Research on Method of State Evaluation and Fault Analysis of Dry-Type Power Transformer Based on Self-Organizing Neural Network

Abstract

Dry-type power transformer was used widely because of its advantages. But unplanned outage effect to construct a strong intelligent power grid because of various stress. Dry-type power transformer’s fault repair time is long and impossible to repair. So it is very important to realize state maintains of dry-type transformer through state monitor and diagnosis. Based on current diagnostic methods, this paper proposed using self-organizing neural network to realize dry-type power transformer the key point temperature parameters of grading evaluation and then to realize the real-time state evaluation and analysis of failure causes. Study results to prolong the dry-type power transformer life and its design production provide theoretical guidance, in order to reduce and avoid dry-type power transformer failure.