Online Power-Transformer Hot-Spot Estimation via Spectrum-Aware Physics-Data Hybrid Learning
Abstract
Abstract The growing share of converter-interfaced wind and photovoltaic generation exposes step-up transformers to rapid load changes and harmonic-rich currents. Conventional thermal models preserve physical interpretability but leave operating-condition-dependent residuals, whereas purely data-driven predictors may lose reliability outside the training regime. This study develops an explainable physics-data hybrid estimator for online power-transformer hot-spot temperature and overtemperature risk. A spectrum-aware thermal recursion converts the measured harmonic-current spectrum into frequency-weighted losses and propagates top-oil and winding states. A conditional C-vine copula then learns residual dependence on load factor, harmonic loss factor, ambient temperature, load-rate change, and the previous reliable residual, producing a median estimate, a 90% prediction interval, and an overtemperature probability. In a 30-day chronological simulation, the method achieved an MAE of 0.636°C, an RMSE of 0.797°C, a coefficient of determination of 0.9928, and 89.3% coverage for the nominal 90% interval. In a 15-day field campaign, the T2 hot-spot MAE and RMSE were 1.87°C and 2.34°C. Recall at the 98°C warning threshold was 12/13 (92.3%; 95% Wilson interval, 66.7%-98.6%). This method can effectively reduce temperature estimation errors under high harmonic loads, load ramping, and cooling state switching conditions.
Read the paper
Related studies
- Preliminary Toxicity Analysis of 3DCRT versus IMRT on the High Dose Arm of the RTOG 0126 Prostate Cancer Trial
- Adaptive Radiotherapy for Head and Neck Cancer: Initial Clinical Outcomes from a Prospective Trial
- American Association of Physicists in Medicine Task Group 263: Standardizing Nomenclatures in Radiation Oncology
- Motion Management Strategies and Technical Issues Associated with Stereotactic Body Radiotherapy of Thoracic and Upper Abdominal Tumors: A Review from NRG Oncology
- High-dose-rate interstitial brachytherapy as a monotherapy for localized prostate cancer: treatment description and preliminary results of a phase I/II clinical trial
- Highly efficient training, refinement, and validation of a knowledge-based plan quality control system for radiotherapy clinical trials
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.