Collective Dynamics of Active Elements: Task Allocation and Pheromone Trailing

摘要

Collective behavior of active elements inspired by mass of biological organisms is addressed. Especially, two topics are focused on among amazing behaviors performed by colony of ants. First, task allocation phenomena are treated from the viewpoint of proportion regulation of population between different states. Using a dynamical model consisting of elements and external ``stock materials'', adaptability against various disturbances is numerically studied. In addition, a dynamical model for a colony ants interacting via two kind of pheromones is studied, in which simulated ants, as a mass, are shown to make an efficient foraging flexibly varying the foraging tactics according to feeding schedules. Finally, experiments are performed with robots moving in virtual pheromone fields simulated by CG and CCD camera feedback system. Trail formation processes are demonstrated by this multi-robot system.