IDS-KG: An industrial dataspace-based knowledge graph construction approach for smart maintenance

Abstract

With the development of information technology in manufacturing enterprises, a large amount of equipment maintenance data and knowledge are recorded. These rich knowledge resources contain a vast amount of semantic and physical associations that have not yet been developed, resulting in a significant gap between equipment maintenance procedures and experiential knowledge. Therefore, this paper proposes a multi-source maintenance data management method called Industrial Dataspace (IDS), and on this basis, proposes a method for constructing an equipment maintenance knowledge graph (IDS-KG) that considers the causal relationships between faults in the equipment maintenance corpus. The method fixes procedural data on the ontology model at the upper layer of the knowledge graph and automatically mines maintenance information from empirical data, and ultimately achieves the fusion management of equipment maintenance procedure knowledge and empirical knowledge. The method is validated in the practical application of nuclear power equipment maintenance, and the experiments show that the method proposed in this paper is able to effectively fuse the procedural data and empirical data and structured as triplets, and at the same time, it is able to identify the hidden causal relationship between failures in the empirical data.