Hierarchically distributed evolutionary algorithm for optimization in nonstationary environments시변환경에서의 최적화를 위한 계층적 분산진화알고리즘

Cited 0 time in webofscience Cited 0 time in scopus
  • Hit : 425
  • Download : 0
Due to their comprehensive and vigorous search capabilities, Evolutionary Algorithms (EAs) have been widely used in a diversity of applications with great success. The majority of research up to now has focused on stationary type environments, in which the evaluation function and problem constraints are fixed over the period of evolution. As a matter of course, most applications of EAs deal with static optimization problems that can be solved off-line processing. However, many real-world problems involve time-varying objectives that should be dynamically optimized. Therefore, there has been a growing interest in time dependent optimization for many real-world applications. The dynamic optimization problem demands that EAs should generate an acceptable solution within an allowable time period. To overcome this difficulty, this thesis proposes an optimization problem solving system characterized by hierarchically distributed evolutionary algorithm (HDEA). HDEA consists of a set of subpoplations and a master evolution system. The master evolution system controls the optimization processing of the subpopulations, while subpopulations evolve independently to find optimal solution using parallel computing system. This enables us to utilize fast local search operators while maintaining sufficient population diversity. The algorithm can be easily implemented using a set of general-purpose computers connected by a network. This extends the practicality of EAs to real-time problems. The efficiency and usefulness of HDEA are analyzed using a set of benchmark problems. Through the comparison between HDEA and traditional EAs, we verify that HDEA is a particularly adaptive tool for optimizing a lot of diversified class of functions. We also apply the algorithm to optimize a nonstationary time series prediction model.
Advisors
Lee, Ju-Jangresearcher이주장researcher
Description
한국과학기술원 : 전기및전자공학전공,
Publisher
한국과학기술원
Issue Date
2005
Identifier
249401/325007  / 000965234
Language
eng
Description

학위논문(박사) - 한국과학기술원 : 전기및전자공학전공, 2005.8, [ viii, 131 p. ]

Keywords

Time Series Prediction; Nonstationary Environments; Evolutionary Algorithm; Distributed Computing; 분산처리; 시계열예측; 시변환경; 진화연산알고리즘

URI
http://hdl.handle.net/10203/35314
Link
http://library.kaist.ac.kr/search/detail/view.do?bibCtrlNo=249401&flag=dissertation
Appears in Collection
EE-Theses_Ph.D.(박사논문)
Files in This Item
There are no files associated with this item.

qr_code

  • mendeley

    citeulike


rss_1.0 rss_2.0 atom_1.0