In order to ensure the safe and efficient operation of the emergency command system of hydropower stations and improve the level of digitalization and intelligentization of emergency responses，the difficulties of traditional emergency responses of hydropower stations were analyzed. The knowledge graph technology was used to standardize the massive heterogeneous information of hydropower stations，and a standardized information base was formed. The research and application of multi-source heterogeneous information fusion technology in the process of accident responses were carried out to realize the real-time acquisition，processing，and analysis of multi-source heterogeneous information. By means of data mining，analysis and processing，fusion and application，the fault early warning model was built to accurately study the nature of the accident and fault location and automatically generate the optimal response guidance scheme，so as to guide emergency response personnel to carry out fault control，equipment isolation，and risk identification safely and efficiently. In addition，it helped deduce the effect of emergency response measures and fault development situations intelligently in the response process and dynamically optimize response measures. At the same time，by building a visual information sharing platform，the key points of accident responses and the progress of measures were displayed dynamically，and the key guidance information was pushed intelligently according to the division of duties of the personnel in the emergency command system，so as to realize the complementary interaction and cooperative command of each level of emergency response and establish an efficient，intelligent，and visual emergency command system.