基于HO-LSSVM的混合陶瓷轴承全寿命周期故障诊断

温金鹏; 李军宁; 罗文广; 白梦莎; 张放

doi:10.13224/j.cnki.jasp.20240545

基于HO-LSSVM的混合陶瓷轴承全寿命周期故障诊断

doi: 10.13224/j.cnki.jasp.20240545

1.
西安工业大学机电工程学院，西安 710021
2.
瓦房店轴承集团有限责任公司，辽宁瓦房店 116300

基金项目: 国家自然科学基金（51505361）；秦创原“科学家+工程师”队伍建设项目（2025QCY-KXJ-048）；陕西省教育厅服务地方专项项目（24JC045）；西安工业大学研究生教育教学研究生教学改革项目（XAGDYJ240804）

详细信息

作者简介:
李军宁（1985-），男，教授，博士，研究领域为工程摩擦学与机械系统状态监测与故障诊断。E-mail：junningli@xatu.edu.cn

中图分类号: V233.4
计量
- 文章访问数: 127
- HTML浏览量: 89
- PDF量: 13
- 被引次数: 0
出版历程
- 收稿日期: 2024-08-05
- 网络出版日期: 2026-03-11

Fault diagnosis of the Hybrid ceramic bearing in whole life cycle based on HO-LSSVM

1.
School of Mechatronic Engineering，Xi’an Technological University，Xi’an 710072，China
2.
Wafangdian Bearing Group Corporation Limited，Wafangdian Liaoning 116300，China

摘要

摘要:
针对最小二乘支持向量机分类模型在滚动轴承全寿命周期故障诊断中准确率较低的问题，提出了一种基于小波阈值去噪、白鲸优化算法-变分模态分解（beluga whale optimization- variational mode decomposition，BWO-VMD）、河马优化算法-最小二乘支持向量机（hippopotamus optimization algorithm-least squares support vector machine，HO-LSSVM）的混合陶瓷轴承全寿命周期故障诊断方法。采用小波阈值去噪方法对振动信号进行去噪处理，使用白鲸优化算法对变分模态分解的参数进行优化，使用河马优化算法优化最小二乘支持向量机参数构建故障诊断模型。基于课题组高性能滚动轴承综合性能试验台采集全寿命周期振动数据，以方均根值和峰值作为全寿命周期故障阶段划分依据，实现混合陶瓷轴承全寿命周期故障状态识别。结果表明所提方法识别准确率高于卷积神经网络等传统方法，最高可达99.38%。
- 混合陶瓷轴承 /
- 全寿命周期 /
- 振动信号 /
- 河马优化算法-最小二乘支持向量机 /
- 故障诊断
Abstract:
In order to solve the problem of low accuracy of the Least squares support vector machine (LSSVM) classification model in the whole life cycle fault diagnosis of rolling bearings, proposes a whole life cycle fault diagnosis method of hybrid ceramic bearings based on wavelet threshold denoising, beluga whale optimization-variational mode decomposition (BWO-VMD), and hippopotamus optimization algorithm-least squares support vector machine (HO-LSSVM). The wavelet threshold denoising method is adopted to denoise the vibration signal. The Beluga Whale Optimization algorithm is used to optimize the parameters of the Variational Mode Decomposition. The Hippo Optimization algorithm is used to optimize the parameters of the least squares support vector machine to construct the fault diagnosis model. Based on the whole life cycle vibration data collected by the high-performance rolling bearing comprehensive performance experimental bench of the research group, the root mean square value and peak value are used as the basis for the division of the whole life cycle fault stage, and the whole life cycle fault state identification of the hybrid ceramic bearing is realized. The results show that the recognition accuracy of the proposed method is higher than that of traditional methods such as the Convolutional Neural Network (CNN), up to 99.38%.
- hybrid ceramic bearings /
- whole life cycle /
- vibration signal /
- HO-LSSVM /
- fault diagnosis

HTML

图 1 总体诊断流程

Figure 1. Overall diagnostic process

下载: 全尺寸图片幻灯片

图 2 HO-LSSVM流程图

Figure 2. Flowchart for HO-LSSVM

下载: 全尺寸图片幻灯片

图 3 测试集分类结果

Figure 3. Classification results of the test set

下载: 全尺寸图片幻灯片

图 4 混淆矩阵分析结果

Figure 4. Results of confusion matrix analysis

下载: 全尺寸图片幻灯片

图 5 高性能航空发动机轴承综合性能试验台

Figure 5. Comprehensive performance experimental rig of high performance aeroengine bearing

下载: 全尺寸图片幻灯片

图 6 混合陶瓷轴承全寿命周期状态

Figure 6. Life cycle state of hybrid ceramic bearing

下载: 全尺寸图片幻灯片

图 7 全寿命周期方均根值和峰值变化情况

Figure 7. Changes of root-mean-square values and peak values over the whole life cycle

