Construction And Application Of Knowledge Graph In The Field Of Fault Maintenance Of CNC Machine Tools

With the proposal of "China’s intelligent manufacturing 2035" and the transformation and upgrading of the domestic industrial field,the manufacturing industry has gradually developed in the direction of digitalization and intelligence.Computer numerical control machine tools(CNC machine tools)are the country’s industrial mother machine and play an important role in the manufacturing field.CNC machine tool failure has a great impact on the normal operation of production enterprises,so the study of its fault and maintenance knowledge is particularly urgent and important for machine tool users and CNC machine tool manufacturers.The fault and maintenance information of CNC machine tools is abundant in the text description of the record operation on site,and the subsequent data mining and visual combing are difficult.In view of this,this paper combines the fault and maintenance record text and knowledge graph of CNC machine tools,and studies and constructs the fault maintenance knowledge graph of CNC machine tools based on the fault and maintenance record text,so as to effectively sort out and display relevant information for manufacturing enterprises,help enterprises improve the fault maintenance efficiency of CNC machine tools through application query and other functions,and provide support for the next decision.The main research contents of this paper are as follows:1.According to the characteristics of fuzzy and different intentions of the fault maintenance record text description record in the preprocessing stage of fault data of CNC machine tools,the semantic analysis of the fault maintenance record text is carried out,and the similarity calculation of the fault text is proposed by using cosine similarity,and the similarity comparison of similar text is based on the formula,which solves the problem of difficult data description,effectively integrates the text information,and experiments prove the effectiveness of the method proposed in this paper.2.Aiming at the entity relationship extraction of fault maintenance knowledge,a PCNN-BLSTM extraction method is proposed,which defines the entity,converts the text information of the fault maintenance record into the form of the feature vector,and enters the result to the PCNN-BLSTM model to extract the entity,and finally enters the fault maintenance data of the CNC machine tools into the model for extraction experiments,and the experiment proves that the entity relationship extraction model proposed in this paper has a good effect.3.The knowledge graph of fault maintenance of CNC machine tools is constructed,and on the basis of data preparation,it is proposed to use echarts software to construct the knowledge graph of fault data,and realize the application display of the knowledge map of fault maintenance of CNC machine tools through echarts language and Java Script language.