The intelligent platform eliminates accidents in the energy sector of enterprises due to accidents and damage to power transformers, promptly identifies defect developments and predicts the trouble-free life of high-voltage equipment.
The Russian energy system is a branched structure that includes hundreds of thousands of power transformers. According to scientists of the International Institute of Smart Materials of the Southern Federal University, mainly medium-power transformers, about 65% of the total, have already exhausted their service life, which was set by the manufacturer – 25 years.
By changing the parameters of power transformers, you can not only determine the location of the damage, but also the rate of growth of the defect, predict the condition of power electrical equipment, and most importantly, prevent the failure of expensive equipment. According to scientists, in some cases, especially in the absence of developments in diagnostics, there is a need to develop means of intelligent decision support.
The proposed hybrid approach of deep learning and HDLSS made it possible to obtain an adequate solution for predicting changes in the parameters of power transformers, as well as by analyzing additional factors to determine the residual life of transformers.
Andrey Chernov, leading researcher at the Laboratory of Artificial Intelligence and Big Data Technologies for Nanodiagnostics of Materials, MII IM SFU:
“The developed expert decision-making system based on artificial intelligence is a modern solution for analyzing multifactor results of diagnostics of power transformers and parameters of current condition monitoring. This system solves two problems: firstly, it will make it possible to predict the condition of power transformers based on the main parameters, identify the development of a defect, and determine the remaining resource for timely repair and replacement of the power transformer; the second task is to eliminate the possibility of an accident due to damage to high-voltage equipment, which is accompanied by an electric arc, ignition of transformer oil, etc.”
The development of scientists from the MII IM SFU is in demand and relevant in the energy sector of enterprises in all fields of activity, and is also significant for the development of domestic decision support systems. Taking into account the cost of a power transformer, this technique will make it possible to cost-effectively approach the decision to replace or repair power transformers.
The study was carried out within the framework of a grant from the Russian Foundation for Basic Research.
#ScienceSFU #powertransformers #SFU