Publication de 12 numéros par an
ISSN Imprimer: 0040-2508
ISSN En ligne: 1943-6009
Indexed in
RECIPE FOR AN EFFICIENT HYBRID GENETIC ALGORITHM
RÉSUMÉ
The elegance demonstrated by Nature when solving real-world problems fascinates and inspires imitation. A number of algorithms that mimic Nature's selection mechanisms have been introduced recently and quickly proved to be powerful instruments for electromagnetic synthesis. There are many papers discussing pros and cons of evolutionary optimization algorithms (EOA) and most of them do this via contrasting one against others whereas their similarity and complementarity often remain in shadow. This paper gives a bird's-eye view on the evolutionary optimization strategy considered as a probability cheating approach. The main questions addressed in this paper are the general logic behind the choice of EOA parameters and the ways of boosting algorithms efficiency. Discussion is supported by introduction of a hybrid genetic algorithm capable of fast optimization of multi-parameter and multi-extrema functions typical for electromagnetics.
-
Dehu Qi , Ron Sun , Cooperative agents through bidding, Proceedings of the 2004 IEEE International Conference on Information Reuse and Integration, 2004. IRI 2004., 2004. Crossref