Answers are grounded in this paper and cite their sources.
Nowadays, there are many different methods of transformer diagnostics. The analysis of used methods and diagnostic systems indicates that a certain complexity of further development of existing methods and diagnostic systems has been achieved. This is due to the complexity of input signals, quite a large number of input factors, nonlinear multiple dynamic interrelationships with other parameters. One of the most promising types of diagnostics, to date, is frequency response analysis. The objective of this paper was to identify various transformer defects by analysing the frequency response. In this paper, frequency response analysis based on the three voltmeter method is used to detect core and winding defects. In a series of experiments, impedance and phase-frequency characteristics of transformers with core and winding defects are obtained. These characteristics show significant differences between the normal and emergency states of the transformers. The obtained characteristics in the form of pictures are the initial data for the convolutional neural network, which determines the type of defect. The use of frequency characteristics of single-phase and three-phase transformers in diagnostics of pre-failure states and failures will allow to create a universal hardware-software complex of diagnostics for transformers of different types and nominal data.
No takes yet. Share an insight, caveat, or question.
Hramyka et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: