一基于ITD 和模糊聚类的齿轮箱故障诊断方法.doc

文档介绍：2013 年摇第 37 卷摇摇摇摇摇摇中国石油大学学报(自然科学版)摇摇摇摇摇摇摇摇 Vol. 37摇 No. 4
摇摇文章编号:1673鄄5005(2013)04鄄0133鄄07摇摇摇 doi:10. 3969 / j. issn. 1673鄄5005. 2013. 04. 022
一于 ITD 和模糊聚类的齿轮箱故障诊断方法
段礼祥, 张来斌, 岳晶晶
(中国石油大学机械与储运工程学院,北京 102249)
摘要:为提高齿轮箱故障诊断的准确性与效率,针对其振动信号非线性和非平稳性的特点,提出将固有时间尺度分
解(ITD)和模糊聚类(FCM)相结合的齿轮箱故障诊断方法。首先对齿轮箱振动信号进行固有时间尺度分解,提取包
含主要故障信息的前 4 个固有旋转分量(PRC),求取 PRC 的特征能量作为故障特征向量。然后利用模糊 C鄄均值聚
类算法对齿轮箱故障进行识别与诊断,并将该方法应用到现场齿轮箱的诊断中。结果表明,诊断结果与实际情况完
全相符,该方法比经验模式分解与模糊聚类相结合的方法具有更高的计算速度和精度,为齿轮箱故障诊断提供了一
种新的有效方法。
关键词:固有时间尺度分解; 模糊 C鄄均值聚类; 齿轮箱; 故障; 诊断
中图分类号:TK 407摇摇摇文献标志码:A
Fault diagnosis method of gearbox based on intrinsic time鄄scale
position and fuzzy clustering
DUAN Li鄄xiang, ZHANG Lai鄄bin, YUE Jing鄄jing
(Faculty of Mechanical and Oil鄄Gas Storage and Transportation Engineering in China University of Petroleum,
Beijing 102249, China)
Abstract: Considering the non鄄linear and non鄄stationary characteristics of vibration signals of gearbox, a new method based
on intrinsic time鄄scale position ( ITD) and fuzzy center鄄means clustering ( FCM) was proposed in order to diagnose
gearbox謖s fault more accurately and effectively. Firstly, original vibration signals of gearbox were posed by ITD meth鄄
od. The first four proper ponents (PRC) containing the main fault information were extracted and the PRC fea鄄
ture energy was calculated as fault feature vectors. Finally, faults of gearbox were identified by using FCM method. The re鄄
sults show that the diagnosis results of gearbox are totally in accordance with the actual situation in the application. The new
method has putation speed and pared with empirical mode position ( EMD), which provides a
new efficient method for gearbox fault diagnosis.
Key words: intrinsic time鄄scale position; fuzzy center鄄means clustering; gearbox; fault; diagnosis
摇摇对齿轮箱进行故障诊断,能够减少齿轮箱故障
带来的经济损失,提高齿轮箱运行的可靠性和安全
性。针对齿轮箱振动信号所具有的强烈的非平稳特
[1]
分布[2] 、EMD 分解[3] 。然而,小波变换是对时频平
面的一种机械格型分解,缺乏对所研究信号的自适
应性[4]
时从时域和频域对时间序列进行分析,但分析多分
量信号时易产生交叉项[5] ;EMD 分解具有很强的自
适应性,