下载: 全尺寸图片幻灯片

图 8 D阶段原始数据与去噪数据时域图

Figure 8. Time domain diagram of raw data and denoised data in stage D

下载: 全尺寸图片幻灯片

图 9 BWO优化VMD适应度曲线

Figure 9. VMD fitness curve optimized by BWO

下载: 全尺寸图片幻灯片

图 10 HO适应度曲线

Figure 10. HO fitness curve

下载: 全尺寸图片幻灯片

图 11 测试集预测结果分析

Figure 11. Analysis of test set prediction results

下载: 全尺寸图片幻灯片

图 12 不同方法平均识别准确率对比图

Figure 12. Comparison of average identification accuracy among different methods

下载: 全尺寸图片幻灯片

表 1 不同故障程度数据集

Table 1. Data sets of different fault degrees

轴承类型	故障位置	故障直径	故障深度	标签
SKF6205 深沟球轴承	正常状态	0	0	1
	内圈滚道1	0.007	0.011	2
	内圈滚道2	0.014	0.011	3
	内圈滚道3	0.021	0.011	4

下载: 导出CSV

表 2 试验轴承基本信息

Table 2. Basic information of experimental bearings mm

轴承型号	d	D	B
NU210ECM	50	90	20

下载: 导出CSV

表 3 全寿命周期阶段划分

Table 3. Stages of the whole life cycle

阶段代号	阶段	故障程度	组数
A	正常状态	无	1～6282
B	轻微退化	轻微故障	6283～11415
C	快速退化	中度故障	11416～14806
D	失效阶段	重度故障	14807～18000

下载: 导出CSV

表 4 VMD最佳参数组合

Table 4. Best combination of VMD parameters

退化阶段	K	α	平均排列熵
A	7	4059	0.62142
B	9	5000	0.6241
C	6	4113	0.62479
D	8	5000	0.62012

下载: 导出CSV

表 5 3次测试集识别准确率

Table 5. Identification accuracy of 3 test sets

选取次数	γ	σ	测试集准确度/%	平均准确度/%
1	188.80	1.40	99.38	99.29
2	181.67	1	99.25
3	151.58	1	99.25

下载: 导出CSV

参考文献(28)

