Robust Non-Intrusive Load Monitoring (NILM) with unknown loads
Bibliographic record
Abstract
A Non-Intrusive Load Monitoring (NILM) method, robust even in the presence of unlearned or unknown appliances (UUAs) is presented in this paper. In the absence of such UUAs, this NILM algorithm is capable of accurately identifying each of the turned-ON appliances as well as their energy levels. However, when there is an UUA or set of UUAs are turned-ON during a particular time window, proposed NILM method detects their presence. This enables the operator to detect presence of anomalies or unlearned appliances in a household. This quality increases the reliability of the NILM strategy and makes it more robust compared to existing NILM methods. The proposed Robust NILM strategy (RNILM) works accurately with a single active power measurement taken at a low sampling rate as low as one sample per second. Here first, a unique set of features for each appliance was extracted through decomposing their active power signal traces into uncorrelated subspace components (SCs) via a high-resolution implementation of the Karhunen-Loeve (KLE). Next, in the appliance identification stage, through considering power levels of the SCs, the number of possible appliance combinations were rapidly reduced. Finally, through a Maximum a Posteriori (MAP) estimation, the turned-ON appliance combination and/or the presence of UUA was determined. The proposed RNILM method was validated using real data from two public databases: Reference Energy Disaggregation Dataset (REDD) and Tracebase. The presented results demonstrate the capability of the proposed RNILM method to identify, the turned-ON appliance combinations, their energy level disaggregation as well as the presence of UUAs accurately in real households.
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".