A Hybrid Deep-Learning and Impedance-Based Scheme for Fault Classification and Location in Transmission Lines: A Physics-Validated, Leakage-Free Evaluation
DOI:
https://doi.org/10.59992/IJCI.2026.v5n10p2Keywords:
Fault classification, fault location, transmission line, convolutional neural network, Takagi method, power system protectionAbstract
Deep-learning fault classifiers and locators for transmission lines are often evaluated on unvalidated simulations, with possible data leakage and without comparison against the impedance-based methods used in practice. This study builds an exact time-domain simulator of a 400 kV, 200 km two-source line, whose fault currents agree with symmetrical-component calculations within 2%, and generates 22,600 independent scenarios covering ten fault types, fault resistance, location, inception angle, loading and source strength. Single-ended 25 ms windows sampled at 4 kHz were used to train one-dimensional convolutional neural networks (CNNs), which were compared with feature-based classifiers and with the reactance and Takagi locators. The CNN classified faults with 99.91% accuracy in 0.4 ms per window, but feature-based classifiers were equally accurate, and all learned classifiers failed for fault resistances beyond the training range until that range was included in training, which in turn increased false alarms under strong noise. CNN regression located faults with a mean absolute error of 5.81 km, better than other learned models but far worse than the Takagi method on the third post-fault cycle. A hybrid scheme, in which the CNN supplies the fault type and the Takagi method the distance, reduced the error to 1.40 km, located 93.8% of faults within 2% of the line length, and remained accurate for weak sources, noise and high fault resistance. Deep learning is thus best used for fast classification, and physics-based methods for precise location.
References
[1] Akhavan, M., & Yazdaninejadi, A. (2026). Enhanced one-ended fault location using superimposed components, source impedance estimation, and distributed line parameters. IEEE Access, 14, 54247–54256. https://doi.org/10.1109/ACCESS.2026.3681303
[2] Akter, S. B., Pias, T. S., Deeba, S. R., Hossain, J., & Rahman, H. A. (2024). Ensemble learning based transmission line fault classification using phasor measurement unit (PMU) data with explainable AI (XAI). PLOS ONE, 19(2), Article e0295144. https://doi.org/10.1371/journal.pone.0295144
[3] Das, S., Santoso, S., Gaikwad, A., & Patel, M. (2014). Impedance-based fault location in transmission networks: Theory and application. IEEE Access, 2, 537–557. https://doi.org/10.1109/ACCESS.2014.2323353
[4] Hossain, M. I., Anonto, H. Z., Riyad, T., Shufian, A., Hossain, M. S., & Pathik, B. B. (2026). Adaptive fault diagnosis in power transmission lines using deep learning and LSTM autoencoders for enhancing grid reliability. International Journal of Electrical Power & Energy Systems, 174, Article 111458. https://doi.org/10.1016/j.ijepes.2025.111458
[5] Jiménez-Aparicio, M., Hernández-Alvidrez, J., Montoya, A. Y., & Reno, M. J. (2022). Embedded, real-time, and distributed traveling wave fault location method using graph convolutional neural networks. Energies, 15(20), Article 7785. https://doi.org/10.3390/en15207785
[6] Kapoor, S., & Narayanan, A. (2023). Leakage and the reproducibility crisis in machine-learning-based science. Patterns, 4(9), Article 100804. https://doi.org/10.1016/j.patter.2023.100804
[7] Kouraichi, M., Mansouri, M., Sakly, A., Trabelsi, M., & Mnif, F. (2026). A comprehensive dataset and simulation framework for fault analysis in power transmission systems: Dataset description and statistical insights. IEEE Access, 14, 37968–38008. https://doi.org/10.1109/ACCESS.2026.3670392
[8] Liang, Y., Zhang, Z., Li, H., Ding, J., Wang, G., & Chen, L. (2022). A robust and accurate discrete Fourier transform–based phasor estimation method for frequency domain fault location of power transmission lines. IET Generation, Transmission & Distribution, 16(10), 1990–2002. https://doi.org/10.1049/gtd2.12408
