Optimizing Self-Organizing Control Architectures with Genetic Algorithms: The Interaction Between Natural Selection and Ontogenesis

To model the relationship between natural selection and ontogenesis, techniques from genetic algorithms and neural networks can be combined. We model this interaction as two complementary processes were the adaptation of the individual takes place to adjust to properties of the system environment interaction that cannot be predicted from the perspective of the genome. Adaptation of the individual during ontogenesis is modeled with a control architecture for an autonomous agent developed in the distributed adaptive control framework. Natural selection is modeled with a parallel genetic algorithm. The results indicate that \species" can evolve as an emergent property of the interaction of these two adaptational processes.