Semantic design space refinement for model-based systems engineering

Abstract

This paper describes a process developed to utilize information within an ontology, a form of a component model library, for design space refinement in conceptual design through the use of an Interactive Reconfigurable Matrix of Alternatives (IRMA). An ontological approach is proposed, as it enhances the capabilities of component model libraries that are already being constructed for use in model-based systems engineering (MBSE) programs. By defining a common vocabulary across multiple design domains, an ontology is capable of driving information consistency and reducing errors that occur due to misinterpretation of common vernacular. Within an ontology, a functional decomposition of any complex system can be represented as a hierarchy of "classes", where a class defines any concept within a given domain. The functions that each system must perform, along with the system components that achieve each function and the design alternatives for each component, are related to the system as properties of the system. This functional decomposition structure of the ontology may be identical to the decomposition that is used in an IRMA, which can use information exported from the ontology to perform design space refinement during conceptual design. The result of this study is a process which is capable of extracting information from an ontology, organizing it, and importing it directly to an IRMA which is fully functional and ready to be used. Such a process becomes a powerful capability when similar systems are designed, as the process can be reused, serving as a starting point for concept selection in conceptual design, resulting in reductions in both development time and cost.