Individual aging in genetic algorithms

Abstract

A concept of age of individuals for measuring their suitability for participation in genetic operations for steady state GAs is introduced. Effective fitness of an individual depends both on its functional value and age. Age of a newly generated individual is taken as zero and every iteration is increased by one. As in nature, adult individuals are considered more fit for genetic operations, compared to young and old ones. The model aims to emulate the natural genetic system in a more natural way. The effectiveness of this concept is demonstrated by solving complex function optimization problems. Results show that the scheme provides enhanced performance and maintains more diversity in the population thereby allowing the species to be robust to trace the changing environment.