作者: Alban Rousset , Bénédicte Herrmann , Christophe Lang , Laurent Philippe
DOI: 10.1007/978-3-319-14325-5_32
关键词: Multi-agent system 、 Qualitative analysis 、 Computer science 、 Supercomputer 、 Complex system 、 Computation 、 Focus (optics) 、 Distributed computing 、 Context (language use)
摘要: Simulation has become an indispensable tool for researchers to explore systems without having recourse real experiments. Depending on the characteristics of modeled system, methods used represent system may vary. Multi-agent are, thus, often model and simulate complex systems. Whatever modeling type used, increasing size precision increases amount computation, requiring use parallel when it becomes too large. In this paper, we focus platforms that support multi-agent simulations. Our contribution is a survey existing their evaluation in context high performance computing. We present qualitative analysis, mainly based platform properties, then comparison using same agent implemented each platform.