Multimodal Mechanism Enhanced Variational Autoencoder for Variable Speed Rolling Bearing Data Augmentation Methods

doi:10.3901/JME.260600

Abstract

Abstract: To address the issues of poor reconstruction quality and inaccurate latent space distribution in Variational Autoencoder (VAE) during data augmentation, a multi-modal mechanism enhanced VAE (MMEVAE) method for time-varying speed rolling bearing data augmentation is proposed. Firstly, the collected vibration signals are short-time Fourier transformed to obtain the time-frequency distribution matrix of the signals, and the entropy of the spectra in each time window is calculated and normalized in turn, so that the NTFES is obtained as the input of the model. Then, GMM is introduced to redefine the potential space, and the number of distributions of the model’s potential space models is evaluated and selected based on BIC and mutual information MI, so that the quality of image reconstruction and model complexity can be improved. Secondly, the SAM is embedded in the model to capture the correlation information between layers more effectively to enhance the quality of image generation. Finally, the effectiveness of the proposed method is verified using a real variable speed bearing fault simulation experiment. The experimental results show that the proposed method can generate high-quality virtual samples to improve the diagnostic accuracy.

Key words: variational autoencoder, multimodal mechanism, imbalanced data, variable speed, fault diagnosis

CLC Number:

TG156

WANG Zhenya, LIU Tao, WU Fengqi, WU Xing. Multimodal Mechanism Enhanced Variational Autoencoder for Variable Speed Rolling Bearing Data Augmentation Methods[J]. Journal of Mechanical Engineering, 2026, 62(11): 353-361.

