Enhancing Transformer Health Monitoring With AI-Driven Prognostic Diagnosis Trends: Overcoming Traditional Methodology’s Computational Limitations
Aniket Vatsa, Ananda Shankar Hati, Akshay Kumar Rathore
Abstract
Power transformers are critical for grids, facilitating transmission, distribution, and voltage conversion. However, their insulation degrades over time, instigating catastrophic failures. This article investigates the emerging trends of artificial intelligence (AI)-driven transformer prognostic health management (TPHM) strategies to mitigate such risks, emphasizing AI’s role in overcoming traditional methodology’s computational limitations. Hybrid approaches enhance dissolved gas analysis (DGA) accuracy, while frequency response analysis (FRA) is interpreted using machine learning (ML). Deep learning automates feature-weight assessment in partial discharge (PD) and insulation analysis, eliminating manual health index (HI) formulation. The implications and strategies for optimizing AI-driven TPHM efficiency are discussed, providing valuable trends in fostering a prognostic diagnosis ecosystem.