Open AccessDesign theoretic analysis of three system modeling frameworks.
Author(s) -
Michael McDonald
Publication year - 2007
Language(s) - English
Resource type - Reports
DOI - 10.2172/908075
Subject(s) - computer science , modular design , scalability , construct (python library) , context (archaeology) , domain (mathematical analysis) , systems engineering , complex system , decomposition , loose coupling , distributed computing , systems modeling , relation (database) , theoretical computer science , software engineering , data mining , artificial intelligence , engineering , programming language , mathematics , ecology , paleontology , mathematical analysis , database , biology
This paper analyzes three simulation architectures from the context of modeling scalability to address System of System (SoS) and Complex System problems. The paper first provides an overview of the SoS problem domain and reviews past work in analyzing model and general system complexity issues. It then identifies and explores the issues of vertical and horizontal integration as well as coupling and hierarchical decomposition as the system characteristics and metrics against which the tools are evaluated. In addition, it applies Nam Suh's Axiomatic Design theory as a construct for understanding coupling and its relationship to system feasibility. Next it describes the application of MATLAB, Swarm, and Umbra (three modeling and simulation approaches) to modeling swarms of Unmanned Flying Vehicle (UAV) agents in relation to the chosen characteristics and metrics. Finally, it draws general conclusions for analyzing model architectures that go beyond those analyzed. In particular, it identifies decomposition along phenomena of interaction and modular system composition as enabling features for modeling large heterogeneous complex systems