A model of artificial genotype and norm of reaction in a robotic system

Abstract

The genes of living organisms serve as large stores of information for replicating their behavior and morphology over generations. The evolutionary view of genetics that has inspired artificial systems with a Mendelian approach does not take into account the interaction between species and with the environment to generate a particular phenotype. In this paper, a genotype model is suggested to shape the relationship with the phenotype and the environment in an artificial system. A method to obtain a genotype from a population of a particular robotic system is also proposed. Finally, we show that this model presents a similar behavior to that of living organisms in what regards the concept of norm of reaction.