To address critical safety and economic challenges in large-scale electrochemical battery energy storage power stations—such as thermal runaway risks, low operational efficiency, and performance degradation—an Intelligent Operation and Maintenance System (iOMS) has been developed. The system adopts a collaborative “cloud-edge-end” architecture supporting multi-source data acquisition and online analysis. At the edge, integrated diagnostic algorithms combining mechanism models and AI-powered expert knowledge graphs enable multi-level warning from cells and modules to racks and ultimately, the containerized system. On the cloud, a group-level supervision platform facilitates data aggregation and strategy deployment. Practical validations show that the system achieves: (1) thermal runaway warnings at least 2 h in advance, (2) over 90% reduction in converter failure rates, (3) 20% increase in station online rate, (4) 30% lower operational workload. The iOMS has been successfully applied in multiple energy storage stations in China, demonstrating effectiveness in optimizing voltage consistency, identifying temperature abnormalities, and detecting PCS defects. It provides a comprehensive solution for safe, efficient, and intelligent operation of large-scale energy storage power stations, significantly enhancing both economic benefits and safety performance.