Hidden genes genetic algorithms for systems architecture optimization

Abstract

The concept of hidden genes was recently introduced in genetic algorithms to handle variable-size design space optimization problems. This paper presents new developments in hidden genes genetic algorithms. Mechanisms for assigning (selecting) the hidden genes in the chromosomes of genetic algorithms are presented. In the proposed mechanisms, a tag is assigned for each gene; this tag determines whether the gene is hidden or not, while they evolve over generations using stochastic operations. These mechanisms are tested on mathematical optimization problems and on a trajectory optimization problem for a space mission to Jupiter. In the conducted tests, one of the proposed hidden genes assignment mechanism has enabled the hidden genes genetic algorithms to find better (lower cost) solutions, while other mechanisms has shown to be able to find close solutions.