Transformer Health Monitoring System

Year : 2025 | Volume : 15 | Issue : 03 | Page : 1 9
    By

    K.G. Dhareni,

  • Arunjith A.S,

  • Logamari Ganesh P,

  • Aameer Arshath S,

  • Dhanush M,

  1. Associate Professor, , Department of Electronics & Telecommunication Engineering, Karpagam College of Engineering, Coimbatore,, Tamil Nadu, India
  2. Student,, Department of Electronics & Telecommunication Engineering, Karpagam College of Engineering, Coimbatore,, Tamil Nadu,, India
  3. Student,, Department of Electronics & Telecommunication Engineering, Karpagam College of Engineering, Coimbatore,, Tamil Nadu,, India
  4. Student,, Department of Electronics & Telecommunication Engineering, Karpagam College of Engineering, Coimbatore,, Tamil Nadu,, India
  5. Student,, Department of Electronics & Telecommunication Engineering, Karpagam College of Engineering, Coimbatore,, Tamil Nadu,, India

Abstract

Rising demands for reliable and efficient power distribution in modern electric control grid increasingly call up for robust monitoring systems for critical substructure. Being a vital part of the power conduction system, transformer are subjected to mechanical, electrical, and environmental stresses, which, if not properly controlled, can cause failures. In this project, we propose a Transformer Health Monitoring System (THMS) using machine learning (ML) models and real-time monitoring method to assess the usable state and evaluate the health of transformer. The temperature, vibration, and crude grade have established themselves as key parameters of transformer performance and possible defect stipulation that the organization will monitor with an array of sensors. Data Points from these sensors are sent to a centralized platform for real-time processing and analysis. Some examples of machine learning algorithmic rules apply to the sensor data for anomaly detection, and fault classification; and predictive sustenance models are provided to examine the datum, discover the anomaly, and forecast possible defect before they find. By leveraging supervised and unsupervised learning techniques, the system can classify fault types, forebode degradation trends, and notice patterns in transformer demeanor. Decision trees, support vector machines (SVM), random forests, and deep learning models are sophisticated methods that increase the effectiveness of fault spotting and condition-based monitoring (CBM). Statistical analysis and machine learning driven insights lead to predictive upkeep recommendations, thereby facilitating proactive maintenance planning and operational optimization. The incorporation of machine learning greatly decreases false alarms and improves the accuracy of defect detection.

Keywords: Transformer health monitoring, machine learning, predictive maintenance, real-time monitoring

[This article belongs to Journal of Power Electronics and Power Systems ]

How to cite this article:
K.G. Dhareni, Arunjith A.S, Logamari Ganesh P, Aameer Arshath S, Dhanush M. Transformer Health Monitoring System. Journal of Power Electronics and Power Systems. 2025; 15(03):1-9.
How to cite this URL:
K.G. Dhareni, Arunjith A.S, Logamari Ganesh P, Aameer Arshath S, Dhanush M. Transformer Health Monitoring System. Journal of Power Electronics and Power Systems. 2025; 15(03):1-9. Available from: https://journals.stmjournals.com/jopeps/article=2025/view=215331


References

  1. Sharma P, Kumar S. An overview of methods and technologies for transformer monitoring and maintenance. Electr Eng J. 2020; 55(2): 121–3.
  2. Jones A, Smith R. Transformer health monitoring system advancements. IEEE Trans Power Deliv. 2021; 36(4): 2041–9.
  3. Singh D, Patel M. IoT and machine learning for power transformer predictive maintenance. Power Electron Int. 2019; 43(7): 1115–22.
  4. Islam MM, Lee G, Hettiwatte SN. A review of condition monitoring techniques and diagnostic tests for lifetime estimation of power transformers. Electr Eng. 2018;100:581-605. doi:10.1007/s00202 017-0532-4.
  5. Thango BA. Dissolved gas analysis and application of artificial intelligence technique for faultdiagnosis in power transformers: A South African case study. Energies. 2022;15:9030. doi:10.3390/en15239030.
  6. Zhang Y, Tang Y, Liu Y, Liang Z. Fault diagnosis of transformer using artificial intelligence: A review. Front Energy Res. 2022; 10: 1006474.
  7. Gridin I. Optimisation of hyperparameters. In: Automated Deep Learning Using Neural Network Intelligence: Develop and Design PyTorch and TensorFlow Models Using Python. Berkeley (CA): Apress; 2022; 31–110. ISBN: 978-1-4842-8149-9.
  8. Bouzar-Benlabiod L, Rubin SH, Benaida A. An overview of deep neural network architecture optimisation. In: Proc IEEE 22nd Int Conf Information Reuse and Integration for Data Science; Las Vegas, NV, USA. 2021 Aug 10–12; 25–32.
  9. Zhang Y, Tang Y, Liu Y, Liang Z. Fault diagnosis of transformer using artificial intelligence: A review. Front Energy Res. 2022;10:1006474. doi:10.3389/fenrg.2022.1006474.
  10. Ghatee M, Bejani MM. A comprehensive analysis of overfitting control in shallow and deep neural networks. Vol. 54. Dordrecht (Netherlands): Springer; 2021. ISBN: 0-12-345678-9.
  11. Sankhwar P. Resiliency for Power and Data Connectivity During Outages by Application of Battery Energy Storage Systems, Use of Electric Vehicle Plug-in Points and Broadband Over Powerline. Int J Innov Sci Res Technol. 2025 Mar 10; 10(2): 1750–9.
Regular Issue Subscription Review Article
Volume 15
Issue 03
Received 12/04/2025
Accepted 20/06/2025
Published 27/06/2025
Publication Time 76 Days
141

Login


My IP

PlumX Metrics