A Declarative Approach for Transforming SysML Models to Executable Simulation Models

Systems Modeling Language (SysML) is an object management group standard for systems-of-systems engineering. It enables the description of complex system models; however, it cannot effectively support all system engineering activities. For instance, system performance evaluation is usually performed via simulation. In this case, the transformation of SysML system models to executable simulation models for specific simulation methodologies and tools is required. Model transformation is a key component for addressing the challenges of seamless integration of SysML model simulation in model-based system engineering. In this paper, we explore a declarative approach, based on the query/view/transformation-relations (QVT-R) standard, for the transformation of SysML models to executable simulation models, fully adhering model-driven architecture (MDA) concepts. It is supported by a framework implemented to provide executable simulation model and code generation. Methodological guidelines, for the effective use of a declarative language as QVT-R, for model transformation, are provided, emphasizing the utilization of existing domain-specific SysML profiles, as well as executable simulation library components. The experience obtained from two different domains, namely, enterprise information and railway transportation systems, modeled as systems-of-systems via SysML, is discussed, based on a quantitative analysis of the respective QVT-R transformations.

Automated summary

SysML is a standard for systems-of-systems engineering that enables the description of complex system models but may not support all system engineering activities. This paper explores a declarative approach based on the QVT-R standard for transforming SysML models into executable simulation models. Methodological guidelines for effective use of a declarative language as QVT-R are provided, emphasizing the utilization of existing domain-specific SysML profiles and executable simulation library components.