An Architecture to Address Uncertain Requirements and Composability for Intelligent Agents in Distributed Simulations

摘要

In this paper we describe an architecture for computer-generated intelligent agent actors (CGAs) that exploits the technological advantages provided by object-orientation, component software, software frameworks, gauges, migration assistants, containerization, and rapid prototyping. These technologies allow the architecture to support composability, flexibility, re-usability, and generality for the CGAs that the architecture supports. The architecture consists of highly modular components and two frameworks where interdependencies are well defined, documented, and minimized. In this paper we discuss the architecture, its constituent elements and their functionality, and discuss some conclusions about CGA architectures.