Power System Knowledge Graph Construction and Fault Disposal Contingency Plan Generation Based on Human-in-the-loop

The application of knowledge graph technology in the field of grid fault disposal plan generation has become a research hotspot. However, the current defects of security, interpretability and generalization have become a bottleneck hindering its practical application. To overcome these defects and make the fault disposal plan more effective, the paper focuses on the introduction of a methodology for generating fault disposal plans after the human presence in the loop. Firstly, the BERT named entity recognition model is used to achieve grid topology entity recognition. Entity relationship extraction is implemented using the Text-CNN relationship extraction model. On basis of the above, a knowledge graph of the grid topology was established. Secondly, introducing the human loop into the construction of the knowledge graph and the generation of the fault disposal plan. Thirdly, the fault disposal plan is visualized. Finally, the validity of the proposed method is validated in a practical power grid with real-operation data.