Fault attribution system and method based on knowledge reasoning

Abstract

The invention discloses a fault attribution system and method based on knowledge reasoning. The system comprises a semantic extraction module, a map matching module, a knowledge reasoning module and a root cause positioning module which are connected in sequence. The semantic extraction module obtains key information under an entity type or a relation type defined in a knowledge graph mode in the fault phenomenon description text through a semantic extraction model trained in advance; the graph matching module is used for matching the extracted entities and the relationships among the entities with the corresponding entities and relationships in the knowledge graph; the knowledge reasoning module obtains more fault attribution information through graph exploration and extension; and the root cause positioning module supplements information to circularly position the causal relationship chain layer by layer until a root cause and a fault position are positioned. According to the invention, the accuracy of fault reason analysis can be improved, and the fault analysis efficiency is greatly improved.

Description

technical field [0001] The invention relates to a fault attribution system and method based on knowledge reasoning, and belongs to the field of natural language processing. Background technique [0002] Under the traditional method, in the production or maintenance of manufacturing enterprises, the quality or inspection engineer needs to rely on their own experience to determine the cause of the failure according to the failure phenomenon, which is highly subjective. It is necessary to know that the same fault may occur with multiple phenomena at the same time. On the one hand, it is difficult to obtain the fault phenomenon comprehensively, and on the other hand, it is not easy to locate multiple fault phenomena to the same fault, which increases the difficulty of accurately analyzing the cause of the fault. It is difficult to guarantee the quality, and the efficiency of accurately locating the fault location is low, which leads to the failure to repair the fault according t...

AI Technical Summary

Problems solved by technology

It should be known that the same fault may occur simultaneously with multiple phenomena. On the one hand, it is difficult to obtain the fault phenomenon comprehensively. On the other hand, it is not easy to locate multiple fault phenomena to the same fault, which increases the difficulty of accurately analyzing the cause of the fault. Fault analysis The quality is difficult to be guaranteed, and the efficiency of accurately locating the fault location is low, resulting in failure to repair in time according to the cause of the fault, seriously affecting normal production

Images