Masters Theses

Abstract

"In this research, Clonal Selection, an immune system inspired approach, is utilized along with Evolutionary Algorithms to solve complex engineering problems such as Intrusion Detection and optimization of Flexible AC Transmission System (FACTS) device placement in a power grid. The clonal selection principle increases the strength of good solutions and alters their properties to find better solutions in a problem space. A special class of evolutionary algorithms that utilizes the clonal selection principle to guide its heuristic search process is termed Clonal EA. Clonal EAs can be used to solve complex pattern recognition and function optimization problems, which involve searching an enormous problem space for a solution. Intrusion Detection is modeled, in this research, as a pattern recognition problem wherein efficient detectors are to be designed to detect intrusive behavior. Optimization of FACTS device placement in a power grid is modeled as a function optimization problem wherein optimal placement positions for FACTS devices are to be determined, in order to balance load across power lines. Clonal EAs are designed to implement the solution models. The benefits and limitations of using Clonal EAs to solve the above mentioned problems are discussed and the performance of Clonal EAs is compared with that of traditional evolutionary algorithms and greedy algorithms"--Abstract, page iii.

Advisor(s)

Tauritz, Daniel R.

Committee Member(s)

McMillin, Bruce M.
Miller, Ann K.

Department(s)

Computer Science

Degree Name

M.S. in Computer Science

Publisher

University of Missouri--Rolla

Publication Date

Summer 2005

Pagination

ix, 84 pages

Note about bibliography

Includes bibliographical references (pages 82-83).

Rights

© 2005 Parthasarathy Kasthurirangan, All rights reserved.

Document Type

Thesis - Open Access

File Type

text

Language

English

Library of Congress Subject Headings

Clonal selection theory
Flexible AC transmission systems
Pattern recognition systems
Evolutionary computation

Thesis Number

T 8776

Print OCLC #

62616855

Electronic OCLC #

1121204399

Share

 
COinS