Conversely, generative encodings allow the reuse of genotype portions that code for similar or identical phenotype components. Direct encodings represent each phenotype component independently in the genotype. There are two main classes of genetic encodings, namely, direct encodings and indirect encodings the latter are also known as generative encodings. Our principal contributions are creating awareness about robot encoding biases, demonstrating how such biases affect evolved morphological, control, and behavioral traits, and finally scrutinizing the trade-offs among different biases. In summary, we demonstrate how two generative encodings of different nature lead to very different robots and discuss these differences. In this paper, we investigate how genetic encodings constrain the space of robot phenotypes and robot behavior. Notably, this anchoring of the literature to performance is limiting, considering that performance provides bounded information about the behavior of a robot system. However, the literature has widely focused on studying the effects that encodings have on performance, i.e., fitness-oriented studies. Genetic encodings and their particular properties are known to have a strong influence on the success of evolutionary systems. Computer Science Department, Vrije Universiteit Amsterdam, Amsterdam, Netherlands.
0 kommentar(er)
