The collective behavior of swarms is extremely difficult to estimate or predict, even when the local agent rules are known and simple. The presented work seeks to leverage the similarities between fluids and swarm systems to generate a thermodynamics-inspired characterization of the collective behavior of robotic swarms. While prior works have borrowed tools from fluid dynamics to design swarming behaviors, they have usually avoided the task of generating a fluids-inspired macroscopic state (or macrostate) description of the swarm. This work will bridge the gap by seeking to answer the following question: is it possible to generate a small set of thermodynamics-inspired macroscopic properties that may later be used to quantify all possible collective behaviors of swarm systems? In this paper, we present three macroscopic properties analogous to pressure, temperature, and density of a gas, to describe the behavior of a swarm that is governed by only attractive and repulsive agent interactions. These properties are made to satisfy an equation similar to the ideal gas law, and also generalized to satisfy the virial equation of state for real gases. Finally, we investigate how swarm specifications such as density and average agent velocity affect the system macrostate.

翻译：集群的集体行为极难估计或预测，即使局部智能体规则已知且简单亦然。本工作旨在利用流体与集群系统之间的相似性，生成一种受热力学启发的机器人集群集体行为表征。尽管先前研究已借用流体动力学工具设计集群行为，但通常避免生成类流体的集群宏观状态（或称宏观态）描述。本文将通过回答以下问题来弥合这一差距：是否可能生成一组受热力学启发的宏观性质，进而用于量化集群系统的所有可能集体行为？本文提出了三种宏观性质——类比气体的压强、温度和密度——用以描述仅由吸引和排斥智能体交互主导的集群行为。这些性质被构造为满足类似理想气体状态方程的公式，并进一步推广至符合真实气体的维里物态方程。最后，我们研究了集群规格（如密度和平均智能体速度）对系统宏观状态的影响。