A Systematic Evaluation of Machine Learning Methods for Fault Detection and Line Identification in Electrical Power Grids
Open accessAbstract
The integration of renewable energy sources into the electrical grid introduces complex challenges in fault detection and coordination of grid recovery mechanisms. Traditional relay protection systems, which operate based on static rules and predefined thresholds, are inadequate for addressing these challenges, particularly in detecting and isolating faults such as short circuits. Consequently, the conventional methodologies applied to electrical network protection frequently fail to achieve optimal performance in fault detection, especially in terms of adherence to safety standards and the selective limitation of damage. Recent research indicates that machine learning (ML)-based approaches can effectively tackle these issues; however, variations in grid configurations and analysis windows have impeded consistent comparative assessments. In this study, we assess the efficacy of various ML models in detecting electrical faults and pinpointing defective transmission lines within a 10 ms measurement interval—a critical time-frame for real-time operational viability, for the first time. The most effective model attained an F1 score of 0.991±0.018 and demonstrated a processing time of 0.342ms±0.509ms.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The RF classifier's reported performance is F1-score 0.998 for FD and 0.985 for FLI, with a runtime of 16.8 ms.
“The reported F1-score is 0.998 for FD, 0.985 for FLI, with a runtime of 16.8 ms.”
From Impact of Data Sparsity on Machine Learning for Fault Detection in Power System Protection (2025) · cites this paper as [8]
The Discussion (page not provided) includes RF among the top-performing models and reports mean F1 scores of up to 0.99 for fault detection and 0.98 for fault-line identification. Neither it nor the Experiments and Results passage (page not provided) gives the claimed RF-specific scores of 0.998 and 0.985 or a 16.8 ms runtime. (LLM, conf=0.83)
Experimental — accuracy not yet measuredThe Random Forest classifier was selected based on a comparative study [8] that demonstrated its favorable trade-off between performance and runtime.
“The Random Forest (RF) classifier is selected based on our comparative study [8], which demonstrated its favorable trade-off between performance and runtime.”
From Impact of Data Sparsity on Machine Learning for Fault Detection in Power System Protection (2025) · cites this paper as [8]
The Experiments and Results section compares RF with other classifiers and lists it among the best performers. The Discussion also reports that RF is among the top-performing models, but identifies MLP, GB, and Stacking—not RF—as combining the highest scores with competitive runtimes. The paper therefore supports using the study to assess RF, but does not establish the claimed favorable performance–runtime trade-off for RF specifically. No page numbers are available in our copy. (LLM, conf=0.84)
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Cites
Scikit-learn: Machine Learning in Python
Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Müller, A., Nothman, J., Louppe, G., et al. (2012) · arXiv · doi:10.48550/arxiv.1201.0490 · Cited by 71,016 · see citing works
Challenges in the decarbonization of the energy sector
Papadis, E., & Tsatsaronis, G. (2020) · Energy · doi:10.1016/j.energy.2020.118025 · Cited by 765 · see citing works
Electric Power Distribution Engineering
Gonen, T. (2015) · doi:10.1201/b16455 · Cited by 287 · see citing works
Wide Area Protection—Technology and Infrastructures
Adamiak, M. G., Apostolov, A., Begovic, M. M., Henville, C., Martin, K. E., Michel, G., PHADKE, A. G., & Thorp, J. S. (2006) · IEEE Transactions on Power Delivery · doi:10.1109/tpwrd.2005.855481 · Cited by 247 · see citing works
Vaish, R., Dwivedi, U. D., Tewari, S., & Tripathi, S. M. (2021) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2021.104504 · Cited by 202 · see citing works
Rizeakos, V., Bachoumis, A., Andriopoulos, N., Birbas, M., & Birbas, A. N. (2023) · Applied Energy · doi:10.1016/j.apenergy.2023.120932 · Cited by 102 · see citing works
Najafzadeh, M., Pouladi, J., Daghigh, A., Beiza, J., & Abedinzadeh, T. (2024) · International Journal of Computational Intelligence Systems · doi:10.1007/s44196-024-00434-7 · Cited by 77 · see citing works
Sapountzoglou, N., Lago, J., Schutter, B. D., & Raison, B. (2020) · Applied Energy · doi:10.1016/j.apenergy.2020.115299 · Cited by 76 · see citing works
