A Hybrid Multi-Criteria Decision-Making Based Fuzzy System for Comprehensive Health Assessment of Power Transformers

Precise evaluation of transformer health using well-structured asset management methodologies is essential, and over the last two decades, several health assessment methods have been developed based on various diagnostic tests. However, the literature on methods to ensure a comprehensive health evaluation with a focus on prioritized correlation analysis is limited. In this context, this work proposes a framework for evaluating the performance of power transformers using 15 diagnostic attributes. The model is designed with Analytical Hierarchy Process (AHP), Grey Relational Analysis (GRA), and Technique for Order Performance by Similarity to Ideal Solution (TOPSIS). The diagnostic attributes are initially normalized, and then prioritized based on closely related characteristics and finally grouped into three specific health degrees using the proposed systematic hybrid decision-making approach. The three health degrees are further coupled with a technical fuzzy rule base to derive the actual health status of power transformers. The hybrid model is validated using known health conditions of 200 transformers collected from Himachal Pradesh State Electricity Board, India, and compared with recent similar health prediction methods published in the literature. The comparison reveals the superiority of the model with 97% accuracy. Moreover, the following are the main advantages of the model over the existing health assessment methods: prioritization among closely related attributes, assessment of samples with respect to positive and negative ideal solutions, no dependency on training data, less complexity, and consideration of visual inspection and maintenance parameters. Furthermore, it facilitates an appropriate alert system for conditional maintenance.