Collective fission behavior in swarming systems with density-based interaction

Abstract

In this paper we revisit the self-organized collective behavior of swarms with density field interaction that is inspired from the model of smoothed particle hydrodynamics. For a homogeneous swarming system a novel collective fission behavior is seen to emerge where equal to or more than two sub-clusters spontaneously come from a single connected cluster. The focus of this study is on the analysis of collective fission behavior in trigger conditions, function manners and dynamic characteristics. For the first time, we are able to predict the stability of a single connected cluster, the trigger conditions of collective fission behavior, the number of sub-clusters and the group size of each sub-cluster. The analysis results also show that the collective fission behavior is irreversible by changing the control parameters and the group size. However, it is reversible when the sensing range is enlarged to encompass all sub-clusters. These characteristics are evaluated in simulations as well as on the real swarm robotic system. The study of collective fission behavior provides an elegant insight for how self-organized segregation emerges in biological swarms and artificial systems.