An Investigation into how concepts of modularity aff�ect the evolution of complex morphologies

Download

PDF - Submitted Version Available under License Creative Commons Attribution Non-commercial. Download (616Kb)

Description/Abstract

There are many di�erent ways in which complex morphologies can be represented. While a simple string representation could be su�cient, often the most impressive arti�cial life simulations utilise. Context Free Grammars (1994, Karl Sims) or Recursive Tree Structures. When modelling a complex morphology using these encodings, it is possible to harness the creatures complex modularity to create more sensible and �t individuals. This article aims to compare and contrast the varying a�ects of evolutionary algorithms which utilise or disregard the organisms modularity.