First published 2 September 2013
Evolution of Spatial Pattern Formation by Autonomous Bio-Inspired Cellular Controllers
Payam Zahadat, Thomas Schmickl, Karl Crailsheim
In this paper, a gene regulatory network called FGRN (Fractal Gene Regulatory Network) and a reaction-diffusion system called AHHS (Artificial Homeostatic Hormone System) are investigated for spatial pattern formation. The two bioinspired controllers possess similar and different features in terms of their underlying processes, structures, and communication abilities. By comparing their behaviours and capabilities in pattern formation, we provide a deeper understanding of the effects of their features. The controllers are evolved and investigated for producing various patterns in presence of different implicit positional information as well as developing a memory in order to keep the desired pattern when the positional information is eliminated. The behaviours of the controllers in each case are discussed and a preliminary test of robustness is performed. The experiments represent a positive impact of diffusion process in AHHS that is compensated by the complex structure of FGRN in producing patterns in presence of positional information and a negative effect of diffusion process in memory capability.