摘要
为实现对历史磁悬浮轴承故障知识的积累、组织和共享,提出一种磁悬浮轴承故障领域的知识图谱构建方法。在分析数据特点的基础上,提出磁悬浮轴承文本数据分类方法和磁悬浮轴承故障数据的语义解析框架。通过改进的跳字模型生成词向量,利用双向LSTM(BiLSTM)模型对相应的数据进行分类,通过双向匹配表KMP算法实现数据内容解析,将解析内容用Neo4j图数据库生成故障图谱并进行应用演示,对该领域知识积累、故障诊断具有一定的指导意义。
In order to realize the accumulation,organization and sharing of historical magnetic bearing failure knowledge,a knowledge graph construction method for the field of magnetic bearing failure is proposed.Based on the analysis of data characteristics,a classification method of magnetic bearing text data and a semantic analysis framework for magnetic bearing fault data are proposed.Among them,the word vector is generated by the improved word-hopping model,and the corresponding data is classified by the bidirectional LSTM(BiLSTM)model.The data content analysis is realized through the two-way matching table KMP algorithm,and the analysis content is used to generate a fault map with the Neo4 j graph database and demonstrate the application,which has certain guiding significance for knowledge accumulation and fault diagnosis in this field.
作者
田野
萧筝
王继业
娄平
严俊伟
TIAN Ye;XIAO Zheng;WANG Ji-ye;LOU Ping;YAN Jun-wei(School of Mechanical and Electronic Engineering,Wuhan University of Technology,Wuhan 430070,China;School of Information Engineering,Wuhan University of Technology,Wuhan 430070,China)
出处
《组合机床与自动化加工技术》
北大核心
2022年第9期160-163,168,共5页
Modular Machine Tool & Automatic Manufacturing Technique
基金
国家自然科学基金项目(51905397)
国家重点研发项目(2018YFB2000103)。
