APPLICATION OF ARTIFICIAL NEURAL NETWORKS FOR ONLINE MONITORING AND DIAGNOSTICS OF POWER TRANSFORMERS
DOI:
https://doi.org/10.47390/ts-v3i10y2025No9Ключевые слова:
power transformers, neural networks, diagnostics, DGA, predictive maintenance, anomaly detection.Аннотация
The application of artificial neural networks (ANNs) is revolutionizing the monitoring and control of power transformer health, shifting the paradigm from scheduled preventive maintenance to intelligent condition-based service. This paper presents a comprehensive study of ANN architectures and their implementation in transformer diagnostics. Using multi-source data such as dissolved gas analysis (DGA), vibration, SFRA, and partial discharge measurements, the study demonstrates how neural networks enable early fault detection, residual life prediction, and optimization of maintenance schedules. The results indicate significant improvement in diagnostic accuracy and cost efficiency compared to traditional threshold-based systems.
Библиографические ссылки
1. Kumar, P., & Singh, B. (2023). Deep learning-based health index estimation for power transformers using hybrid autoencoder–LSTM architecture. IEEE Transactions on Power Delivery, 38(4), 2995–3007.
2. Zhao, Y., Chen, H., & Li, X. (2022). Anomaly detection in transformer monitoring data using probabilistic neural networks. Electric Power Systems Research, 211, 108441.
3. Sharma, N., & Das, S. (2021). Bayesian inference and fuzzy logic-based reliability assessment of substation transformers. International Journal of Electrical Power & Energy Systems, 133, 107205.
4. Liu, J., Zhang, W., & Gao, D. (2023). A hybrid SCADA-IoT architecture for intelligent transformer diagnostics. IEEE Access, 11, 107931–107945.
5. Sarkar, A., & Bose, T. (2022). Data-driven prognostics of power transformers using adaptive neuro-fuzzy inference systems. IET Generation, Transmission & Distribution, 16(14), 2793–2804.
6. Kandula, S., & Rajan, R. (2020). Transformer fault diagnosis through dissolved gas analysis using ensemble machine learning models. IEEE Transactions on Dielectrics and Electrical Insulation, 27(6), 1995–2006.
7. Mohanty, S., & Tripathy, M. (2024). Integration of digital twins for real-time condition monitoring of high-voltage transformers. Energies, 17(3), 872.
8. Zhou, K., & Yan, C. (2021). Predictive maintenance of power transformers using long short-term memory networks. Journal of Electrical Engineering & Technology, 16(2), 901–911.
9. Alonso, P., & García, D. (2023). Reliability-centered maintenance optimization for transformers using fuzzy Bayesian networks. Electric Power Components and Systems, 51(7–8), 655–670.
10. Bhattacharya, R., & Patel, D. (2022). Internet of Things-enabled diagnostic framework for online monitoring of power transformers. IEEE Internet of Things Journal, 9(18), 17612–17623.*
This work is licensed under a 