Roadmap and research issues of multiagent social simulation using high-performance computing
Roadmap and research issues of multiagent social simulation using high-performance computing
复制标题
DOI：
10.1007/s42001-017-0011-8
复制
发表时间：
2018-01-01
影响因子：
3.2
通讯作者：
Hattori, Hiromitsu
中科院分区：
文献类型：
--
作者：
Noda, Itsuki;Ito, Nobuyasu;Hattori, Hiromitsu
在这篇文章中，我们提出的路线图和研究问题有关的多智能体社会模拟，以说明在该领域的技术成果的方向。与物理仿真相比，社会仿真还处于仿真模型的建立阶段。我们专注于四个问题，即“未确定的模型”，“意识效应”，“模糊的边界”，和“不完整的数据”，并考虑如何克服这些问题，使用高性能计算的巨大计算能力。我们选择了三个应用程序，即疏散，道路交通和市场，并估计所需的计算成本的真实的应用程序。此外，我们调查的应用方面的研究问题，并对多智能体社会模拟未来可能的工作进行分类。
In this article, we present roadmaps and research issues pertaining to multiagent social simulation to illustrate the directions of technological achievements in that domain. Compared with physical simulation, social simulation is still in the phase of establishing simulation models. We focus on four issues, namely "undetermined model", "awareness effects", "obscure boundary", and "incomplete data", and consider ways to overcome these issues using the massive computational power of high-performance computing. We select three applications, namely evacuation, road traffic, and market, and estimate the required computational cost of real applications. Moreover, we investigate research issues on the application side and categorize possible future works on multiagent social simulations.