The Potential of FMI for the Development of Digital Twins for Large Modular Multi-Domain Systems

Abstract

Digital twins enable the observation, prediction, and optimization of a physical system and thus allow to realize their full potential. However, their functionality is mainly based on simulation models of the entire system behavior. For modular multi-domain systems, this requires the extensive use of dynamically composed models that are made up of individual component models. The FMI-Standard forms a solid foundation for this problem and is very well known in the automotive engineering fields. However, composed system models using FMI are not widely adapted in renewable energy and wind energy yet. So far, the coupling of simulation models is limited. This paper discusses the strategy of building digital twins from individual FMUs with predefined model interfaces based on an ontology for renewable energy systems. An accelerated development is enabled by the exchange of sub-models in the digital twin without adjustments of interface. An example for the proposed process is given by the composed simulation model of a hydrogen generation process based on wind energy.

Dynamic system models and their simulation in the Semantic Web

“…This is enabled by annotating the component models with model-specific and domain-specific information and then using a reasoner to infer connections between models. Wiens et al present similar work for creating digital twins of wind turbines as an example of large, modular multi-domain systems [48]. Both base their implementations on FMI as the format for the component models and the System Structure and Parameterization (SSP) standard [26] for the specification of the topology, in other words the connections between the FMUs.…”