Reinventing the wheel, or evolving to the wheel

Conteúdo do artigo principal

Francisco Antonio Pereira Fialho
Luis Alberto Gómez

Resumo

Genetic Algorithms is based on the assumption that nature can teach us how to develop artifacts able to comply with restrictions imposed by the environments. Those artifacts can be concrete or mental. In this article the authors propose the concept of Geometric Genetic Programming (GGP), which opens a wide variety of opportunities for applying the concepts of Genetic Programming (GP) to engineering problems. This papers shows how it is possible to evolve to well known, or even completely new designs starting just with a set of physical laws and a fitness function. To show how GGP works the challenge is to produce the most common machine in nature, the wheel.

Detalhes do artigo

Seção
Engenharia do Conhecimento

Artigos mais lidos pelo mesmo(s) autor(es)

1 2 > >>