Graph-Based Automatic Identification and Correction of Bus Topology in 500kV Substations

Under the digital transformation of power systems, there is a growing demand for intelligent power grid analysis. This requires automatic and precise batch modeling over massive power equipment with standardized topology and measurement data entry. However, due to limitations in data management, power grid companies usually have errors and non-standardized naming of topological data. This work introduces a graph database to construct a semi-realtime topology data governance method. It considers semantic recognition, isolated electrical component verification, quantified relationships between electrical components, and power system operation regulations as constraints to perform automatic identification and correction of bus topology structure in substations. The method comprises a data identification module and an error data correction module to achieve automatic generation and intelligent analysis of substation topology. While a relational database relies on complex joins on data tables to process data, a graph database can efficiently perform batch synchronous parallel calculations. The method adopts a graph database instead of the commonly used relational database. The minimum grain of the data system is extended from discrete numbers to nodes and edges in the graph. The method validation section determines the 500kV-side bus topology structure through quantitative analysis of the topological data. Such a method adapts well to the development trend of the Internet and meets the needs of large-scale multi-dimensional cross-system real-time queries of the power automation analysis business.