基于PCA-SVDD的滚动轴承性能退化评估

吴义岚; 廖爱华; 丁亚琦

基于PCA-SVDD的滚动轴承性能退化评估

1.
上海工程技术大学城市轨道交通学院，上海 201620
2.
上海地铁维护保障有限公司车辆分公司，上海 200235

基金项目: 国家自然科学基金资助项目（51605274）；上海申通地铁集团有限公司科研计划资助项目（JS-KY15R024-4）

详细信息

作者简介:
吴义岚（1993−），男，在读硕士，研究方向为滚动轴承的性能退化评估.E-mail:502716969@qq.com

通讯作者:
廖爱华（1978−），女，副教授，博士，研究方向为车辆工程.E-mail:aiwa7816@126.com

中图分类号: TH 133.33
计量
- 文章访问数: 205
- HTML全文浏览量: 132
- PDF下载量: 65
- 被引次数: 0
出版历程
- 收稿日期: 2020-07-24
- 刊出日期: 2020-12-30

Evaluation of Rolling Bearing Performance Degradation Based on PCA-SVDD

WU Yilan^1
,,
LIAO Aihua^{1
, ,},
DING Yaqi²

1.
School of Urban Rail Transportation, Shanghai University of Engineering Science, Shanghai 201620, China
2.
Vehicle Branch of Shanghai Metro Maintenance Guarantee Co., Ltd., Shanghai 200235, China

摘要

摘要: 针对滚动轴承早期微弱故障难以及时发现的问题，提出一种基于主成分分析(PCA)和支持向量数据描述(SVDD)的滚动轴承性能退化评估模型. 使用主成分分析法对滚动轴承振动信号时域和频域的特征指标进行加权融合，构建一个可以有效全面描述滚动轴承运行状况的综合特征指标，将正常状态样本的综合特征指标输入SVDD模型完成评估模型的构建，通过设置健康报警阈值判定轻微故障出现时间，并采用滚动轴承全寿命试验数据进行验证. 结果表明，与以峭度指标、均方根值作为SVDD模型的特征指标输入相比，该评估模型可以更早检测到滚动轴承早期微弱故障的发生，也能更准确地描述滚动轴承整体退化程度.
- 主成分分析 /
- 支持向量数据描述 /
- 性能退化评估
Abstract: Aiming at the problem that it is difficult to detect the early weak faults of rolling bearing in time, a rolling bearing performance degradation evaluation model based on principal component analysis (PCA) and support vector data description (SVDD) was proposed. PCA method was used to weighted fusion of the characteristic indexes of rolling bearing vibration signal in time domain and frequency domain, and a comprehensive characteristic index which can effectively and comprehensively describe the operation status of rolling bearing was constructed. The comprehensive characteristic indexes of normal samples were input into the SVDD model to complete the construction of evaluation model. The occurrence time of minor fault was determined by setting health alarm threshold, and the whole life test data of the rolling bearing was used for verification. The results show that compared with the SVDD model that the kurtosis index and the root mean square value as the characteristic index input, the evaluation model can detect the occurrence of early weak faults of rolling bearing earlier, and can describe the overall degradation of rolling bearing more accurately.
- principal component analysis (PCA) /
- support vector data description (SVDD) /
- performance degradation evaluation

HTML全文

图 1 SVDD二维空间原理图

Figure 1. Schematic diagram of SVDD two-dimensional space

下载: 全尺寸图片幻灯片

图 2 性能退化评估流程

Figure 2. Evaluation process of performance degradation

下载: 全尺寸图片幻灯片

图 3 滚动轴承测试装置

Figure 3. Testing device of rolling bearing

下载: 全尺寸图片幻灯片

图 4 PCA约减融合后的特征指标

Figure 4. PCA reduced feature indicators after fusion

下载: 全尺寸图片幻灯片

图 5 以R为阈值时PCA-SVDD模型性能退化评估结果

Figure 5. Performance degradation evaluation results of PCA-SVDD model with R as the threshold

下载: 全尺寸图片幻灯片

图 6 按$ 3\sigma $法则设置阈值时PCA-SVDD模型性能退化评估结果

Figure 6. Performance degradation assessment results of PCA-SVDD model when threshold is set according to $ 3\sigma $ rule

