Intelligent Decision-Making Method for Fault Handling Based on Knowledge Graph

Abstract

The scale of the power grid continues to expand and the stability characteristics become more and more complex. Fast and effective fault handling is an important means to ensure the safe and stable operation of the power grid. A fault handling intelligent decision-making method integrating knowledge graph and data mining algorithm is proposed, and a two-level scheduling knowledge graph covering multiple subgraphs is constructed. The upper atlas calls the lower business sub atlas based on the fault handling rules obtained from historical data learning and training. The lower-level map realizes business functions such as on-duty strategy analysis, stable limit matching, and disposal strategy generation of the security control system. Mining fault handling rules based on indicators such as sensitivity, and using massive historical operation data for dynamic update of two-level knowledge graphs to solve the problem that offline handling rules do not match fault scenarios. Finally, the effectiveness of the method is verified by the actual power grid operation data.