Transformer High Voltage Equipment System Based on Machine Learning

Online monitoring and fault diagnosis of high-voltage power equipment in the power system is an important part of achieving the development of intelligent power systems. However, there are a wide variety of insulation faults in the power system that are not universally applicable, making existing detection technologies difficult to adapt to the development needs of smart grids. Therefore, this article is based on a convolutional neural network (CNN) method for monitoring the status of transformer high-voltage equipment. By extracting features from the voltage and current signals of transformer high-voltage equipment during operation and inputting them into the CNN network for classification, real-time monitoring of the operating status of transformer high-voltage equipment is achieved. The experimental results show that the method applied in this article can quickly and accurately classify the operating status of transformer high-voltage equipment, and the fault recognition accuracy based on CNN model can reach over $\mathbf{9 7. 9 \%}$. This method is superior to traditional methods in terms of accuracy, detection speed, and cost consumption in fault identification, and the effectiveness and potential of CNN in transformer fault diagnosis.