DocumentCode :
2351026
Title :
Fitness Evaluation Expansion to Enhance GA´S Performance in Evolvable Hardware
Author :
Benkhelifa, Elhadj ; Pipe, Anthony ; Nibouche, Mokhtar ; Dragffy, Gabriel
Author_Institution :
Bristol Robot. Lab., Univ. of the West of England, Bristol
fYear :
2008
fDate :
22-25 June 2008
Firstpage :
27
Lastpage :
31
Abstract :
In this paper the authors introduce a novel experimental method when using genetic algorithms (GAs) to design and optimise digital hardware (HW). This approach proved to enhance the GA´s performance. It produced more design solutions than selected comparable techniques used by others and maximises optimisation. The novel aspect of our algorithm works at the fitness evaluation stage, where every single unit´s output in the evolved array is taken as a potential solution, instead of specifying specific cells for desired outputs. This will be referred to in this paper as fitness evaluation expansion (FEE). The FEE approach isevaluated experimentally by evolving a 2bit Multiplier and 2bit binary Adder.
Keywords :
adders; genetic algorithms; logic design; multiplying circuits; binary adder; digital hardware; evolutionary electronics; evolvable hardware; fitness evaluation expansion; genetic algorithms; multiplier; word length 2 bit; Adaptive systems; Algorithm design and analysis; Circuits; Design methodology; Design optimization; Genetic algorithms; Hardware; Humans; NASA; Space technology; Evolutionary Electronics; Evolvable Hardware; Fitness Evaluation Expansion; GA Perfomance; Genetic Algorithms;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Adaptive Hardware and Systems, 2008. AHS '08. NASA/ESA Conference on
Conference_Location :
Noordwijk
Print_ISBN :
978-0-7695-3166-3
Type :
conf
DOI :
10.1109/AHS.2008.50
Filename :
4584251
Link To Document :
بازگشت