ModSim@SusTech 2026

At the conference, he presented two contributions from ongoing research at the Modeling and Simulation Lab. The first paper, “Quantification and Surrogate Model Estimation of Spatial-Temporal Carbon Intensity Factors” by D. Bayer, J. Schiller, T. M. Pham and M. Pruckner, addresses the modeling of spatially and temporally resolved carbon intensity factors. The work focuses on how such factors can be quantified and approximated using surrogate models, enabling their integration into simulation and optimization frameworks for future energy systems.

The second paper, “Benchmarking Transformer and xLSTM for Time-Series Forecasting of Heat Consumption” by M. Wahl, D. Bayer, S. Rausch, and M. Pruckner, investigates modern deep learning architectures for heat consumption forecasting. In particular, the study compares transformer-based models with xLSTM architectures for time-series forecasting tasks in the energy domain. The work was carried out in collaboration with RAUSCH Technology GmbH, a project partner of the ModSim Lab, represented by M. Wahl and S. Rausch.

The conference provided valuable opportunities to discuss current challenges and future research directions in sustainable energy systems, including data-driven forecasting, digital twin architectures, and carbon-aware modeling approaches.