Abstract: Real-time measurement and simulation information of large power grids based on the massive and high-dimensional, advanced information technologies such as big data, artificial intelligence and knowledge graph, deeply mining and characterizing the knowledge of power grid operation situation and displaying it intuitively, are the core of building a large power grid online intelligent active safety defense system. First, under the ideological system of the intelligent panoramic system, adopting the fusion mechanism of "physical system + information system + knowledge graph", facing the construction needs and dispatching operators of the intelligent panoramic power grid, and comprehensively considering the stable state evaluation indicators and optimal control strategy information of different operating scenarios of the power grid, this paper gives the basic concept and connotation of the knowledge graph of power grid operation situation and introduces its core elements and construction goals in detail. Then, relying on the smart grid dispatching spatio-temporal big data support platform, the software functional architecture of the power grid operation situation knowledge graph is designed, including multi-source information fusion, spatio-temporal big data platform, situation graph engine, knowledge association reasoning, graph business application and friendly human-computer interaction. On this basis, the up-bottom construction process of the knowledge graph of power grid operation situation is given, and the logical framework, technical framework and a series of technologies for operation situation knowledge graph construction such as knowledge modeling, knowledge extraction, knowledge fusion, knowledge processing, and knowledge updating are systematically introduced. Finally, an intelligent construction and visualization system for the knowledge graph of the operation situation of the large power grid is developed. The system has the characteristics of simplicity, intuitive operation, and user-friendliness. It has been tested and applied in a provincial power grid, and the functional framework and technical system proposed in this paper have been verified. It can also provide technical support for the realization of the information-driven intelligent panoramic active safety defense of the large power grid.