Hierarchical System Concepts for Simulation of High Autonomy Systems

Abstract

A high autonomy system is an intelligent system with the self-determination power. Such a system must perform a variety of functions such as reasoning, planning, sensing, control, and other activities necessary to achieve predefined system objectives over an extended period of time under uncertainties in its environment. Simulation modeling can be a powerful tool for design of such complex autonomy systems. This paper describes the hierarchical, modular system concepts for simulation of high autonomy systems with an example of an intelligent control system called the AIDECS(AI-based, Distributed Environmental Control System). The paper emphasizes on development of methods, in an object-oriented manner, for isomorphic models. We show how polymorphism and inheritance in the object-oriented programming can be exploited to develop such methods in DEVS-Scheme, a realization of the DEVS formalism in a LISP-based, object-oriented framework.