ISOTROPIC CELLULAR AUTOMATON FOR MODELING EXCITABLE MEDIA

EXCITABLE media, exemplified by the chemical system of the Belousov-Zhabotinsky (BZ) reaction, are often modelled theoretically through the integration of sets of partial differential equations that describe their dynamics. An alternative approach is to use cellular automata1-4, which sacrifice insight into the detailed physical mechanisms for the benefit of being able to reproduce the observed patterns of behaviour at low computational cost. But the cellular automata used in previous work have been anisotropic (mainly square or hexagonal), leading to the problem that this anisotropy tends to be propagated into the patterns produced. Here we describe a means of generating isotropic cellular automata, which are able to reproduce a wide range of observed modes of behaviour ranging from spiral-type BZ patterns to structures reminiscent of turbulence. © 1990 Nature Publishing Group.