Question Answering Algorithm for Grid Fault Diagnosis based on Graph Neural Network

Abstract

Due to the existence of uncertain factors such as the power grid system itself, natural climate change and human factors, various faults will still occur in the power grid system. If the fault alarm is not responded to in time, it is likely to cause grid instability or even collapse, resulting in inestimable losses. By building a knowledge graph for massive power grid operation and maintenance information, we can achieve fast and accurate fault information reasoning and traceability, and retrieve reasonable fault resolution measures. Use artificial intelligence technology and big data to assist power grid systems to achieve more efficient operation and maintenance. Realizing the intelligent fault diagnosis of power grid is an urgent problem to be solved at present. With the rapid development and application of artificial intelligence technology, if artificial intelligence and big data technology can be applied to the fault diagnosis and analysis of power grids, this situation of relying on manual analysis will be broken, and the efficient processing of massive operation and maintenance data will be realized.