Fuse: a multi-agent simulation environment
Kensuke Kuramoto, Masakazu Furuichi · Winter Simulation Conference · 2013
We propose a new integrated multi-agent simulation environment which is designed for hierarchical organization behavior modeling. Recently, multi-agent simulations are getting more important in many fields such as planning of a large-scale disaster evacuation and military operations (Cil 2010, Nakajima 2008, Persons 2005). Since such simulations are characterized that they are composed of multiple of heterogeneous organizations, and the number of agents is very large, we have to provide an effective and efficient simulation environment. We have focused on a decision making model of organizations in the real world (Kuramoto 2012), and have implemented FUSE in Java. We have also proposed CaSPA which is based on a goal directed reasoning algorithm, and implemented reasoning rules in Java to show the basic functionality. Moreover, we have expanded foreign language interface to FUSE, and implemented CaSPA's reasoning rules in JRuby to show the applicability and expandability of CaSPA.