05 Zeitschriften, fortlaufende Sammelwerke
Refine
Document Type
- Article (peer reviewed) (6)
- Preprint (2)
- conferencepaper (1)
Language
- English (9)
Has Fulltext
- no (9)
Is part of the Bibliography
- no (9)
Keywords
- Transfer Learning (3)
- Building Thermal Dynamics (2)
- Machine Learning (2)
- Simulation (2)
- Building Dynamics (1)
- Building Similarity (1)
- Building Simulation (1)
- Change in Occupancy (1)
- Concept Drift (1)
- Continual Learning (1)
Parameter estimation for dynamical systems remains challenging due to non-convexity and sensitivity to initial parameter guesses. Recent deep learning approaches enable accurate and fast parameter estimation but do not exploit transferable knowledge across systems. To address this, we introduce a transfer-learning-based neural parameter estimation framework based on a pretraining-fine-tuning paradigm. This approach improves accuracy and eliminates the need for an initial parameter guess. We apply this framework to building RC thermal models, evaluating it against a Genetic Algorithm and a from-scratch neural baseline across eight simulated buildings, one real-world building, two RC model configurations, and four training data lengths. Results demonstrate an 18.6-24.0% performance improvement with only 12 days of training data and up to 49.4% with 72 days. Beyond buildings, the proposed method represents a new paradigm for parameter estimation in dynamical systems.
Recent advancements in foundation models for tabular data, such as TabPFN, demonstrated that pretrained Transformer architectures can approximate Bayesian inference with high predictive performance. However, Transformers suffer from quadratic complexity with respect to sequence length, motivating the exploration of more efficient sequence models. In this work, we investigate the potential of using Hydra, a bidirectional linear-time structured state space model (SSM), as an alternative to Transformers in TabPFN. A key challenge lies in SSM’s inherent sensitivity to the order of input tokens – an undesirable property for tabular datasets where the row order is semantically meaningless. We investigate to what extent a bidirectional approach can preserve efficiency and enable symmetric context aggregation. Our experiments show that this approach reduces the order-dependence, achieving predictive performance competitive to the original TabPFN model.
Data-driven modeling of building thermal dynamics is emerging as an increasingly important field of research for largescale intelligent building control. However, research in data-driven modeling using machine learning (ML) techniques requires massive amounts of thermal building data, which is not easily available. Neither empirical public datasets nor existing data generators meet the needs of ML research in terms of data quality and quantity. Moreover, existing data generation approaches typically require expert knowledge in building simulation. To fill this gap, we present a thermal building data generation framework which we call BuilDa. BuilDa is designed to produce synthetic data of adequate quality and quantity for ML research. The framework does not require profound building simulation knowledge
to generate large volumes of data. BuilDa uses a single-zone Modelica model that is exported as a Functional Mockup Unit (FMU) and simulated in Python. We demonstrate BuilDa by generating data and utilizing it for a transfer learning study involving the fine-tuning of 486 data-driven models.
Transfer Learning (TL) is currently the most effective approach for modeling building thermal dynamics when only limited data are available. TL uses a pretrained model that is fine-tuned to a specific target building. However, it remains unclear how to proceed after initial fine-tuning, as more operational measurement data are collected over time. This challenge becomes even more complex when the dynamics of the building change, for example, after a retrofit or a change in occupancy. In Machine Learning literature, Continual Learning (CL) methods are used to update models of changing systems. TL approaches can also address this challenge by reusing the pretrained model at each update step and fine-tuning it with new measurement data. A comprehensive study on how to incorporate new measurement data over time to improve prediction accuracy and address the challenges of concept drifts (changes in dynamics) for building thermal dynamics is still missing. Therefore, this study compares several CL and TL strategies, as well as a model trained from scratch, for thermal dynamics modeling during building operation. The methods are evaluated using 5–7 years of simulated data representative of single-family houses in Central Europe, including scenarios with concept drifts from retrofits and changes in occupancy. We propose Seasonal Memory Learning (SML), a CL strategy that provides greater accuracy improvements than existing CL and TL methods, while maintaining low computational effort. SML outperformed the benchmark of initial fine-tuning by 42.5% without concept drifts and 48.3% with concept drifts.
Investments in research and development (R&D) are crucial for innovation and economic growth, yet dis-tinctions between high-tech and mid-tech industries may lead tovarying financial outcomes. Building onFuest et al.(2024), this study examines the financial performance of publicly listed high-and mid-techcompanies from a shareholder'sperspective. Using data from theEU Industrial R&D Investment Scoreboard and LSEGWorkspace for 2,349 listed companies, market-capitalization-weighted portfolios were constructed and benchmarked against global and local market indices. Performance was evaluated using theSharpe ratio and multi-factor asset pricing models (Carhart four-factor and Fama-French five-factor), distinguishing between crisis and non-crisis periods, as well as regional and sectoral variations. Results showthat, at the aggregate level,high-tech portfolios consistentlyand significantly outperform mid-tech and "other" categories on absolute and risk-adjusted bases. Mid-tech indus-tries exhibitresilience during crisis but lower alphasthan high-techindustries,and sectoral heterogeneity high-lightsdiversification benefits. Furthermore, theanalysis reveals investment opportunities in R&D-intensive sec-tors, emphasizing crisis resilience and regional competitiveness.
