Integrating Computational Design Support in Model-Based Systems Engineering Using Model Transformations

Abstract

Model-based systems engineering (MBSE) combines the rigor of systems engineering with formal models to support communication in multidisciplinary engineering. With industrial adoption of MBSE, the maturity of modeling environments supporting MBSE increased. Still, efficient means to integrate computational design methods in MBSE are missing. Here, we present a method that enables systems engineers to directly integrate computational methods for solving design tasks. The method relies on established semantics of the systems modeling language (SysML) and therefore can be directly integrated with existing system models so to avoid redundant knowledge formalizations for computational methods. Next, model transformations are applied to generate the mathematical model based on the relevant parts of the system model. These temporary models are used to solve the design task and generate output that is fed back to the system model. Therefore, the proposed method contributes by relying on a single and comprehensible knowledge formalization understandable to engineers. Further, it enables systems engineers to formalize design tasks for automated reasoning themselves by bundling the complexity of the mathematical modeling within the model transformations. An industrial case for designing sealing elements for piping is used to illustrate the potential of the proposed approach. Future work needs to further elaborate on automated selection of appropriate mathematical methods as well as computational support for the identification of opportunities for integration of computational design methods readily while developing a system model.