LOSGAN: latent optimized stable GAN for intelligent fault diagnosis with limited data in rotating machinery

Many recent studies have focused on imbalanced rolling bearing data for fault diagnosis. Complementing the imbalance dataset through data augmentation methods excellently solves this problem superior. In this paper, a patch variational autoencoding generative adversarial network (PVAEGAN) is proposed. Firstly, overlap sampling is designed to preprocess the input samples to alleviate noise interference. Secondly, the PVAEGAN is constructed, and the matrix discriminative output of the model allows it to focus on more features of the data during training. Thirdly, a stability-enhancing structure is designed for PVAEGAN to improve the stability of network parameter variations and inter-network stability for better model results. Furthermore, to verify the use of the multi-class comparison method, experiments are conducted. The results indicate that PVAEGAN can augment imbalanced datasets more effectively and with better robustness than other existing models.

Section: Introductionmentioning

confidence: 99%

Data-augmented patch variational autoencoding generative adversarial networks for rolling bearing fault diagnosis

Wang¹,

Jiang²,

Liu³

et al. 2023

“…Nowadays, in the context of industrial big data, the rapid development of artificial intelligence and machine learning is gradaully making fault diagnosis intelligent. Data-driven intelligent fault diagnosis algorithms are rapidly becoming a research hotspot [2]. In recent years in particular, neural networks represented by deep learning (DL) algorithms have have been very successful in fault diagnosis [3].…”

Section: Introductionmentioning

confidence: 99%

Research on bearing fault diagnosis method based on transformer neural network

Yang

Cen

Liu

et al. 2022

Attention mechanism (AM) has been widely used for fault diagnosis and health identification in industrial equipment. The existing researches have only used AM in combination with deep networks, or to replace certain components of these deep networks. This reliance on deep networks severely limits the feature extraction capability of AM. In this paper, a bearing fault diagnosis method based on signal Transformer neural network (SiT) with pure AM is proposed. Firstly, the raw one-dimensional vibration time-series signal is segmented and a new segmented learning strategy is introduced. Secondly, linear encoding and position encoding are performed on the segmented subsequences. Finally, the encoded subsequence is fed to the Transformer for feature extraction to achieve fault identification. The validity of the proposed method is verified using Case Western Reserve University (CWRU) dataset and self-priming centrifugal pump bearing dataset. Compared with other existing methods, the proposed method still achieves the highest average diagnostic accuracy without any data preprocessing. The results demonstrate that the proposed SiT based on pure AM can extract features and identify faults from the raw vibration signal, and has superior diagnostic performance.

“…They introduced variational auto-encoder into the fault diagnosis framework to realize data amplification by vibration signal generation, then an enhanced fault diagnosis approach is proposed by combining with a convolution neural network. Liu et al [20] developed an intelligent fault diagnosis method for rotating machinery under small sample size conditions, i.e. a multi-module learning method based on the latent optimized stable generative adversarial (LOSGAN), which combines the data-generation capability of a GAN with the powerful feature-extraction capability of a deep residual network.…”

Section: Introductionmentioning

confidence: 99%

Generative adversarial network and transfer-learning-based fault detection for rotating machinery with imbalanced data condition

Liu

2022

Intelligent fault diagnosis achieves tremendous success in machine fault diagnosis because of the outstanding data-driven capability. However, the severe imbalanced dataset in practical scenarios of industrial rotating machinery is still a big challenge for the development of intelligent fault diagnosis method. In this paper, we solve this issue by constructing a novel deep learning model incorporated with a transfer learning method based on the Time-GAN and Efficient-Net models. Firstly, the proposed model so called Time-GAN-TL extends the imbalanced fault diagnosis of rolling bearings by using time series generative adversarial network. Secondly, balanced vibration signals are converted into two-dimensional images for training and classification by implementing the Efficient-Net into the transfer learning method. Finally, the proposed method is validated using two-types of rolling bearing experimental data. The high-precision diagnosis results of the transfer learning experiments and the comparison with other representative fault diagnosis classification methods reveal the efficiency, reliability, and generalization performance of the presented model.