Please use this identifier to cite or link to this item:
|Title:||Analysis of genotype size for an evolvable hardware system|
|Keywords:||Evolvable hardware;Genotype size;Computational intelligence;Design of logic circuits|
|Publisher:||World Academy of Science, Engineering and Technology|
|Citation:||5th International Enformatika Conference (IEC 05), Prague, Czech Republic, pp. 74 - 79, 26-28 August 2005|
|Abstract:||The evolution of logic circuits, which falls under the heading of evolvable hardware, is carried out by evolutionary algorithms. These algorithms are able to automatically configure reconfigurable devices. One of main difficulties in developing evolvable hardware with the ability to design functional electrical circuits is to choose the most favourable EA features such as fitness function, chromosome representations, population size, genetic operators and individual selection. Until now several researchers from the evolvable hardware community have used and tuned these parameters and various rules on how to select the value of a particular parameter have been proposed. However, to date, no one has presented a study regarding the size of the chromosome representation (circuit layout) to be used as a platform for the evolution in order to increase the evolvability, reduce the number of generations and optimize the digital logic circuits through reducing the number of logic gates. In this paper this topic has been thoroughly investigated and the optimal parameters for these EA features have been proposed. The evolution of logic circuits has been carried out by an extrinsic evolvable hardware system which uses (1+lambda) evolution strategy as the core of the evolution.|
|Appears in Collections:||Dept of Electronic and Computer Engineering Research Papers|
Items in BURA are protected by copyright, with all rights reserved, unless otherwise indicated.