[9] Lima, D. A. C., Bernardon, D. P., Morais, A. P., Oliveira, A. L., Hokama, W. S., Conceição, J. B. R., & Sartori, Â. F. (2022). Review of bus differential protection using IEC 61850. Energies, 15(24), Article 9537. https://doi.org/10.3390/en15249537
[10] Machlev, R., Heistrene, L., Perl, M., Levy, K. Y., Belikov, J., Mannor, S., & Levron, Y. (2022). Explainable artificial intelligence (XAI) techniques for energy and power systems: Review, challenges and opportunities. Energy and AI, 9, Article 100169. https://doi.org/10.1016/j.egyai.2022.100169
[11] Mishra, S., Gupta, S., & Yadav, A. (2023). Traveling-wave based fault location estimation in FACTS compensated transmission system: Recent status, challenges and future research directions. MethodsX, 11, Article 102365. https://doi.org/10.1016/j.mex.2023.102365
[12] Nayak, P., Das, S. R., Mallick, R. K., Mishra, S., Althobaiti, A., Mohammad, A., & Flah, A. (2024). 2D-convolutional neural network based fault detection and classification of transmission lines using scalogram images. Heliyon, 10(19), Article e38947. https://doi.org/10.1016/j.heliyon.2024.e38947
[13] Ogar, V. N., Hussain, S., & Gamage, K. A. A. (2022). Transmission line fault classification of multi-dataset using CatBoost classifier. Signals, 3(3), 468–482. https://doi.org/10.3390/signals3030027
[14] Ogar, V. N., Hussain, S., & Gamage, K. A. A. (2023). The use of artificial neural network for low latency of fault detection and localisation in transmission line. Heliyon, 9(2), Article e13376. https://doi.org/10.1016/j.heliyon.2023.e13376
[15] Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Prettenhofer, P., Weiss, R., Dubourg, V., Vanderplas, J., Passos, A., Cournapeau, D., Brucher, M., Perrot, M., & Duchesnay, É. (2011). Scikit-learn: Machine learning in Python. Journal of Machine Learning Research, 12, 2825–2830.
[16] Rezapour, H., Jamali, S., & Bahmanyar, A. (2023). Review on artificial intelligence-based fault location methods in power distribution networks. Energies, 16(12), Article 4636. https://doi.org/10.3390/en16124636
[17] Shadi, M. R., Ameli, M.-T., & Azad, S. (2022). A real-time hierarchical framework for fault detection, classification, and location in power systems using PMUs data and deep learning. International Journal of Electrical Power & Energy Systems, 134, Article 107399. https://doi.org/10.1016/j.ijepes.2021.107399
[18] Shakiba, F. M., Azizi, S. M., Zhou, M., & Abusorrah, A. (2023). Application of machine learning methods in fault detection and classification of power transmission lines: A survey. Artificial Intelligence Review, 56(7), 5799–5836. https://doi.org/10.1007/s10462-022-10296-0
[19] Shakiba, F. M., Shojaee, M., Azizi, S. M., & Zhou, M. (2022). Generalized fault diagnosis method of transmission lines using transfer learning technique. Neurocomputing, 500, 556–566. https://doi.org/10.1016/j.neucom.2022.05.022
[20] Sinha, P., Paul, K., Saiprakash, C., Abdelaziz, A. Y., Omar, A. I., Su, C.-L., & Elsisi, M. (2023). Identification of cross-country fault with high impedance syndrome in transmission line using tunable Q wavelet transform. Mathematics, 11(3), Article 586. https://doi.org/10.3390/math11030586
[21] Su, C., Yang, Q., Wu, X., Lai, C. S., & Lai, L. L. (2023). A two-terminal fault location fusion model of transmission line based on CNN-multi-head-LSTM with an attention module. Energies, 16(4), Article 1827. https://doi.org/10.3390/en16041827
[22] Sun, P., Liu, X., Lin, M., Wang, J., Jiang, T., & Wang, Y. (2023). Transmission line fault diagnosis method based on improved multiple SVM model. IEEE Access, 11, 133825–133834. https://doi.org/10.1109/ACCESS.2023.3333939
[23] Takagi, T., Yamakoshi, Y., Yamaura, M., Kondow, R., & Matsushima, T. (1982). Development of a new type fault locator using the one-terminal voltage and current data. IEEE Transactions on Power Apparatus and Systems, PAS-101(8), 2892–2898. https://doi.org/10.1109/TPAS.1982.317615
[24] Yuvaraju, V., Thangavel, S., & Golla, M. (2024). Applications of artificial intelligence and PMU data: A robust framework for precision fault location in transmission lines. IEEE Access, 12, 136565–136587. https://doi.org/10.1109/ACCESS.2024.3464088
