摘要
为提高轴承故障诊断的准确率,以灰色关联理论和信息熵理论为基础,提出了基于灰关联信息熵提取属性特征的支持向量机决策树多故障分类器。该分类器可以实现对轴承的多故障类型的分类,并对轴承的各类故障进行了分类实验。验证结果表明,该方法可有效地进行故障状态识别,达到了准确进行机械系统多故障诊断的目的。
To improve the accuracy of bearing fault diagnosis,a support vector machine decision tree multi-fault classifier,which extracts attribute characteristics by using grey correlation Information entropy,is proposed based on grey correlation theory and information entropy theory.The classifier achieves the classification of multiple fault types of bearing,verifying various faults of bearing.The result shows that the method can recognize the fault condition effectively,achieving the purpose of the accurate multi-fault diagnosis of mechanical systems.
出处
《振动．测试与诊断》
EI
CSCD
北大核心
2013年第3期504-508,531-532,共5页
Journal of Vibration,Measurement & Diagnosis
基金
国家自然科学基金资助项目(61175080)
关键词
支持向量机(SVM)
决策树
轴承
灰色关联
信息熵
support vector machine,decision tree,bearing,grey correlation,information entropy
