To address challenges in relay protection operation and maintenance

including low efficiency

difficulty in detecting hidden defects

and insufficient accuracy in locating complex faults

a multi-source data knowledge graph system for operation and maintenance decision-making is proposed. By integrating multi-source heterogeneous data from dispatching

protection information

and fault recording

this system constructs a panoramic knowledge graph covering associations between primary and secondary equipment

protection action logic

and the temporal evolution of faults. It is applied to coordinated intelligent diagnosis of primary and secondary systems in power grid faults and intelligent identification of hidden defects in protection devices. Application results show that the proposed method can significantly improve fault-handling efficiency

enable early warning of potential risks

alleviate on-site operation and maintenance pressure

and provide a practical and feasible technical pathway for intelligent relay protection operation and maintenance.