Cross-domain fault diagnosis of bearing using improved semi-supervised meta-learning towards interference of out-of-distribution samples

被引：55

作者：

Lin, Jian ^{[1
]}

Shao, Haidong ^{[1
]}

Min, Zhishan ^{[1
]}

Luo, Jingjie ^{[1
]}

Xiao, Yiming ^{[1
]}

Yan, Shen ^{[1
]}

Zhou, Jian ^{[1
]}

机构：

[1] Hunan Univ, Coll Mech & Vehicle Engn, Changsha 410082, Peoples R China

来源：

KNOWLEDGE-BASED SYSTEMS | 2022年 / 252卷

基金：

中国国家自然科学基金;

关键词：

Cross-domain bearing fault diagnosis; Out-of-distribution samples; Label allocation strategy; Scalable distance metric function; Meta-learning; NETWORK;

D O I：

10.1016/j.knosys.2022.109493

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

The study of cross-domain semi-supervised fault diagnosis of bearings using meta-learning technique has important practical significance. However, existing methods fail to consider the interference problem caused by unlabelled out-of-distribution samples and are difficult to flexibly adapt to different diagnostic scenarios. For this purpose, a new method called improved semi-supervised meta-learning (ISSML) model is proposed in this study. First, a label allocation strategy is designed to fully utilise the information of unlabelled samples and effectively suppress the interference of out-of-distribution samples. Next, a scalable distance metric function is defined to flexibly evaluate the similarity between fault samples and efficiently extract the generic characteristics of the overall diagnostic task space. The proposed method is used to analyse experimental bearing datasets, and its superiority is demonstrated through several cross-domain few-shot scenarios and different interferences of out-of-distribution samples. (c) 2022 Elsevier B.V. All rights reserved.

引用

页数：11

共 39 条

[1] Unsupervised domain-share CNN for machine fault transfer diagnosis from steady speeds to time-varying speeds [J].

Cao, Hongru ;

Shao, Haidong ;

Zhong, Xiang ;

Deng, Qianwang ;

Yang, Xingkai ;

Xuan, Jianping .

JOURNAL OF MANUFACTURING SYSTEMS, 2022, 62 :186-198

[2] Mechatronics Equipment Performance Degradation Assessment Using Limited and Unlabeled Data [J].

Ding, Peng ;

Jia, Minping .

IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS, 2022, 18 (04) :2374-2385

[3] Meta deep learning based rotating machinery health prognostics toward few-shot prognostics [J].

Ding, Peng ;

Jia, Minping ;

Zhao, Xiaoli .

APPLIED SOFT COMPUTING, 2021, 104

[4] Statistical Alignment-Based Metagated Recurrent Unit for Cross-Domain Machinery Degradation Trend Prognostics Using Limited Data [J].

Ding, Peng ;

Jia, Minping ;

Ding, Yifei ;

Zhao, Xiaoli .

IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT, 2021, 70

[5] Intelligent Fault Diagnosis of Rotary Machines: Conditional Auxiliary Classifier GAN Coupled With Meta Learning Using Limited Data [J].

Dixit, Sonal ;

Verma, Nishchal K. ;

Ghosh, A. K. .

IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT, 2021, 70

[6] Meta-learning as a promising approach for few-shot cross-domain fault diagnosis: Algorithms, applications, and prospects [J].

Feng, Yong ;

Chen, Jinglong ;

Xie, Jingsong ;

Zhang, Tianci ;

Lv, Haixin ;

Pan, Tongyang .

KNOWLEDGE-BASED SYSTEMS, 2022, 235

[7] Semi-supervised meta-learning networks with squeeze-and-excitation attention for few-shot fault diagnosis [J].

Feng, Yong ;

Chen, Jinglong ;

Zhang, Tianci ;

He, Shuilong ;

Xu, Enyong ;

Zhou, Zitong .

ISA TRANSACTIONS, 2022, 120 :383-401

[8]

Finn C, 2017, PR MACH LEARN RES, V70

[9] An enhanced convolutional neural network with enlarged receptive fields for fault diagnosis of planetary gearboxes [J].

Han, Yan ;

Tang, Baoping ;

Deng, Lei .

COMPUTERS IN INDUSTRY, 2019, 107 :50-58

[10] Ensemble transfer CNNs driven by multi-channel signals for fault diagnosis of rotating machinery cross working conditions [J].

He Zhiyi ;

Shao Haidong ;

Zhong Xiang ;

Zhao Xianzhu .

KNOWLEDGE-BASED SYSTEMS, 2020, 207

← 1 2 3 4 →