摘要
提出了一种全新的分类框架,将故障诊断方法整体分为两大类,即定性分析的方法和定量分析的方法.对现有的方法在此框架下进行了划分,并详细介绍了每种方法的基本思想、研究进展和典型应用,其中重点讨论了数据驱动的方法.最后,简述了故障预测的研究现状,并探讨了故障诊断研究存在的问题和未来的发展方向.
A novel classification framework is proposed, which divides fault diagnosis approaches into two classes: qualitative analysis approaches and quantitative analysis approaches. The basic idea, main research progresses, and typical applications of each method are discussed in detail, with emphasis on the data-driven approaches. The state-of-the-art of fault prediction is also outlined. Finally, some problems and development trends of the research on fault diagnosis are pointed out.
出处
《自动化学报》
EI
CSCD
北大核心
2009年第6期748-758,共11页
Acta Automatica Sinica
基金
国家重点基础研究发展计划(973计划)(2009CB320602)
国家自然科学基金(60721003
60736026)资助~~
关键词
动态系统
故障诊断
故障预测
数据驱动
Dynamic systems, fault diagnosis, fault prediction, data-driven