False Data Injection Attack Detection and Localization Framework in Power Distribution Systems Using a Novel Ensemble of CNNs and Explainable Artificial Intelligence
Bibliographic record
Abstract
Cyber-physical power systems are vulnerable to cyber-attacks, especially false data injection attacks (FDIAs). FDIAs against distribution system state estimation (DSSE), which alter state estimation (SE) by changing meter readings, have received researchers’ attention in recent years. A common defense against FDIAs in the literature is the use of labeled data to train classifiers as FDIA detectors. However, this approach's performance can be limited by the highly imbalanced nature of FDIA datasets. The black box characteristics of the machine learning models can make them hard to trust and adopt in important applications. Hence, we propose an innovative explainable artificial intelligence (XAI)-enhanced ensemble-based detection and localization model that leverages convolutional neural networks (CNNs) and support vector machine (SVM). The ensemble model uses SVM to merge various spatiotemporal CNNs’ outputs. Training these CNNs on under-sampled subsets of the majority class and using their ensemble addresses class imbalance. This paper leverages XAI to enhance the interpretability of the detection process and improve localization accuracy. The localization process uses the outputs of an XAI technique, gradient-weighted class activation mapping, to aid the majority-voting-based localization ensemble model. Our model can also detect FDIAs during the distribution feeder's topology changes. Extensive simulations on IEEE 13-bus and IEEE 123-bus feeders prove the proposed under-sampling-based detection approach as an alternative to prevalent over-sampling methods like generative adversarial networks (GANs), offering a novel solution to class imbalance challenges. The paper also provides a comprehensive analysis of the proposed spatiotemporal model's performance, demonstrating its superiority over temporal CNNs.
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 | 0.001 |
| 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".