A formal framework for capturing knowledge to transform structural models into analysis models

Abstract

During the systems design process, there are a multitude of analyses and computer simulations that are performed to evaluate a particular design or architecture. This paper focuses on automating this process by defining a formal framework for capturing and applying the knowledge needed to automatically generate system-level analysis models from system-level descriptive models. The framework builds on the similarities that exist between analytical and descriptive models when considered from a systems perspective, namely, as consisting of sub-systems or components and the interactions between them. The relationships between analytical and descriptive models are captured at the component level in multi-aspect component models (MAsCoMs). The information in MAsCoMs is represented formally in the Object Management Group's Systems Modeling Language and can then be applied automatically through the use of generic model transformations. The transformations apply to all models in a certain domain, such as dynamic simulation modelling. In this paper, the approach is demonstrated for a hydraulic system by generating a system-level dynamic simulation from a descriptive model of the hydraulic circuit.

Keywords:

• systems engineering
• SysML
• model composition
• model to model transformation