An approach to neuro-fuzzy monitoring of power transformers

Abstract

The article proposes a model of an adaptive network-based fuzzy inference system for power transformer monitoring based on the analysis of dissolved gases in transformer oil. The determination of the type and nature of the developing defect is carried out using the method of gas concentration ratios. The neuro-fuzzy system was tested on the results of dissolved gases analysis of power transformers operated at power plants. The proposed model of the neuro-fuzzy system with good accuracy allows you to determine the nature of the fault in the transformer. The proposed model of a neurofuzzy system can be used to build a continuous online monitoring system for power transformers