Function optimization in nonstationary environment using steady state genetic algorithms with aging of individuals

Abstract

The authors explore the utility of the concept of aging of individuals in the context of steady state GAs for nonstationary function optimization. Age of an individual is used as an additional factor in addition to the objective functional value in order to determine its effective fitness value. Age of a newly generated individual is taken as zero, and in every iteration it is increased by one. Individuals undergoing genetic operations are selected based on the effective fitness value, which changes dynamically. This helps to maintain diversity in the population and is useful to trace changes in environment. Simulation results show some promise for the utility of the present technique for nonstationary function optimization.