HOME    About this site    mypage    Japanese    library    university    Feedback

University of the Ryukyus Repository >
Faculty of Engineering >
Bulletin >
Bulletin of the Faculty of Engineering, University of the Ryukyus >
No. 59 (2000/03) >

 
Title :A Hybrid Optimization Method Using a Real-coded Multi-parent GA, Simplex & Simulated Annealing with Applications in Resolution of Overlapped Signals
Authors :Mendoza, Noel
Chen, Yen-Wei
Nakao, Zensho
Adachi, Tatsuhiro
Masuda, Yoshihisa
Authors alternative :陳, 延偉
仲尾, 善勝
Issue Date :Mar-2000
Abstract :This paper presents a GA algorithm for function optimization that combines the features of Simplex crossover and Simulated Annealing. Hybridization with a local search tool like Simplex directs the GA to more promising search spaces while preventing premature local minima convergence. The use of Simulated Annealing(SA) as a mutation operator further enhances the GA's overall capability. This elitist real-coded multi-parent model is able to address the issue of slow convergence rate very well and is robust in optimizing epistatic multimodal problems. In addition to test functions, our model was applied to the resolution of overlapping signals.
Type Local :紀要論文
ISSN :0389-102X
Publisher :琉球大学工学部
URI :http://hdl.handle.net/20.500.12000/1958
Citation :琉球大学工学部紀要 no.59 p.87 -99
Appears in Collections:No. 59 (2000/03)

Files in This Item:

File Description SizeFormat
No59p87.pdf1600KbAdobe PDFView/Open