대한전기학회:학술대회논문집 (Proceedings of the KIEE Conference)
- 대한전기학회 1995년도 하계학술대회 논문집 B
- /
- Pages.1006-1010
- /
- 1995
유성생식 유전알고리즘 : 다중선택과 이배성이 탐색성능에 미치는 영향
Sexual Reproduction Genetic Algorithms: The Effects of Multi-Selection & Diploidy on Search Performances
- 류근배 (효성중공업(주) 기술연구소) ;
-
최영준
(효성중공업(주) 기술연구소)
;
- 김창업 (효성중공업(주) 기술연구소) ;
- 이학성 (효성중공업(주) 기술연구소) ;
- 정창기 (한국전력공사 기술연구원)
- Ryu, K.B. (R&D Institute, Hyosung Industries Co., Ltd) ;
-
Choi, Y.J.
(R&D Institute, Hyosung Industries Co., Ltd)
;
- Kim, C.E. (R&D Institute, Hyosung Industries Co., Ltd) ;
- Lee, H.S. (R&D Institute, Hyosung Industries Co., Ltd) ;
- Jung, C.K. (Research Center, Korea Electric Power Co.)
- 발행 : 1995.07.20
초록
This paper describes Sexual Reproduction Genetic Algorithm(SRGA) for function optimization. In SRGA, each individual utilize a diploid chromosome structure. Sex cells(gametes) are produced through artificial meiosis in which crossover and mutation occur. The proposed method has two selection operators, one, individual selection which selects the individual to fertilize, and the other, gamete selection which makes zygote for offspring production. We consider the effects of multi-selection and diploidy on search performance. SRGA improves local and global search(exploitation and exploration) and show optimum tracking performance in nonstationary environments. Gray coding is incorporated to transforming the search space and Genic uniform distribution method is proposed to alleviate the problem of premature convergence.
키워드