Seven Constructs for Modeling Single-Agent Behavior

Abstract

Introduces stepwise the seven basic constructs used to model single-agent systems by Abstract State Machines (ASMs). It starts from scratch and is focussed on the central role of the ASM abstraction and refinement concept. Via small control system examples, it is explained how to use those constructs a) to build abstract models for such systems, and b) to refine them in a controllable manner to executable code. It is illustrated how ASMs support reliable requirements capture, controlled reuse of models, integration of data and control features, separation of normal behavior from error handling, and componentwise construction.