下载: 全尺寸图片幻灯片

图 7 峭度为特征指标的性能退化评估结果

Figure 7. Performance degradation evaluation result with kurtosis index as characteristic index

下载: 全尺寸图片幻灯片

图 8 均方根值为特征指标的性能退化评估结果

Figure 8. Performance degradation evaluation results with root mean square value as characteristic index

下载: 全尺寸图片幻灯片

图 9 第200组数据的包络谱

Figure 9. Envelope spectrum of the 200th group data

下载: 全尺寸图片幻灯片

图 10 第574组数据的包络谱

Figure 10. Envelope spectrum of the 574th group data

下载: 全尺寸图片幻灯片

参考文献(18)

[1]	LOUTAS T H, ROULIAS D, GEORGOULAS G. Remaining useful life estimation in rolling bearings utilizing data-driven probabilistic e-support vectors regression[J] . IEEE Transactions on Reliability,2013,62(4):821 − 832. doi: 10.1109/TR.2013.2285318
[2]	李超, 郭瑜. 基于SK等指标和SVM的滚动轴承性能退化评估研究[J] . 电子科技,2020,33(1):6 − 12.
[3]	王恒, 周易文, 季云, 等. 结合狄利克雷过程和连续隐马尔科夫模型的滚动轴承性能退化评估[J] . 吉林大学学报(工学版),2019,49(1):117 − 123.
[4]	任利娟. 滚动轴承性能退化评估与剩余寿命预测[D]. 济南: 山东大学, 2019.
[5]	杨潇谊, 吴建德, 马军. 基于散布熵和余弦欧氏距离的滚动轴承性能退化评估方法[J] . 电子测量与仪器学报,2020,34(7):15 − 24.
[6]	TAX D M J, DUIN R P W. Support vector data description[J] . Machine Learning,2004(54):45 − 66.
[7]	李勇发, 左小清, 杨芳, 等. 基于小波奇异谱及SVDD的轴承故障检测方法[J] . 轴承,2016(8):46 − 49. doi: 10.3969/j.issn.1000-3762.2016.08.014
[8]	王发令, 周建民, 张臣臣, 等. 基于AR能量比和SVDD的滚动轴承性能退化评估[J] . 机床与液压,2020,48(12):103 − 111. doi: 10.3969/j.issn.1001-3881.2020.12.015
[9]	王立祥, 廖爱华, 丁亚琦. 基于比例故障率模型的转向架牵引电机滚动轴承可靠性评估[J] . 测控技术,2018,37(1):14 − 18. doi: 10.3969/j.issn.1000-8829.2018.01.004
[10]	孟文俊, 张四聪, 淡紫嫣, 等. 滚动轴承寿命动态预测新方法[J] . 振动. 测试与诊断,2019,39(3):652 − 658; 678.
[11]	郭伟超, 赵怀山, 李成, 等. 基于小波包能量谱与主成分分析的轴承故障特征增强诊断方法[J] . 兵工学报,2019,40(11):2370 − 2377. doi: 10.3969/j.issn.1000-1093.2019.11.022
[12]	LIN T, CHEN G, OUYANG W L, et al. Hyper-Spherical distance discrimination: A novel data description method for aero-engine rolling bearing fault detection[J] . Mechanical Systems & Signal Processing,2018,109:330 − 351.
[13]	WANG B X, PAN H X, YANG W. Robust bearing degradation assessment method based on improved CVA[J]. IET Science, Measurement & Technology, 2017, 11(5): 637-645.
[14]	肖顺根, 马善红, 宋萌萌, 等. 基于 EEMD 和 PCA 滚动轴承性能退化指标的提取方法[J] . 江南大学学报(自然科学版),2015,14(5):572 − 579.
[15]	QIU H, LEE J, LIN J, et al. Wavelet filter-based weak signature detection method and its application on rolling element bearing prognostics[J] . Journal of Sound and Vibration,2006,289(4/5):1066 − 1090.
[16]	杨国安. 滚动轴承故障诊断实用技术[M]. 北京: 中国石化出版社, 2012
[17]	田福庆, 罗荣, 贾兰俊, 等. 机械故障非平稳特征提取方法及其应用[M]. 北京: 国防工业出版社, 2014.
[18]	徐清瑶. 基于支持向量数据描述的滚动轴承性能退化评估[D]. 南昌: 华东交通大学, 2015.