A kind of intelligent dynamic industrial event knowledge graph and its application in process stability evaluation
Qingzong Li,Pingyu Jiang,Jianwei Wang,Maolin Yang,Yuqian Yang
DOI: https://doi.org/10.1007/s10845-024-02325-2
IF: 8.3
2024-03-05
Journal of Intelligent Manufacturing
Abstract:Event knowledge graph (EKG) is a method of representing real-world entities, events, their attributes, and the relations between them in a graph structure. The EKG has been applied in manufacturing industry to empower intelligence manufacturing. But there are limitations of generic EKG in addressing manufacturing issues. Because in the manufacturing industry, there is not only text-type knowledge but also signals, images, videos, etc. In particular, some of the relations between entities/events of the knowledge are in the form of formulas, functions, and even trained artificial intelligence models. This kind of knowledge is called Functional Knowledge in this paper. The generic EKG is suitable for representing text-type knowledge but not Functional Knowledge. Thus, the research aims to present a new kind of EKG that has the ability to represent various types of knowledge, especially Functional Knowledge. In this regard, an intelligent dynamic Industrial Event Knowledge Graph (IEKG) is proposed. Firstly, Functional Relation, Functional Triple, and a knowledge representation model based on property graphs for the schema layer of IEKG are proposed for representing Functional Knowledge. Secondly, a dynamic construction method of the instance layer of IEKG based on the event triggering mechanism is proposed, which enables the IEKG to be constructed dynamically with the production operation. Third, the constructed IEKG is applied in production monitoring using a novel graph similarity-based process stability evaluation method. Finally, a web application encapsulating our theory was developed and applied on a kneading machine in a prebaked carbon anode factory. The result shows that our proposed method has the ability to represent Functional Knowledge. Compared to the existing EKG, it has a better and broader ability of knowledge representation. The application of process stability evaluation demonstrates the potential of IEKG in addressing manufacturing issues.
engineering, manufacturing,computer science, artificial intelligence