Research On Construction And Application For Knowledge Graph Of CNC Machine Tools Fault Repair

With the continuous popularity of the Internet and the development of artificial intelligence technology,the amount of data has increased exponentially.In the era of big data,how to quickly obtain valuable information from massive data and effectively use it is one of the important issues in the field of data mining and analysis.A picture is worth a thousand words.Expressing complex text data in the form of a picture can enable people to obtain the required information faster and more accurately,and better understand the information.The knowledge graph is born.In recent years,with the introduction and development of "Internet +" and "Made in China 2025",knowledge graphs have been favored by the industrial field,and intelligent manufacturing has become the key to the transformation of traditional industries.In this paper,the construction and application of the fault maintenance knowledge graph of Computer numerical control machine tools(CNC machine tools)in the industrial field are studied.The main research results are as follows.(1)Aiming at the fault maintenance field of CNC machine tools,a named entity recognition method based on the BLSTM-LCRF model is proposed.The BLSTM neural network is employed to learn the hidden features of the text automatically.Besides,the output of the BLSTM neural network is used as the input of the LCRF and decoded through the LCRF to obtain the optimal global sequence.This method solves the problems of the need for a large number of manual annotations,the difficulty of maintenance,the difficulty of obtaining long-distance dependence,and the high cost in the early named entity recognition method.Experiments show that the proposed method achieves better recognition performance on the CNC machine tool fault repair data set.(2)An entity relationship extraction method based on PCNN-BLSTM-Attention deep neural network model is proposed.First,defining the entity-relationship category,the text information is converted into word vectors and spliced with the position feature vector as the joint feature vector at the same time.The joint feature vector is introduced into the PCNN-BLSTM-Attention model for entity relationship extraction.Entering different relationship extraction models of the CNC machine tool fault repair data set for comparative experiments,the experiments show that the proposed method is more effective than other comparable models for extraction.(3)A knowledge graph based on the field of CNC machine tool fault repair is constructed and applied it simply.First,the data obtained in the first two chapters are converted into CSV text format,and the file encoding form of CSV file is changed to the UTF-8 encoding form by using the "Notepad++" text editor;Then the sorted file is imported into the Neo4 j graph database,and the whole picture of the CNC machine tools fault maintenance knowledge graph is displayed by using cypher language.Finally,the knowledge graph of CNC machine tools fault maintenance is simply applied by cypher language,which provides a basis for the diagnosis of CNC machine tools and improves the efficiency of fault repair.