Unsupervised automated fault detection and diagnosis for light commercial buildings’ HVAC systems
Bibliographic record
Abstract
• The proposed unsupervised method enhances traditional PCA fault detection and diagnosis, particularly when there is no information about the unfaulty condition of the HVAC system (which is very common in existing buildings). • PCA, with some modifications, can be applied to detect and diagnose various faults beyond sensor faults, especially when there are changes in the dynamic behavior of the HVAC system. • The proposed unsupervised method demonstrates transferability and scalability across different HVAC configurations in light commercial buildings, validated on two completely different configurations. Fault detection in light commercial building HVAC systems can significantly improve the energy efficiency of this class of buildings. A light commercial building is a commercial structure with fewer than six stories and a floor plan area of less than 2500 ft². Data extracted from existing buildings in this class are generally unlabeled, raw, and characterized by many inconsistencies and discontinuities, making Automated Fault Detection and Diagnosis (AFDD) particularly challenging. This study aims to develop an unsupervised AFDD method tailored for light commercial buildings, which is transferable among different HVAC configurations within this building class. The method is designed to handle unlabeled, incomplete, and raw datasets provided by their Building Energy Management Systems (BEMS). Principal Component Analysis (PCA) was selected as the core method due to its scalability and transferability. Specific techniques were introduced to address time series analysis and fault detection and diagnosis (FDD) based on the dynamics of the system, using appropriate window sizing. The method was validated using two different light commercial buildings with distinct configurations and data availability. The primary building, an office in Montreal, Canada, and the secondary building, a small industrial facility in Ireland, served as the test cases. The proposed method demonstrated promising results in detecting and isolating faulty inputs, providing information on the severity levels and locations of faults. It successfully identified whether faults were at the level of the central system or within specific zones in both studied cases.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".