Hybrid AI-physical Modelica Modelling using SMArtiInt and ClaRa

Use Case für SMArtiInt und ClaRa

Hybrid AI-physical Modelica Modelling - A power plant simulation use case for SMArtiInt and ClaRa Lib .

You have a lot of system-data, but you need furthermore a system model for the next development steps? XRG provides methods and tools for a proper hybrid AI-physical model generation. You can start right away by downloading the Modelica libraries SMArtIInt and ClaRa. If requested, we can provide our services as well to significantly reduce the effort for digital twin generation.  

XRG established a first AI assisted generation and maintenance of digital twin power plant system model within DIZPROVI, a joint research project with the university Zittau/Goerlitz, CombTec GmbH and LEAG funded by the federal Ministry of Education and Research, Germany. 

Using the Modelica libraries SMArtIInt and the ClaRa we demonstrate the integration of an operation data-based AI model for heat flow into the physical based plant boiler system model. This shall pave the way to AI-Modelica hybrid models, which combine insights from available data with the predictive power and reliability of physical laws.