Abstract: Marine ranching equipment monitoring system is an important part of marine ranching equipment system, and also a guarantee system for normal operation and equipment safety of marine ranching．In order to enhance comprehensive understanding of equipment monitoring system for marine ranching managers and front-line operators and improve level of equipment maintenance, break through blind supervision caused by fragmentation of relevant information and imperfect knowledge system, knowledge graph of monitoring system was constructed．Top-down and bottom-up methods were adopted to obtain comprehensive information from five aspects, namely primary equipment, secondary equipment, secondary equipment function, secondary equipment working principle and application requirements of secondary equipment in monitoring system, so as to establish knowledge correlation and solve problem of knowledge fragmentation in monitoring system of marine ranching equipment．In the research, model layer was constructed from top to bottom, and framework of knowledge graph was completed through ontology modeling．Data layer was constructed from bottom up to realize data acquisition, knowledge extraction, knowledge representation and knowledge fusion, and atlas data was stored in the Neo4j database．Knowledge graph constructed provided a reference basis for subsequent design and development of intelligent question and answer system, intelligent search system, fault diagnosis system and intelligent recommendation system for marine ranching equipment．