Abstract: To address the efficiency bottlenecks in business instruction generation and link state monitoring under the "Four Synchronizations" mechanism of the State Grid, this study proposes an intelligent system based on an operations and maintenance knowledge base. The system integrates equipment status, link quality, business rules, and historical data to construct a dynamic knowledge graph, leveraging LSTM-GNN models, edge computing, and clustering optimization techniques to achieve optimal instruction generation, real-time link state monitoring, and fault diagnosis. Experiments demonstrate a 38% reduction in instruction execution time, a link data loss rate reduced to 2.0%, and a recommendation adoption rate of 94%, significantly enhancing grid business efficiency and resource utilization.