基于EMD信息熵和支持向量机的往复压缩机轴承故障诊断Fault Diagnosis for Reciprocating Compressor Bearings based on EMD-information Entropy and SVM
王金东;代梅;夏法锋;赵海峰;
WANG Jin-dong;DAI Mei;XIA Fa-feng;ZHAO Hai-feng;Northeast Petroleum University;
摘要(Abstract)：
往复压缩机工况恶劣、结构复杂、易损件多等特点,增加了压缩机故障诊断难度。将EMD信息熵和支持向量机(SVM)技术相结合,应用于压缩机轴承故障诊断。通过EMD对压缩机轴承信号进行分解,计算其信息熵值,并提取出能反映轴承工作状态的信息熵,将其作为特征向量训练SVM网络。结果表明,EMD信息熵和支持向量机相结合的方法,可以准确识别压缩机轴承故障。
The bad work conditions,complicated structure and more easy-wearied parts,etc.,which add to the difficulties of fault diagnosis for reciprocating compressor. A compound fault diagnosis technique based on EMD-Information Entropy and Support Vector Machine( SVM) are applied to fault diagnosis of reciprocating compressor bearings. The signals of reciprocating compressor bearings were divided by the method of Empirical Mode Decomposition. The Information Entropy was calculated and the characteristics which represent the compressor working condition were extracted. The Information Entropy can be as a vector to train Support Vector Machine Network. The results show that the method of combining EMD-Information Entropy and Support Vector Machine can accurately identify failure of compressor bearings.
关键词(KeyWords)：
经验模态分解;信息熵;支持向量机;往复压缩机;故障诊断
EMD;information entropy;SVM;reciprocating compressor;fault diagnosis
基金项目(Foundation): 黑龙江省教育厅科学技术研究项目(12521051)
作者(Authors):
王金东;代梅;夏法锋;赵海峰;
WANG Jin-dong;DAI Mei;XIA Fa-feng;ZHAO Hai-feng;Northeast Petroleum University;
参考文献(References)：
- [1]周敏,王庆锋,刘文彬,等.往复压缩机故障诊断模型技术研究及应用[J].压缩机技术,2012,(4):13-15.
- [2]江志农,张进杰,敖静晖.基于支持向量机的往复压缩机示功图识别研究[J].流体机械,2012,40(5):28-32.
- [3]樊长博,张来斌,王朝晖,等.基于EMD与功率谱分析的滚动轴承故障诊断方法研究[J].机械强度,2006,28(4):628-631.
- [4]张超,陈建军,徐亚兰,等.基于EMD分解和奇异值差分谱理论的轴承故障诊断方法[J].振动工程学报,2011,24(5):539-545.
- [5]赵海洋,纪彦东,王金东,等.基于间隙运动副的往复压缩机传动机构动力学分析[J].流体机械,2013,41(3):22-26,40.
- [6]张宝军,张国丰,严智,等.核电厂主螺栓超声自动检测技术研究与实现[J].压力容器,2013,30(5):68-73.
- [7]韩中合,韩悦,朱霄珣.基于IMF信息熵与SVM的转子振动故障智能诊断方法[J].华北电力大学学报,2012,39(4):81-85.