Fault Detection in Solar Thermal Systems using Probabilistic Reconstructions

2025-11-13

AI Summary

Fault detection in solar thermal systems (STS) remains challenging due to installation, maintenance, or operational errors—especially under limited labeled data and across heterogeneous systems. Method: This paper proposes a lightweight, probability-based anomaly detection framework leveraging existing sensor time-series data. It integrates deep learning-based reconstruction with heteroscedastic uncertainty modeling to enable low-cost, fully automated, human-in-the-loop-free monitoring. Contribution/Results: By jointly exploiting probabilistic reconstruction errors and calibrated uncertainty estimates for anomaly discrimination, the method significantly improves generalization in few-shot and cross-system scenarios. Evaluated on the real-world PaSTS dataset, it achieves high-accuracy qualitative and quantitative fault detection across multiple typical failure modes—including leakage, pump failure, and sensor drift—outperforming both conventional and state-of-the-art deep learning baselines. The framework demonstrates strong robustness, computational efficiency, and practical deployability in industrial STS monitoring.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Data Mining & Knowledge Management: Anomaly/Outlier Detection
• Intelligent Robots: State Estimation

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Energy management for devices in mobile Web and WoT environments
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

Solar thermal systems (STS) present a promising avenue for low-carbon heat generation, with a well-running system providing heat at minimal cost and carbon emissions. However, STS can exhibit faults due to improper installation, maintenance, or operation, often resulting in a substantial reduction in efficiency or even damage to the system. As monitoring at the individual level is economically prohibitive for small-scale systems, automated monitoring and fault detection should be used to address such issues. Recent advances in data-driven anomaly detection, particularly in time series analysis, offer a cost-effective solution by leveraging existing sensors to identify abnormal system states. Here, we propose a probabilistic reconstruction-based framework for anomaly detection. We evaluate our method on the publicly available PaSTS dataset of operational domestic STS, which features real-world complexities and diverse fault types. Our experiments show that reconstruction-based methods can detect faults in domestic STS both qualitatively and quantitatively, while generalizing to previously unseen systems. We also demonstrate that our model outperforms both simple and more complex deep learning baselines. Additionally, we show that heteroscedastic uncertainty estimation is essential to fault detection performance. Finally, we discuss the engineering overhead required to unlock these improvements and make a case for simple deep learning models.

Problem

• Detecting faults in solar thermal systems automatically
• Identifying abnormal system states using probabilistic reconstructions
• Addressing efficiency loss from installation and operational faults

Innovation

• Probabilistic reconstruction framework for anomaly detection
• Heteroscedastic uncertainty estimation improves fault detection
• Simple deep learning models generalize across solar systems