Constructing Method of FOD Knowledge Graph Based on Civil Aircraft Assembly Process

This paper analyzes the causes of Foreign Object Debris (FOD) generated during the civil aircraft assembly, collects samples of FOD at the assembly site, and proposes a knowledge graph classification and detection scheme for FOD. This method uses temporal knowledge graph analysis to threshold different scenes in time and space, and analyzes the characteristics of FOD from the perspectives of physical stoichiometry and appearance features to achieve more accurate FOD recognition and detection. This method takes entities and timestamps as inputs to the entity embedding function to preserve the temporal perception features of entities at any point in time. Key techniques such as energy spectrum, spectroscopy, and mass spectrometry were used to qualitatively and quantitatively analyze 103 FOD image data. Design a data knowledge base to store and retrieve graph data, and ultimately build a comprehensive and efficient FOD knowledge graph manual.