[1]	KANKAR P K, SHARMA S C, HARSHA S P. Rolling element bearing fault diagnosis using wavelet transform[J]. Neurocomputing, 2011, 74(10): 1638-1645. doi: 10.1016/j.neucom.2011.01.021
[2]	MBA D. The use of acoustic emission for estimation of bearing defect size[J]. Journal of Failure Analysis and Prevention, 2008, 8(2): 188-192. doi: 10.1007/s11668-008-9119-8
[3]	RANDALL R B, ANTONI J. Rolling element bearing diagnostics: a tutorial[J]. Mechanical Systems and Signal Processing, 2011, 25(2): 485-520. doi: 10.1016/j.ymssp.2010.07.017
[4]	FENG Zhigang, WANG Shouqi, YU Mingyue. A fault diagnosis for rolling bearing based on multilevel denoising method and improved deep residual network[J]. Digital Signal Processing, 2023, 140: 104106. doi: 10.1016/j.dsp.2023.104106
[5]	BRUSA E, DELPRETE C, GARGIULI S, et al. Screening of discrete wavelet transform parameters for the denoising of rolling bearing signals in presence of localised defects[J]. Sensors, 2023, 23(1): 8. doi: 10.3390/s23010008
[6]	LU Wenqing, ZHANG Laibin, LIANG Wei, et al. Research on a small-noise reduction method based on EMD and its application in pipeline leakage detection[J]. Journal of Loss Prevention in the Process Industries, 2016, 41: 282-293. doi: 10.1016/j.jlp.2016.02.017
[7]	GUO Hualing, BIN Z, LIU Liping, et al. Study on denoising method of surface defect signal of rail based on CEEMD and wavelet soft threshold[J]. Acoustical Physics, 2023, 69(6): 929-935. doi: 10.1134/S1063771022600504
[8]	WANG Lihua, ZHAO Xiaoping, WU Jiaxin, et al. Motor fault diagnosis based on short-time Fourier transform and convolutional neural network[J]. Chinese Journal of Mechanical Engineering, 2017, 30(6): 1357-1368. doi: 10.1007/s10033-017-0190-5
[9]	QI Bowei, LI Yuanyuan, YAO Wei, et al. Application of EMD combined with deep learning and knowledge graph in bearing fault[J]. Journal of Signal Processing Systems, 2023, 95(8): 935-954. doi: 10.1007/s11265-023-01845-z
[10]	DRAGOMIRETSKIY K, ZOSSO D. Variational mode decomposition[J]. IEEE Transactions on Signal Processing, 2014, 62(3): 531-544. doi: 10.1109/TSP.2013.2288675
[11]	LI Junning, CHEN Wuge, HAN Ka, et al. Fault diagnosis of rolling bearing based on GA-VMD and improved WOA-LSSVM[J]. IEEE Access, 2020, 8: 166753-166767. doi: 10.1109/ACCESS.2020.3023306
[12]	ZHU Xingtong, XIONG Jianbin, LIANG Qiong. Fault diagnosis of rotation machinery based on support vector machine optimized by quantum genetic algorithm[J]. IEEE Access, 2018, 6: 33583-33588. doi: 10.1109/ACCESS.2018.2789933
[13]	ZHOU Junbo, XIAO Maohua, NIU Yue, et al. Rolling bearing fault diagnosis based on WGWOA-VMD-SVM[J]. Sensors, 2022, 22(16): 6281. doi: 10.3390/s22166281
[14]	DONOHO D L. De-noising by soft-thresholding[J]. IEEE Transactions on Information Theory, 1995, 41(3): 613-627. doi: 10.1109/18.382009
[15]	LI Junning, LUO Wenguang, BAI Mengsha, et al. Fault diagnosis of high-speed rolling bearing in the whole life cycle based on improved grey wolf optimizer-least squares support vector machines[J]. Digital Signal Processing, 2024, 145: 104345. doi: 10.1016/j.dsp.2023.104345
[16]	LU Jingyi, LIN Hong, YE Dong, et al. A new wavelet threshold function and denoising application[J]. Mathematical Problems in Engineering, 2016, 2016(1): 3195492.
[17]	ZHONG Changting, LI Gang, MENG Zeng. Beluga whale optimization: a novel nature-inspired metaheuristic algorithm[J]. Knowledge-Based Systems, 2022, 251: 109215. doi: 10.1016/j.knosys.2022.109215
[18]	HEIDARI A A, PAHLAVANI P. An efficient modified grey wolf optimizer with Lévy flight for optimization tasks[J]. Applied Soft Computing, 2017, 60: 115-134. doi: 10.1016/j.asoc.2017.06.044
[19]	LI Hua, LIU Tao, WU Xing, et al. An optimized VMD method and its applications in bearing fault diagnosis[J]. Measurement, 2020, 166: 108185. doi: 10.1016/j.measurement.2020.108185
[20]	BANDT C, POMPE B. Permutation entropy: a natural complexity measure for time series[J]. Physical Review Letters, 2002, 88(17): 174102. doi: 10.1103/PhysRevLett.88.174102
[21]	YANG Yu, YUDEJIE, CHENG Junsheng. A roller bearing fault diagnosis method based on EMD energy entropy and ANN[J]. Journal of Sound and Vibration, 2006, 294(1/2): 269-277. doi: 10.1016/j.jsv.2005.11.002
[22]	AMIRI M H, MEHRABI HASHJIN N, MONTAZERI M, et al. Hippopotamus optimization algorithm: a novel nature-inspired optimization algorithm[J]. Scientific Reports, 2024, 14: 5032. doi: 10.1038/s41598-024-54910-3
[23]	SUYKENS J A K, VANDEWALLE J. Least squares support vector machine classifiers[J]. Neural Processing Letters, 1999, 9(3): 293-300. doi: 10.1023/A:1018628609742
[24]	JIN Zhihao, CHEN Guangdong, YANG Zhengxin. Rolling bearing fault diagnosis based on WOA-VMD-MPE and MPSO-LSSVM[J]. Entropy, 2022, 24(7): 927. doi: 10.3390/e24070927
[25]	SMITH W A, RANDALL R B. Rolling element bearing diagnostics using the Case Western Reserve University data: a benchmark study[J]. Mechanical Systems and Signal Processing, 2015, 64/65: 100-131.
[26]	LV Mingzhu, LIU Shixun, SU Xiaoming, et al. General log-linear weibull model combining vibration and temperature characteristics for remaining useful life prediction of rolling element bearings[J]. Shock and Vibration, 2020, 2020: 8829823. doi: 10.1155/2020/8829823
[27]	SHEN Weijie, XIAO Maohua, WANG Zhenyu, et al. Rolling bearing fault diagnosis based on support vector machine optimized by improved grey wolf algorithm[J]. Sensors, 2023, 23(14): 6645. doi: 10.3390/s23146645
[28]	ZHANG Lipin, LU Chen, TAO Laifa. Curve similarity recognition based rolling bearing degradation state estimation and lifetime prediction[J]. Journal of Vibroengineering, 2016, 18(5): 2839-2854. doi: 10.21595/jve.2016.17377