Add to citation manager EndNote|Reference Manager|ProCite|BibTeX|RefWorks

URL: http://www.cjmenet.com.cn/EN/10.3901/JME.260600

http://www.cjmenet.com.cn/EN/Y2026/V62/I11/353

References

[1] HOU Bingchang，WANG Dong，PENG Zhike，et al. Adaptive fault components extraction by using an optimized weights spectrum based index for machinery fault diagnosis[J]. IEEE Transactions on Industrial Electronics，2024，71(1)：985-995.
[2] 黄海松，魏建安，任竹鹏，等. 基于失衡样本特性过采样算法与SVM的滚动轴承故障诊断[J]. 振动与冲击，2020，39(10)：65-74，132. HUANG Haisong，WEI Jianan，REN Zhupeng，et al. Rolling bearing fault diagnosis based on unbalance sample characteristic oversampling algorithm and SVM[J]. Journal of Vibration and Shock，2020，39(10)：65-74，132.
[3] AZAMFAR M，SINGH J，BRAVO-IMAZ I，et al. Multi sensor data fusion for gearbox fault diagnosis using 2-D convolutional neural network and motor current signature analysis[J]. Mechanical Systems and Signal Processing，2020，144(1)：108664-108688.
[4] 王振亚，刘韬，伍星. 互信息规范的卷积神经网络及其在轴承故障特征提取中的应用[J]. 机械工程学报，2024，60(12)：137-146. WANG Zhenya，LIU Tao，WU Xing. Convolutional neural network with mutual information specification and its application in bearing fault feature extraction[J]. Journal of Mechanical Engineering，2024，60(12)：137-146.
[5] WANG Youming，CHENG Liu. A combination of residual and long-short-term memory networks for bearing fault diagnosis based on time-series model analysis[J]. Measurement Science and Technology，2021，32(1)：1-10.
[6] WANG Zhenya，LIU Tao，WU Xing，et al. Application of an oversampling method based on gmm and boundary optimization in imbalance-bearing fault diagnosis[J]. IEEE Transactions on Industrial Informatics，2024，20(2)：1931-1940.
[7] 刘东东，程卫东，万广通. 基于故障特征趋势线模板的滚动轴承故障诊断[J]. 机械工程学报，2017，53(9)：83-91. LIU Dongdong，CHENG Weidong，WAN Guangtong. Fault diagnosis of rolling bearing based on fault characteristic trend line template[J]. Journal of Mechanical Engineering，2017，53(9)：83-91.
[8] 张宗振，王金瑞，韩宝坤，等. 非线性稀疏盲解卷积的轴承早期故障诊断方法[J]. 机械工程学报，2023，59(16)：157-166. ZHANG Zongzhen，WANG Jinrui，HAN Baokun，et al. Early bearing fault diagnosis based on nonlinear sparse blind deconvolution[J]. Journal of Mechanical Engineering，2023，59(16)：157-166.
[9] 邵海东，颜深，刘政武，等. 面向小样本场景的风机多增量故障诊断[J]. 机械工程学报，2024，60(8)：1-10. SHAO Haidong，YAN Shen，LIU Zhengwu，et al. For small sample of the scene fan fault diagnosis more incremental[J]. Journal of Mechanical Engineering，2024，60(8)：1-10.
[10] LIU Jie，ZHANG Changhe，JIANG Xingxing. Imbalanced fault diagnosis of rolling bearing using improved MsR-GAN and feature enhancement-driven CapsNet[J]. Mechanical Systems and Signal Processing，2022，168(1)：108644-108667.
[11] FAN S K S，TSAI D M，YEH P C. Effective variational autoencoder based generative models for highly imbalanced fault detection data in semiconductor manufacturing[J]. IEEE Transactions on Semiconductor Manufacturing，2023，36(2)：205-214.
[12] 王志颖，李天福，许文纲，等. 降噪混合注意力变分自编码器及其在轴向柱塞泵故障诊断中的应用[J]. 机械工程学报，2024，60(4)：167-177. WANG Zhiying，LI Tianfu，XU Wengang，et al. Noise reduction hybrid attention variational autoencoder and its application in fault diagnosis of axial piston pump[J]. Journal of Mechanical Engineering，2024，60(4)：167-177.
[13] LI Mengnan，LI Kun，YE Zhen，et al. Bearing fault diagnosis under small sample data based on SE-VAE and M1DCNN[J]. Mechanical Science and Technology for Aerospace Engineering，2024，43(5)：773-780.
[14] ZHAO Dongfang，LIU Shulin，MIAO Zhonghua，et al. Subdomain adaptation joint attention network enabled two-stage strategy towards few-shot fault diagnosis of LRE turbopump[J]. Advanced Engineering Informatics，2024，60(1)：102366-102380.
[15] WANG Zhenya，LIU Tao，WU X，et al. A fault diagnosis method based on NTFES-FCCT for variable working condition bearing signals[J]. IEEE Sensors Journal，2024，24(13)：20989-20998.
[16] 尹爱军，陈小敏，谭建，等. 深度概率优化的VAE轴承状态评估[J]. 振动与冲击，2021，40(20)：186-192. YIN Aijun，CHEN Xiaomin，TAN Jian，et al. VAE bearing state evaluation with deep probability optimization[J]. Journal of Vibration and Shock，2021，40(20)：186-192.
[17] YANG Dan，PENG Xin，SU Cheng，et al. Regularized wasserstein distance-based joint distribution adaptation approach for fault detection under variable working conditions[J]. IEEE Transactions on Instrumentation and Measurement，2024，73(1)：1-11.
[18] 邵海东，李伟，刘翊，等. 时变转速下基于双阈值注意力生成对抗网络和小样本的转子-轴承系统故障诊断[J]. 机械工程学报，2023，59(12)：215-24. SHAO Haidong，LI Wei，LIU Yi，et al. Fault diagnosis of rotor-bearing system based on dual threshold attention generation adversarial network and small sample under time-varying speed[J]. Journal of Mechanical Engineering，2023，59(12)：215-224.
[19] DOU Huili，LIU Yuanyuan，CHEN Sirui，et al. A hybrid CEEMD-GMM scheme for enhancing the detection of traffic flow on highways[J]. Soft Computing，2023，27(21)：16373-16388.
[20] LI Wei，ZHONG Xiang，SHAO Haidong，et al. Multi-mode data augmentation and fault diagnosis of rotating machinery using modified ACGAN designed with new framework[J]. Advanced Engineering Informatics，2022，52(1)：101552-101568.