The development of the power system cannot be separated from real-time monitoring, analysis, and control of the transmission network, and online monitoring is an important part of it. Deep learning technology is an advanced technology applied in the power industry, which can automatically identify and classify different types of transmission equipment. The application of deep learning technology in transmission networks can solve some common problems, so this article analyzes its application in transmission equipment identification. This paper mainly adopts the experimental comparative law method and comparison method, and conducts experimental operations through convolutional neural network, improved convolutional neural network and YOLO (You Only Look Once) related algorithms. The experimental results show that the Faster R-CNN (Region-based Convolutional Neural Network) model has an accuracy rate of over 35 for identifying transmission equipment.