Neural network synthetic dataset generation for fault detection in district heating substations
(2025)
District heating systems (DHS) play a vital role in sustainable heating solutions and the decarbonization of the energy sector. However, inefficiencies due to undetected faults in substations result in high return temperatures, increasing heat losses, and limiting the integration of renewable energy sources. The lack of publicly available labeled datasets poses a significant challenge for fault detection using supervised learning models. To address this issue, this study explores three machine learning-based synthetic data generation techniques – time series forecasting, generative adversarial networks (GANs), and fault signature transfer.
These methods aim to increase publicly available data either by sharing the generating model or a synthetic dataset. The novelty lies in the combination of advanced supervised machine learning methods being applied to a large, fully labeled data set to create new, equally labeled data for publication, as, to our knowledge, no such dataset has been compiled before. We evaluate our methods on the first-of-its-kind ILSE dataset, which includes real-world smart meter data from 547 substations and 1,162 reviewed faults from a German DHS network, including detailed root cause information. Overall, time series forecasting achieves an MAPE of 3% to 10% for inlet and outlet temperature and 25% to 40% for heat load and flow rate, both of which are within year-to-year variance. For GANs, specifically TimeGAN, we found a discriminative score of about 0.10 compared to 0.24 in the original publication when tested on Energy benchmark data. Fault signature transfer has yet to yield usable results, most likely due to the high variance in the fault signatures, fault duration, and overlapping or multiple root causes. Finally, fault data in the synthetic data is not yet good enough for practical use, e.g. training a fault detector.
Exploring stakeholder perspectives: Enhancing robot acceptance for sustainable healthcare solutions
(2023)
The pandemic has highlighted the fact that healthcare systems around the world are under pressure. Demographic change is leading to an increasing shortage of care workers in most countries, and the demographic challenge is only just beginning in most societies. While robots are widely used in industry, robotic support in healthcare is still limited to very specialized robots in the operating theatre. The question of what type of deployment is likely to be successful in a healthcare scenario is not only a technological or economical question but also one of technology acceptance. The answer to this question supports entrepreneurial opportunities to develop sustainable healthcare solutions.
In this paper, we analyze the acceptance of robots in elderly care from the perspective of patients, patient families, and geriatric care professionals. To understand the various positions and to identify the suitability of existing acceptance models, we applied stakeholder mapping to conduct qualitative interviews with 14 people with different knowledge backgrounds and levels of involvement in care situations, based on 9 videos showing different robots and application scenarios.
The results confirmed that existing technology acceptance models need to be extended by factors such as robot appearance. We found that the background knowledge of the respondents influences the results of the questions about e.g. safety concerns. In addition, we found that the contribution to patients' self-determination and independence is an important factor that is not included in existing technology acceptance models. Finally, the discovery of a significant discrepancy between the self-perception and the external perception of the different stakeholders regarding the acceptance of a service robot can be explained by the stakeholder positions involved in caring for the benefit of a specific patient.
These findings encourage further research, especially with the underlying assumption that technology acceptance in healthcare is not just a patient issue, but a stakeholder issue. Stakeholder mapping is a valid tool to analyze the interdependencies for the acceptance of robots. Therefore, we suggest using a tool such as stakeholder mapping to further analyze these issues.
Fault and anomaly detection in district heating substations: A survey on methodology and data sets
(2023)
District heating systems are essential building blocks for affordable, low-carbon heat supply. Early detection and elimination of faults is crucial for the efficiency of these systems and necessary to achieve the low temperatures targeted for 4th generation district heating systems. Especially methods for fault and anomaly detection in district heating substations are currently of high interest, as faults in substations can be repaired quickly and inexpensively, and smart meter data are becoming widely available. In this paper, we review recent scientific publications presenting data-driven approaches for fault and anomaly detection in district heating substations with a focus on methods and data sets. Our review indicates that researchers use a wide variety of methods, mostly focusing on unsupervised anomaly detection rather than fault detection. This is due to a lack of labeled data sets, preventing the use of supervised learning methods and quantitative analysis. Together with the lack of publicly available data sets, this impedes the accurate comparison of individual methods. To overcome this impediment, increase the comparability of different methods and foster competition, future research should focus on establishing publicly available data sets, and industry-relevant metrics as benchmarks.
Real-world domestic electricity demand datasets are the key enabler for developing and evaluating machine learning algorithms that facilitate the analysis of demand attribution and usage behavior. Breaking down the electricity demand of domestic households is seen as the key technology for intelligent smart-grid management systems that seek an equilibrium of electricity supply and demand. For the purpose of comparable research, we publish DEDDIAG, a domestic electricity demand dataset of individual appliances in Germany. The dataset contains recordings of 15 homes over a period of up to 3.5 years, wherein total 50 appliances have been recorded at a frequency of 1 Hz. Recorded appliances are of significance for load-shifting purposes such as dishwashers, washing machines and refrigerators. One home also includes three-phase mains readings that can be used for disaggregation tasks. Additionally, DEDDIAG contains manual ground truth event annotations for 14 appliances, that provide precise start and stop timestamps. Such annotations have not been published for any long-term electricity dataset we are aware of.
