eess.SYApr 7, 2026

Thermal-GEMs: Generalized Models for Building Thermal Dynamics

Authors: Felix Koch, Fabian Raisch, Benjamin Tischler

Organizations: Technical University of Applied Sciences Rosenheim, Rosenheim, Germany · Technical University of Munich, Munich, Germany

Read PDF
Ask AI
Cite

Copy citation

DOINot available for this paper

Abstract

Data-driven models for building thermal dynamics are a scalable approach for enabling energy-efficient operation through fault detection & diagnosis or advanced control. To obtain accurate models, measurement data from a target building spanning months to years are required. Transfer Learning (TL) mitigates this challenge by employing pretrained models based on single or multiple source buildings. General multi-source TL models promise to outperform single-source TL, but alternative multi-source modeling architectures remain to be explored, and evaluation on real-world data is missing. Moreover, time series foundation models (TSFM) have emerged as candidates for the best-performing general models. Hence, we conduct a first, comprehensive assessment of general modeling approaches for building thermal dynamics, including multi-source TL and TSFMs. Our assessment includes ablations using four state-of-the-art multi-source TL architectures and evaluations on synthetic as well as real-world data. We demonstrate that multi-source TL models are highly effective in accurately modeling buildings in real-world applications, yielding up to 63% lower forecasting errors compared to single-source TL. Moreover, our results suggest a trade-off between multi-source TL models exclusively pretrained with building data and TSFMs pretrained with a multitude of different time series, revealing that data from 16-32 source buildings must be available over 1 year for pretraining multi-source TL models to consistently outperform TSFMs as evaluated using the mean absolute error. These findings provide practical guidance for selecting modeling strategies based on the number of source buildings available for pretraining multi-source TL models.

Figures & tables

Figure 1. Overview of modeling strategies regarding our research questions.
ObservableDataset
NameTypeD1D2D3
Indoor HumidityFeature✓×\times×\times
Thermostat Heat SetpointFeature✓×\times✓
Thermostat Cool SetpointFeature✓×\times×\times
Supply Air Fan RuntimeFeature✓×\times×\times
Presence of OccupantsFeature✓×\times×\times
Table 1. Features included in the three datasets.
 EcoBeeIdealEuro. House
# Rooms Total932227450
# Sources Total892202400
# Targets402550
% Targets4.311.011.1
Table 2. Source and target split for all datasets
Figure 2. Seasonal Fine-Tuning split for the experiments
Figure 3. Zero-Shot MAE of all architectures as MSGMs
Figure 4. MAE of Single-Source TL versus MSGM for different fine-tunings on real-world data (Ecobee). MSGM beats single-source TL once errors are above the dotted line.
Figure 5. Zero-shot performance: TSFMs vs. the best-performing MSGMs (Transformer for Euro, LSTM for Ecobee and Ideal).
Figure 6. Zero-shot performance: Per-building TSFM compared to MSGM (each data point is a target building)
Figure 7. Zero-shot performance: Influence of the number of sources on the MSGM for buildings. The best-performing TSFM with respect to each dataset is used as a baseline.
Appendix figures & tables4 assets

Supplementary material from the paper’s appendix.

Appendix

HyperparameterLSTMTransformerMambaxLSTM
FromToStepFromToStepFromToStepFromToStep
Hidden Size6425632642563264256326425632
# Layers241241281241
Batch Size3212832321283232128323212832
Learning Rate.0001.001.0001.0001.001.0001.0001.001.0001.0001.001.0001
Dropout0.00.10.010.00.10.01---0.00.10.01
Appendix
Table 3. Hyperparameter Search Space for the Multi-Source General Models.
ParameterLSTMTransformerMambaxLSTM
EcobeeIdealEuroEcobeeIdealEuroEcobeeIdealEuroEcobeeIdealEuro
Hidden Size96128192192128192160128969612864
# Layers333433343333
Batch Size1281289612812812812812896128128128
Use AMPTrueTrueTrueTrueTrueTrueTrueTrueTrueFalseFalseFalse
Learning Rate.0009.001.001.0001.0005.0004.0006.0002.0002.0002.0005.0005
Appendix
Table 4. Hyperparameters for all MSGM architectures for each dataset.
Figure 8. RMSE of Single-Source TL versus MSGM for different fine-tunings on real-world data (Ecobee). MSGM beats single-source TL once errors are above the dotted line.
Figure 9. Zero-shot performance RMSE: Influence of the number of sources on the MSGM for buildings. The best-performing TSFM with respect to each dataset is used as a baseline.
More in this topic

Time Series Forecasting

Time series forecasting predicts future values from ordered observations using statistical, neural, and probabilistic models to support decisions under uncertainty.

