Decomposition method in system dynamics modeling & block diagram conversion

Abstract

This paper shows a research of the dynamic mechanism underlying the related behaviors of the Gartner’s hype cycle by introducing the technique of system dynamics. We suggest a decomposition method that first prepares basic components showing common behavior patterns, combines basic and intermediate components according to target system behavior, and finally construct a minimal model to fit it. The Hype Cycle minimal model that is established from this procedure helps to understand the behavior of the expectation and adoption with regard to emerging technologies. In addition, the minimal model can be the center for business managers or researchers to decide whether and when they adopt new technology. It can be expected that the Hype Cycle model made by decomposition method is applicable to other fields of research having similar behavior, which can be used as a new modeling pattern. This paper suggests other modeling patterns such as J-curve, N-shaped curve as well. Those modeling patterns become intermediate modeling components for decomposition method, and make it easy to model new system. Thus, we hope that decomposition method is generic enough to apply system dynamics modeling for various target system. Another research that this paper suggests is block diagram conversion. In general, there are several ways to model the same system by different diagrams. Stock and flow diagram, mostly used in system dynamics modeling, represents a set of differential equations. However, those equations can be modeled by block diagram in simulink, which provides us with the potential to find correspondence between. Pointing out this relation, this paper suggests a conversion rule that can change stock and flow diagram into block diagram with system behavior maintained.