Data-driven fault detection of rotating machinery using synthetic oversampling and generative adversarial network

Zhongze Han; Chen Shen; Yong Zhang; Haoran Wang; Lianqing Yu

doi:10.1117/12.2674215

28 June 2023 Data-driven fault detection of rotating machinery using synthetic oversampling and generative adversarial network

Zhongze Han, Chen Shen, Yong Zhang, Haoran Wang, Lianqing Yu

Author Affiliations +

Proceedings Volume 12720, 2022 Workshop on Electronics Communication Engineering; 127200W (2023) https://doi.org/10.1117/12.2674215
Event: 2022 Workshop on Electronics Communication Engineering (WECE 2022), 2022, Xi'an, China

Abstract

In the fault detection of mechanical rotating parts such as gears and bearings using data-driven methods, the working condition data sets detected by sensors have the characteristic of imbalanced proportion of positive and negative samples, which makes it difficult for machine learning classification methods to accurately identify fault samples. This paper proposes a method combining Synthetic Minority Over-Sampling Technique (SMOTE) with Generative Adversarial Networks (SMOTIFIED-GAN) to pre-process the training data. The input random noise of generator in GAN is replaced by the synthetic sample of SMOTE, which makes the samples more consistent with the real distribution, to solve the imbalanced class problem, thus improve the fault diagnosis ability of the classification model. This method is applied to experimental gearbox fault datasets. The classification ability and robustness of the detection model in different imbalance rates are analysed, and the results show that SMOTIFIED-GAN improved the detection rate. Compared with the traditional methods, the F1 score of fault detection results is improved, and the feasibility of the proposed method.

Citation Download Citation

Zhongze Han, Chen Shen, Yong Zhang, Haoran Wang, and Lianqing Yu "Data-driven fault detection of rotating machinery using synthetic oversampling and generative adversarial network", Proc. SPIE 12720, 2022 Workshop on Electronics Communication Engineering, 127200W (28 June 2023); https://doi.org/10.1117/12.2674215

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available