Gearbox Condition Monitoring Based on Curvilinear Component Analysis

Gearbox Condition Monitoring Based on Curvilinear Component Analysis
复制标题

基于曲线分量分析的齿轮箱状态监测

DOI:
10.4028/www.scientific.net/kem.245-246.249
复制
发表时间:
2003-07
期刊:
Key Engineering Materials
影响因子:
--
通讯作者:
Shiyuan Liu
Shiyuan Liu
中科院分区:
其他
文献类型:
--
作者:
Guanglan Liao;Tielin Shi;Shiyuan Liu

文献摘要

参考文献

相似文献

齿轮箱传动系统是机械中最基本和最重要的部件之一,在世界各地的工业部门都有应用。齿轮箱在运行过程中出现故障将导致严重的损坏。为了及时监测齿轮箱状态,本文研究了一种称为“曲线分量分析”(CCA)的新技术,该技术完成了两个任务:输入空间子流形的矢量量化和这些量化向量向输出空间的非线性投影,从而提供了子流形的揭示。利用CCA方法对不同工况下工业齿轮箱振动信号提取的特征数据进行了分析。由分析结果可知,在选择二维输出空间的情况下,CCA方法将高维特征数据投影到输出空间中,形成多个聚类区域,每个聚类区域分别表示一个特定的齿轮箱状态。因此,将出现疲劳裂纹或断齿时的齿轮箱工况与正常工况进行比较,可以清楚地识别出来。利用特征数据图像点在二维空间e中的轨迹,直观地观察齿轮箱工况的变化,及时监测齿轮箱早期故障的发展。
A gearbox transmission system is one of the fundamental and most i portant parts of machinery and is employed in industrial sectors worldwide. Gearbox fa ults during operation will result in serious damage. In order to monitor gearbox condition in time, a novel technique called “curvilinear component analysis” (CCA) is investigated in this paper, which performs two tasks: vector quantization of the submanifold in the input space and non-linear projec tion of these quantizing vectors toward an output space, thus providing a revealing unfolding f the submanifold. The feature data extracted from industrial gearbox vibration signa l measured under different operating conditions are analysed using the CCA method. It follows from the analytical results that with the choice of a two-dimensional output space, the CCA method project s th high-dimensional feature data into the output space and forms several clustering reg ons, each indicative of a specific gearbox condition, respectively. Therefore, the gearbox operating condition w h a fatigue crack or broken tooth compared with the normal condition is identified clearly. More ove , with the trajectory of the image points for the feature data in the two-dimensional spac e, the variation of gearbox conditions is observed visually, and the development of gearbox early-stag e failures is monitored in time.
DOI: 10.1006/jsvi.1998.2161
发表时间: 1999-06-17
影响因子: 4.7
作者:
Jiang, JD;Chen, J;Qu, LS
通讯作者: Qu, LS
DOI: --
发表时间: 1998-11
期刊: --
影响因子: --
作者:
D. Akhmetov;土手 康彦
通讯作者: D. Akhmetov;土手 康彦
DOI: --
发表时间: 2001-09
期刊: --
影响因子: --
作者:
A. Starr
通讯作者: A. Starr
DOI: 10.1115/1.3269317
发表时间: 1986-04-01
期刊: JOURNAL OF VIBRATION ACOUSTICS STRESS AND RELIABILITY IN DESIGN-TRANSACTIONS OF THE ASME
影响因子: --
作者:
MCFADDEN, PD
通讯作者: MCFADDEN, PD
DOI: 10.1007/bf00337288
发表时间: 1982-01-01
影响因子: 1.9
作者:
KOHONEN, T
通讯作者: KOHONEN, T