A Data-Driven Framework for Transformer Health Monitoring in Distribution Networks
Bibliographic record
Abstract
Distribution networks typically contain a large number of distribution transformers that are critical for high-quality uninterrupted energy delivery. Distribution transformers are the last portion in a power delivery system, their health is essential to ensure the reliability of distribution networks, and their failure may cause huge technical and economic problems to power utilities and customers. The distribution transformer failure may be influenced by its maintenance history, risk index in a distribution network associated with keraunic level, average number of lightning strikes, protection devices employed, dissolved gas analysis (DGA) data, overloading, and phase unbalance etc. To prevent sudden failures of transformers while meeting growing electricity demands, transformer health monitoring and failure prediction become increasingly important in operation and maintenance procedures for power utilities. To enhance reliable operation of transformers in distribution networks, an effective data-driven framework for the distribution transformer heath monitoring is proposed in this thesis, which is computationally efficient, accurate, reliable, and comprehensive to monitor the health of distribution transformers based on readily available lightning strike data, distribution networks risk index, DGA data, historical maintenance data, advanced metering infrastructure (AMI) data, roof-top photovoltaics (PV) production data, and electric vehicle (EV) charging stations data. To develop the proposed framework, first, a comprehensive literature review is conducted on data-driven approaches for distribution transformer health monitoring to understand research gaps; second, a novel distribution transformer failure prediction method through a hybrid one-class deep support vector data description (SVDD) is proposed using lightning strike data and distribution networks risk index data, and this method helps distribution transformer predictive maitenance; third, a novel data-driven hybrid approach is proposed to predict the health index (HI) of a transformer using DGA data, where a robust technique is used to handle missing or erroneous data due to sensor or data transfer issues, and five case studies with different percentages of missing data are used to validate the proposed approach; fourth, the combined impact of plug-in electric vehicles (PEV) charging load and roof-top PV is investigated through a case study in Saskatoon, Canada, the results show that the increased PEV adoption exacerbates the transformer stress, while PV integration mitigates these effects, especially during summer, and this highlights the need to encourage rooftop PV adoption to balance the PEV charging demand on transformers; finally, a novel hierarchical load forecast aggregation technique in distribution transformers using AMI data recorded by Saskatoon Light and Power in Canada is proposed to accurately predict each customer’s load, each transformer phase load and the total transformer load, which helps to determine the transformer overloading and load unbalance.
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 machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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 source (direct Gemma or distilled Codex), 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".