CityDPC: A Python library for handling 3D city model datasets
This study presents CityDPC, a Python library for geometric computations on CityGML and CityJSON datasets.
View PublicationAll Publications
Identifying Reference Districts by means of Machine Learning and Open-Source Data
This study addresses neighborhood-level sustainability in urban development as a critical lever in Germany’s energy transition goals.
View Publication
Low-order building model for district energy simulations in Matlab
The paper presents a low-order building model for district energy simulations in Matlab, based on the VDI 6007 standard and implemented using SimScape, with validation against ASHRAE Standard 140 showing good accuracy for heat transfer dynamics. It concludes that SimScape is the most flexible and suitable toolbox for this purpose, despite minor shortcomings in thermal storage representation.
View Publication
AN OPEN-SOURCE CITYGML ENRICHMENT TOOL (CITYENRICH)
The paper presents CityEnrich, an open-source tool that allows users to manually enrich CityGML LoD2 building models with energy-specific data in compliance with the Energy ADE standard, improving the accuracy of Urban Building Energy Modelling (UBEM) simulations compared to archetype-based approaches.
View Publication
Automatic Model Selection and the Use of Representative Days in Approximate Model Predictive Control Applications for Heat Pump Systems
This paper presents a framework for approximated model predictive control (AMPC) using machine learning, focusing on automatic algorithm selection and the use of representative days for training, with application to a heat pump system. The results show that Gradient Boosting performs best and that using just ten clustered days for training can achieve prediction accuracies close to full-year models, especially when considering imbalanced data handling.
View Publication
Urban Energy Simulations using Open CityGML Models: A Comparative Analysis
This paper compares simulated and measured heating demands for 500 buildings using open-source CityGML data enriched with building archetypes, revealing that the lack of critical attributes like construction year, usage type, and refurbishment state causes large simulation discrepancies. It concludes that accurate energy modeling at urban scale depends heavily on including such attributes in open datasets to enable reliable archetype assignment and simulation validation.
View Publication