Department of Mechanical Engineering, California State University, Northridge, California 91330, USA.
Engineering Mechanics Program, Virginia Tech, Blacksburg, Virginia 24061, USA.
Phys Rev E. 2019 Dec;100(6-1):062415. doi: 10.1103/PhysRevE.100.062415.
In the present study, we consider two independent sensing modes (auditory and visual) in Vicsek-like models and compare the emergent group-level behaviors in terms of polarization, cohesion, and cluster size. The auditory and visual modes differ in the determination of particle neighbors, which at the level of groups results in higher polarization, lower cohesion, and larger cluster size for the auditory mode relative to the visual. With the increase in average density of the particles, these differences are more pronounced. These differences are due to the fact that these sense modalities robustly generate distinct spatial distributions of the particles. We demonstrate the use of a data-driven approach, called transfer entropy, to distinguish the sensing modalities by considering only a pair of particle trajectories. Such an approach could be applicable to real-world systems, where it may be a challenge to measure the position and velocity of every particle within a swarm.
在本研究中，我们考虑了 Vicsek 模型中的两种独立的感应模式（听觉和视觉），并比较了在极化、内聚和簇大小方面的群体级行为。听觉和视觉模式在确定粒子邻居方面有所不同，这导致在群体水平上，听觉模式的极化程度更高，内聚程度更低，簇大小更大。随着粒子平均密度的增加，这些差异更加明显。这些差异是由于这些感觉模式能够稳健地产生粒子的不同空间分布。我们展示了一种数据驱动的方法，称为转移熵，通过仅考虑一对粒子轨迹来区分感应模式。这种方法可以应用于现实世界的系统，在这些系统中，测量群体中每个粒子的位置和速度可能是一项挑战。
