At present

the scale of the substation system is expanding

and the operating environment of the current transformer is complex. Traditional manual maintenance is difficult to capture subtle abnormalities in time

which is easy to cause the expansion of equipment failures. Moreover

manual diagnosis depends on experience and has poor consistency

making it difficult to adapt to the operation and maintenance needs of high density and high reliability of the power grid. Intelligent diagnostic technology breakthroughs are urgently needed. This paper builds an intelligent diagnostic model. By collecting the operation data of the transformer

it extracts current

temperature and other abnormal characteristics and trains the model in combination with an improved machine learning algorithm. The test shows that model diagnosis can quickly identify fault types

provide an accurate basis for maintenance

and significantly improve the operation and maintenance efficiency and safety of the substation system.