based Self-Tuning Controller
Design
-
- 3.4.1 Conceptual Fundamentals of the Genetic-Based
Self-Tuning Controller
- 3.4.2 Operational Aspect of the Genetic-Based
Self-Tuning Controller
- 3.5 Conclusions
- Chapter 4 : Application to a Load Frequency
Control System
-
- 4.1 Introduction
- 4.2 Background of the Problem Domain-LFC
- 4.3 Application to One-Area LFC System
-
- 4.3.1 Designing Genetic-Based Self-Tuning
Load Frequency Controller
- 4.3.2 Sensitivity Analysis
- 4.3.3 Output Responses
- 4.4 Application to Two-Area Interconnected LFC
System
- 4.5 Conclusions
- Chapter 5 : Conclusion and Recommendations
-
- 5.1 Conclusions
- 5.2 Recommendations
- Appendix A : Schemata Theorems
- Appendix B : Load Frequency Control Model
- Appendix C : Runge-Kutta Methods
- References
|
