Pattern dynamics and optimization by reaction diffusion systems

Reaction-diffusion systems show a fast and rather complex response on patterns produced by external space- and/or time-dependent perturbations. For example, one-component autocatalytic reactions rapidly find the loci where the given space-dependent reaction rates have relatively high values by following a kind of Darwinian strategy (combining self-reproduction and diffusion). It is shown that a simulation of this strategy in combination with annealing (decreasing the diffusion rates in time) may be used as an alternative to thermodynamic annealing strategies. Many-component reactions, such as the light-sensitive Belousov-Zhabotinsky reaction, show a more complex response to patterns impressed by illumination, for example. The response behavior and possible applications to dynamic information processing are discussed.