Summary of Conformalized Prediction Of Post-fault Voltage Trajectories Using Pre-trained and Finetuned Attention-driven Neural Operators, by Amirhossein Mollaali et al.
Conformalized Prediction of Post-Fault Voltage Trajectories Using Pre-trained and Finetuned Attention-Driven Neural Operators
by Amirhossein Mollaali, Gabriel Zufferey, Gonzalo Constante-Flores, Christian Moya, Can Li, Guang Lin, Meng Yue
First submitted to arxiv on: 31 Oct 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper presents a novel methodology for predicting voltage trajectory intervals in power systems, leveraging the Quantile Attention-Fourier Deep Operator Network (QAF-DeepONet). The QAF-DeepONet is designed to capture complex dynamics and estimate quantiles of target trajectories without distributional assumptions. The proposed operator regression model maps observed voltage data to unobserved post-fault trajectories. A pre-training and fine-tuning process addresses limited data availability, using federated learning with neighboring buses for privacy. After pre-training, the model is fine-tuned with target bus data, adapting to unique dynamics and operating conditions. Conformal prediction integrates into the fine-tuned model for coverage guarantees. Performance evaluation on the New England 39-bus test system uses Prediction Interval Coverage Probability (PICP) and Prediction Interval Normalized Average Width (PINAW). The results demonstrate the proposed approach’s practical and reliable uncertainty quantification in predicting voltage trajectory intervals. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper helps us better predict what happens to power systems after a fault occurs. They created a new kind of network called QAF-DeepONet that can capture complex patterns in voltage data without making any assumptions about how the data is distributed. The network uses a combination of old and new data to make predictions, which it does by mapping observed data to unobserved future data. To keep the model private, they used a special technique called federated learning with neighboring power systems. The results show that their approach can accurately predict voltage trajectory intervals and provide reliable uncertainty estimates. |
Keywords
» Artificial intelligence » Attention » Federated learning » Fine tuning » Probability » Regression