See all 854 papers

Explore similar work

CardsList
Jan 23, 2025

GenTL: A General Transfer Learning Model for Building Thermal Dynamics

Transfer Learning (TL) is an emerging field in modeling building thermal dynamics. This method reduces the data required for a data-driven model of a target building by leveraging knowledge from a source building. Consequently, it enables the creation of data-efficient models that can be used for advanced control and fault detection & diagnosis. A major limitation of the TL approach is its inconsistent performance across different sources. Although accurate source-building selection for a target is crucial, it remains a persistent challenge. We present GenTL, a general transfer learning model for single-family houses in Central Europe. GenTL can be efficiently fine-tuned to a large variety of target buildings. It is pretrained on a Long Short-Term Memory (LSTM) network with data from 450 different buildings. The general transfer learning model eliminates the need for source-building selection by serving as a universal source for fine-tuning. Comparative analysis with conventional single-source to single-target TL demonstrates the efficacy and reliability of the general pretraining approach. Testing GenTL on 144 target buildings for fine-tuning reveals an average prediction error (RMSE) reduction of 42.1 % compared to fine-tuning single-source models.

Fabian Raisch, Thomas Krug, Christoph Goebel +1

Technical University of Applied Sciences Rosenheim Rosenheim, Germany · Technical University of Munich Munich, Germany

Fine-TuningTransfer Learning
AskPDF
May 2, 2026

Toward a foundational thermal model for residential buildings

The building energy community lacks a foundational thermal model, i.e., a single pretrained model capable of generalizing across diverse buildings, climates, and control strategies without building-specific calibration. Achieving this vision requires architectural principles that capture universal thermal dynamics rather than memorizing building-specific patterns. We take a step toward this goal by presenting a physics-informed transformer architecture that embeds domain knowledge, e.g., derivative enrichment and Euler-based numerical integration, into a decoder-only framework. We incorporate static building features extracted from simulation models and employ Rotary Position Embedding attention to capture temporal dependencies. Evaluated on the CityLearn dataset spanning 247 residential buildings across three climate zones, our model achieves one-step prediction accuracy (RMSE of 0.30°C in Texas, 0.29°C in Vermont) while outperforming both traditional baselines and fine-tuned Time-Series Foundation Models. We also demonstrate zero-shot transferability: models trained on as few as two buildings generalize to unseen buildings and climate zones without fine-tuning. Despite the limitation of simulated residential buildings, our results establish physics-informed architectural principles as a promising foundation for universal building thermal models.

Ting-Yu Dai, Kingsley Nweye, Dev Niyogi +1

The University of Texas at Austin · Eindhoven University of Technology

Time Series ForecastingPhysics-Informed ML
AskPDF
Sep 11, 2026

Very Exciting: Zero-Shot Model Predictive Control of Buildings via Excitation-Based Generalized Transfer Learning Models

The widespread adoption of data-driven, energy-efficient model predictive control (MPC) in buildings remains hindered by substantial effort to collect data and train models for individual buildings. Transfer learning (TL) has consequently gained increasing attention for target building modeling, as it reduces data requirements and modeling effort by reusing pretrained source models. However, these TL models are typically evaluated only on prediction accuracy in the target, without testing downstream control performance. To address this gap, we apply a state-of-the-art TL approach - pretraining a generalized model on multiple source buildings using standard operational data - within an MPC setup in a target building. We show that this approach is insufficient to achieve satisfactory control performance. As a solution, we introduce generalized models pretrained on excitation-based operational source data - purposefully probed inputs that explore the building's state-action space. For evaluation, we apply the generalized models via zero-shot (i.e., without fine-tuning) to 32 simulated target buildings and assess MPC performance. Our results show that excitation-based generalized models achieve the strongest control performance among all benchmarks, outperforming an online linear model-based MPC and a PI controller by 6.4% and 36.9%, respectively. By combining strong control performance with the ability to generalize across multiple buildings, without requiring any target-specific data, our approach reduces MPC setup cost and simplifies its widespread deployment in the building sector.

Fabian Raisch, Felix Koch, Zack Xuereb Conti +2

Technical University of Applied Sciences Rosenheim, Hochschulstraße 1, Rosenheim, 83024, Germany · Technical University of Munich, Arcisstraße 21, Munich, 80333, Germany · Karlsruhe Institute of Technology, Hermann-von-Helmholtz-Platz 1, Eggenstein-Leopoldshafen, 76344, Germany +1

Building Energy ManagementModel Predictive ControlTransfer Learning
AskPDF