Adaptive Thermal Building Models And Methods for Scalable Simulations of Multiple Buildings using Modelica

For building performance simulations of multiple buildings, varying objectives and data availability lead to different requirements for various model applications. Flexibility regarding spatial discretization, parameterization and process automation can offer an alternative to specifically tailored models for each application. We present adaptive low order thermal network models with variable discretization regarding number of zones, number of wall elements and wall discretization. These Modelica simulation models come with parameterization processes in Python that allow data enrichment using statistical data. Process automation takes care of model generation, parallel simulation and results analysis. All models and processes are open source and freely available.