Abstract: The coal mine electrical substation is an important part of large coal mine power supply system, and the accurate recognition of pressure panel switch state of electrical substation is a crucial aspect of the detection of power supply status in coal mines. However, the increasing number of pressure panel switches on substation cabinets has made traditional manual inspections and visual inspections inadequate due to challenges with data management and inspection quality control. In view of the above problems, a recognition method for the operating state of substation pressure panel switches was proposed in this study, which was based on YOLOv5 neural network model. In this study, the Pytorch deep learning framework was employed for model training and the method also included a preprocessing algorithm for images of the panel switches. The resulting best performing model was capable of detecting pre-processed images of the panel switches and evaluating the detection results. Experimental results demonstrate that the proposed method has the characteristics of fast detection speed and high accuracy.