Elektrische Kraftwerke und Netze
Oeding, D., & Oswald, B. R. (2016) · Springer eBooks · doi:10.1007/978-3-662-52703-0 · Cited by 66 · see citing works
Mbey, C. F., Kakeu, V. J. F., Boum, A. T., & Souhe, F. G. Y. (2023) · The Journal of Engineering · doi:10.1049/tje2.12324 · Cited by 40 · see citing works
Protection of Distribution Systems with Distributed Energy Resources
Bak‐Jensen, B., Browne, M., Calone, R., González, R. C., Craib, A., Donnart, G., Dumitrașcu, D.-G., Engel, M., Hanuš, R., Høidalen, H. K., et al. (2015) · VBN Forskningsportal (Aalborg Universitet) · Cited by 36 · see citing works
Hou, S.-Z., Guo, W., Wang, Z., & Liu, Y.-T. (2021) · IEEE Sensors Journal · doi:10.1109/jsen.2021.3133352 · Cited by 32 · see citing works
Protection Coordination of AC/DC Intersystem Faults in Hybrid Transmission Grids
Prommetta, J., Schindler, J., Jaeger, J., Keil, T., Butterer, C., & Ebner, G. (2020) · IEEE Transactions on Power Delivery · doi:10.1109/tpwrd.2020.3000731 · Cited by 23 · see citing works
Hybrid fuzzy evaluation algorithm for power system protection security assessment
Meyer, G. J., Lorz, T., Wehner, R., Jaeger, J., Dauer, M., & Krebs, R. (2020) · Electric Power Systems Research · doi:10.1016/j.epsr.2020.106555 · Cited by 13 · see citing works
Protection Algorithms for Closed-Ring Grids With Distributed Generation
Biller, M., & Jaeger, J. (2022) · IEEE Transactions on Power Delivery · doi:10.1109/tpwrd.2022.3144004 · Cited by 11 · see citing works
Faulted Line Identification in Power Network using Unsupervised Machine Learning
Kumar, A. R., & Kundu, P. (2023) · International Conference on Power Systems (ICPS) · doi:10.1109/icps60393.2023.10428707 · Cited by 1 · see citing works
Flagged first, then by citations
Research works printed but not linked: 3
- W.-K. Chen, The electrical engineering handbook. Elsevier Academic Press, 2005, OCLC: 57371415.
- IEEE Power and Energy Society, "IEEE guide for determining fault location on AC transmission and distribution lines." [Online]. Available: http://ieeexplore.ieee.org/document/7024095/
- M. Wang, G. Kordowich, and J. Jäger, "A generic data generation framework for short circuit detection training of neural networks," in PESS + PELSS 2022; Power and Energy Student Summit. VDE, 2022, pp. 49-54. [Online]. Available: https://ieeexplore.ieee.org/document/ 10104226
Cited by
Sorted by citations
Impact of Data Sparsity on Machine Learning for Fault Detection in Power System Protection
Oelhaf, J., Kordowich, G., Kim, C., Pérez-Toro, P. A., Maier, A., Jäger, J., & Bayer, S. (2025) · 2025 33rd European Signal Processing Conference (EUSIPCO) · doi:10.23919/eusipco63237.2025.11226584
“For further details on the parameter randomization process, refer to our previous study [8].”
InfluentialSelf-citationLightweight Feature-Focused FCNN for Power Equipment Health Monitoring on Edge Devices
Mu, C., Wang, Z., Wang, P., Yang, R., Wang, K., & Wang, B. (2026) · Lecture notes in electrical engineering · doi:10.1007/978-981-92-1635-2_15
- Authors
- Oelhaf, J., Kordowich, G., Pérez-Toro, P. A., Arias-Vergara, T., Maier, A., Jäger, J., & Bayer, S.
- Affiliations
- Friedrich-Alexander-Universität Erlangen-Nürnberg (DE)
- Friedrich-Alexander-University Erlangen-Nuremberg,Pattern Recognition Lab,Erlangen,Germany
- Institute of Electrical Energy Systems, Friedrich-Alexander-University Erlangen-Nuremberg,Erlangen,Germany
- Venue
- ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
- Publisher
- IEEE
- Published
- 2025-04-06
- Pages
- 1-5
- ISSN
- 1520-6149 (ISSN-L)
- OpenAlex
- W4408353127
- Cited by
- 2
- References
- 16
- Citations by year
- 2026: 1 · 2025: 1
- FWCI
- 4.37
- Type
- conference paper
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- IEEE International Conference on Acoustics Speech and Signal Processing (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Funders
- Deutsche Forschungsgemeinschaft (535389056)
- Topics
- Physical Sciences › Engineering › Control and Systems Engineering › Power Systems Fault Detection
- Physical Sciences › Engineering › Electrical and Electronic Engineering › Electrical Fault Detection and Protection
- Physical Sciences › Engineering › Electrical and Electronic Engineering › Power System Optimization and Stability
- Keywords
- fault detection, electrical power grid, machine learning, short-circuit faults, relay protection, renewable energy integration, real-time fault detection, model comparison
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
