Axiomatic-Design-Theory-Based Approach to Modeling Linear High Order System Dynamics

This paper presents a systematic approach to modeling linear high-order system dynamics for the purpose of control. The philosophy behind this approach is that system modeling can be made analogous to system design. In particular, functional requirements (in design) are analogous to dynamics (in modeling) of the system to be modeled and design parameters (in design) correspond to models (in modeling). With this philosophy, we assume that a model can be decomposed into a set of basic models. We then apply the axiomatic design theory (ADT) developed by Suh in the late 1980s, in particular independence axiom of ADT, to determine these basic models. Four experiments were conducted, and the results have shown that our approach is very promising, as opposed to the existing approaches, in terms of model accuracy and model development efficiency.