An intermittent behavior is shown to appear in a minimal model for a system of self-driven interacting particles. While the system is in the ordered phase, with most particles moving in approximately in the same direction, it displays a series of intermittent bursts during which the order is lost. The intermittency is characterized and its statistical properties are found analytically for a reduced system containing only two particles. On large systems, the particles aggregate into clusters. The role of these clusters in the system dynamics is studied. The cluster sizes are shown to follow a power-law distribution. The transition probabilities between different cluster sizes are found to also follow power-law distributions and to satisfy detailed balance. The results are discussed using a dynamic interaction-network perspective.
Copyright. All Rights Reserved.