A fault detection method for automated industrial equipment based on multi attribute decision fusion in knowledge graph

Automated industrial equipment is an important production equipment in modern industry, but the occurrence of equipment failures may seriously affect production capacity. A method based on multi-attribute decision fusion was studied and designed for fault detection of automation industrial equipment. During the process, a mapping structure between the data layer and the pattern layer of the knowledge graph was designed. Knowledge extraction was performed on unstructured and semi-structured texts, and the fault knowledge graph was established through knowledge verification operations. Then, the fault alarm data was processed using Cypher query language, and the semantics were blurred using fuzzy set theory. Finally, the correctness of the fault chain was analyzed through attribute weights and attribute value matrices. Then it searched for the source fault node of the fault. The experimental results showed that the research method maintains an average accuracy of 0.8046 or above in the mean accuracy test when the number of traceability fault chains is 17-18. In the analysis of actual fault detection effectiveness, the research method focused on the fault detection time of the 8-station robotic arm swing plate robot when the number of fault nodes involved increased to 12, which was only 72ms. This indicated that the research method can effectively detect faults in automated industrial equipment and has more accurate detection accuracy.