Fuzzy logic approach for transformer remnant life prediction and asset management decision

Abstract

The quality of insulation system within power transformer reflects the overall health condition of the asset. A timely and reliable maintenance decision along with the remaining operational life estimation of the asset can be identified through the measurements of some parameters reflecting the degradation rate of the transformer dielectric insulation. However, the process is not always straightforward due to the complexity of the insulation structure and degradation process. This paper proposes a new fuzzy logic approach to provide a proper asset management decision and predict the remaining operational life of a power transformer based on some insulating oil tests. The developed fuzzy logic model is validated through field data collected for various transformers of preknown health condition and life span. Results show that the proposed model is reliable and can be facilitated to provide a timely asset management decision with less reliance on expert personnel.