摘要
针对机床故障诊断中人工诊断效率低、实时性差、误差大的缺点,研究提出使用分级特征库进行机床故障诊断,该研究将特征分为典型特征和非典型特征两级,在底层汇聚节点完成振动特征提取和典型特征(即A级特征)机床故障诊断,实现故障现场预警,提高了现场预警的实时性。在控制中心完成对非典型特征(即B级特征)综合处理和分析,同时对两级特征数据采用机器学习中SVM算法完成故障特征库的完善和更新,提高了机床故障诊断的准确性。通过实验证明该研究在基于特征库的机床故障诊断领域具有理论研究价值和工程应用价值,提高了数控机床故障诊断的实时性和精确性。
In order to resolve the disadvantages of low efficiency, poor real-time performance and large error in the artificial diagnosing the faults of machine tools. The study classified the features into two levels: typical features and atypical ones. It extracted vibration features and diagnosed the typical feature (A-level fea- ture) faults in the sink node of bottom layer to implement on-the-spot warning and improve the real-time performance. It handled and analyze comprehensively the atypical features (B-level feature) in the control center. Simultaneously, by using SVM arithmetic in machine learning on those features, complete and renew the fault feature database, improve the accuracy in fault diagnosis. Experiments proved that the study definitely has an important value in theory and application and can improve the real-time performance and accuracy in diagnosing the faults of digital machine tools
出处
《组合机床与自动化加工技术》
北大核心
2016年第5期87-90,共4页
Modular Machine Tool & Automatic Manufacturing Technique
基金
"沈阳特种专用数控机床产业集群国产数控系统创新应用示范"国家科技支撑计划(2012BAF13B08)
辽宁省社会科学规划基金(L15BTQ001)
关键词
分级特征
实时预警
故障诊断
hierarchical features
real-time warning
fault diagnosis