Identification of Primary Wiring Diagram in Substation Based on YOLOv5

To solve the substation electrical wiring diagrams are difficult to identify. Based on YOLOv5 algorithm, a hierarchical classification method using a tree structure was proposed for the detection of primary wiring diagrams in substations, resulting in a method for extracting internal information from wiring diagrams. This method is based on the positional correlation between different electrical elements and achieves segmentation and recognition of tuple regions in electrical diagrams. The wiring diagram is identified step by step according to voltage level, wiring type, and device order. After each recognition, the output is extracted and classified as the input for the next recognition, solving the problem of differential identification of small targets and differential differentiation of connection relationships, reducing interaction, and improving recognition speed and effectiveness. It can effectually extract information from the wiring diagram, secure accessibility, and quickly and efficiently complete information.