Research on intelligent identification method of power grid wiring diagram based on deep convolution neural network

The continuous evolution of smart grid technology has promoted the intelligent level of power dispatching automation systems to meet more stringent requirements and challenges. In the context of the growing demand for unified grid standards in power systems, there is an increasing importance in the collaborative integration and sharing of models, data, and graphs. This paper proposes a set of intelligent identification schemes for power grid graphics, combining convolutional neural network and OCR technology with CAD and PDF graphics files commonly used in power grids. The proposed solution is capable of accurately identifying information such as standard components, topology associations, and device labels in power grid graphs, and automatically converting this information into standard CIM/G file format. Furthermore, by introducing natural language processing technology, this paper effectively establishes a close relationship between graphic elements in CIM/G files and power grid model data, and achieves accurate recognition of graphic elements in power grid wiring diagrams, accurate extraction of character information, and comprehensive analysis of topology structure. The implementation of these technological innovations has established a solid foundation for advanced functions, such as automated management and rapid response to failures of the smart grid. The model association method employs image recognition to construct a power grid pattern recognition platform that assists dispatch centre management. The platform has been shown to refine graphics, integrate AI efficiency, provide a user-friendly interface, reduce labour requirements, ensure information accuracy and stabilise the operation of smart grids. Through the specific application in the engineering field, it is verified that this method improves the efficiency of model correlation, reduces the work of manual maintenance, and improves maintenance efficiency.