| Popis: |
Digital twins are promising several benefits, including better monitoring and control, improved performance, and lower maintenance costs of their physical counterparts. However, there are still some challenges associated with the development and management of digital twins that need to be addressed before the actual realisation of digital twins’ applications. In this study, we look at the data management issue of digital twins that are based on system simulation.The problem with most of the commercial modelling and simulation software is the use of proprietary data models and specific formats to store these models, resulting in poor interoperability. Since digital twins of certain assets need to be operated for a couple of decades, there is a possibility that the underlying application software may not be available through the assets’ life cycles. In order to avoid unknown risks or business disruption, it is advisable to preserve information about the simulation models of digital twins in long-lasting formats. This is to ensure that in situations where the original software is inappropriate/inaccessible for opening/running a digital twin’ simulation model, one should be able to revive the model, using the preserved information, and adopt alternative tools.This work illustrates the long-term data preservation of digital twins by exploring two Modelica system simulation models. Modelica is an open-source language for modelling complex systems and provides detailed information about the models of components that can be preserved in user-specific formats. The models were built in OMEdit environment, a free tool for Modelica implementation. Data models of the corresponding simulation models were developed in Web Ontology Language (OWL), which is used for standardised representation of information on the Semantic Web. Protégé ontology editor was used to design four ontologies that provide the necessary concepts and relationships for describing the simulation models. Using these concepts and relationships, the OWL data models were generated. It was found that OWL lacks certain features and the visualisation of data models gets complex as the model size grows. |
