Master's Thesis Yingying Yang


Generation of Modelica Simulation Models from Ontology-based Building System Models

Brick-Schema Copyright: EBC Brick-Schema Model of a Heat Pump

Building Energy Performance Simulation (BEPS) is a key element in the operation of high-performance buildings. However, using BEPS tools such as Modelica language to generate building simulation models are time-consuming and error-prone processes. Reusing the existing data point labels of building management systems (BMS) as basis for BEPS has the potential to make BEPS model development easier, faster and more reliable. In addition, it provides the possibility to integrate the time series data of BMS into the subsequent simulation. Therefore, this master thesis proposes a Pythonbased model transformation tool-chain to semi-automatically convert ontology-based building system models using Brick Schema (BrickModel) into Modelica simulation models. Brick-Model is generated based on the Information of data point labels and can describe the topology of heating, ventilation, and air conditioning (HVAC) system. Furthermore, Brick-Model can be easily extended to store the additional data required by Modelica. The Information stored in BrickModel can be queried and extracted with SPARQL Protocol And RDF Query Language (SPARQL). Through a template-based method, Modelica code can be generated. The tool-chain is successfully tested and applied to modelling selected parts of HVAC system of the main building of E.ON Energy Research Center, in which the data points of BMS are named by buildings unified data point naming schema for operation management(BUDO).