A novel multi-adversarial cross-domain neural network for bearing fault diagnosis

Jin, Guoqiang; Xu, Kai; Chen, Huaian; Jin, Yi; Zhu, Changan

doi:10.1088/1361-6501/abd900

Cited by 25 publications

(17 citation statements)

References 46 publications

(58 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Domain adaptation is often defined as a branch of transfer learning. The source and target have the same task T s = T t and the same feature space, i.e., X s = X t , but different marginal distributions, i.e., P(X s ) = P(X t ) [53], [56], [57], [58]. However, this distinction is not consistent in the literature.…”

Section: A: Inductive Transfer Learningmentioning

confidence: 97%

“…Jin et al [162] combined AdaBN with an MMD approach (B2.1, B3), and Jin et al [58] added adversarial learning (B2.2, B3). Both considered the condition diagnosis of bearings.…”

Section: E Combined Approachesmentioning

confidence: 99%

“…Wang et al [287] conducted examinations with the CWRU and IMS data and a bearing test rig dataset provided by Jiangnan University. Xiang et al [288] used the CWRU, XJTU, and Paderborn University bearing datasets, and Jin et al [58] used the CWRU data and data from an internal test rig. Feng et al [283] also transferred knowledge between datasets from three bearing test rigs.…”

Section: A Similar Bearingsmentioning

confidence: 99%

See 2 more Smart Citations

Using Data From Similar Systems for Data-Driven Condition Diagnosis and Prognosis of Engineering Systems: A Review and an Outline of Future Research Challenges

Braig

Zeiler

2023

IEEE Access

View full text Add to dashboard Cite

Prognostics and health management (PHM) is an engineering approach dealing with the diagnosis, prognosis, and management of the health state of engineering systems. Methods that can analyze system behavior, fault conditions, and degradation are crucial for PHM applications, as they create the basis for determining, predicting, and monitoring the health of engineering systems. Data-driven methods have been proven to be suitable for automated diagnosis or prognosis due to their pattern recognition and anomaly detection abilities. Moreover, they do not require knowledge of the underlying degradation process. However, training data-driven methods usually requires a large amount of data, whose collection, cleansing, organization, and preparation are generally very time-consuming and costly. There are usually little or no run-to-failure data available at market launch, especially for new systems such as new machine generations. Nevertheless, related systems, hereinafter referred to as similar systems, often already exist, differing only in some technical characteristics. In this paper, the similar system problem is defined, and explanations of the difficulties that arise when using data from similar systems are presented. Furthermore, it is discussed why the usage of these data offers great potential for condition diagnosis and prognosis of engineering systems. An overview of data-driven methods that can be used to utilize data from similar systems is provided, and the methods that such systems already consider are highlighted. Two related research areas are identified, namely, fleet learning and transfer learning. In the paper, it is shown that similar system approaches will become an important branch of research in PHM. However, some difficulties must be overcome.INDEX TERMS Condition diagnosis, condition prognosis, data-driven methods, fleet learning, prognostics and health management (PHM), similar system approach, similar system problem, transfer learning.

show abstract

Section: A: Inductive Transfer Learningmentioning

confidence: 97%

“…Jin et al [162] combined AdaBN with an MMD approach (B2.1, B3), and Jin et al [58] added adversarial learning (B2.2, B3). Both considered the condition diagnosis of bearings.…”

Section: E Combined Approachesmentioning

confidence: 99%

Section: A Similar Bearingsmentioning

confidence: 99%

See 1 more Smart Citation

Using Data From Similar Systems for Data-Driven Condition Diagnosis and Prognosis of Engineering Systems: A Review and an Outline of Future Research Challenges

Braig

Zeiler

2023

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Guo et al [18] fused the features of time domain, frequency domain and time-frequency domain, and inputted the fused features into stacked autoencoder (SAE) network to realize bearing fault diagnosis. Jin et al [19] proposed an end-to-end multi adversarial crossdomain neural network to achieve bearing fault diagnosis between cross-domains.…”

Section: Introductionmentioning

confidence: 99%

An intelligent diagnosis method of rolling bearing based on multi-scale residual shrinkage convolutional neural network

Zhao¹,

Zhang²

2022

Meas. Sci. Technol.

View full text Add to dashboard Cite

The vibration signals of rolling bearings are affected by changing operating conditions and environmental noise, so they are characterized by multi-scale complexity. Deep residual shrinkage network can achieve bearing fault diagnosis in strong noise environment, but ignore the multi-scale complexity feature. To address this problem, we propose a multi-scale residual shrinkage convolutional neural network (MRSCNN) for fault diagnosis of rolling bearing. In this method, a multi-scale residual shrinkage layer based on multi-scale learning and a residual shrinkage block is constructed. By stacking multiple multi-scale residual shrinkage layers, the features of vibration signals are automatically learned from the input data. In addition, to establish the connection for different vibration signals and to reduce the number of parameters in the network, we design a separable convolution block using residual connections and separable convolution. By verifying the effectiveness of the method in CWRU and MFPT datasets, the results show that the method in this paper not only has good noise resistance in strong noise environments, but also has high diagnostic accuracy and good generalization performance in different load condition domains. The proposed method is compared with three other deep learning methods under the same experimental conditions, and the results show that the method in this paper is superior in rolling bearing fault diagnosis.

show abstract

“…Inspired by the generative adversarial network (GAN) [26], adversarial learning is introduced to solve the DA diagnosis problem [27,28]. By replacing the raw GAN distance with Wasserstein distance, Zhang et al [29] devised an adversarial network to extract the shared features from the source and target domains for the cross-domain fault diagnosis of bearings.…”

Section: Introductionmentioning

confidence: 99%

A game theory enhanced domain adaptation network for mechanical fault diagnosis

Sun¹,

Liu²,

Liu³

et al. 2022

Meas. Sci. Technol.

View full text Add to dashboard Cite

Transfer learning (TL) technology have been successfully applied to address the domain adaptation (DA) problem in machinery fault diagnosis. However, partial DA problem is more suitable for industrial applications, where the target data only covers a subset of the source classes, which makes it difficult to know where to transfer the target data. To overcome this problem, a novel game theory enhanced domain adaptation network (GT-DAN) is proposed in this paper. Based on different metrics, including Maximum mean discrepancy (MMD), Jensen-Shannon (JS) divergence and Wasserstein distance, three attention matrices are constructed to describe the distribution discrepancies between the source domain and the target domain, and the optimal coordination between these attention matrices is achieved by the combination weighting based on game theory to generate the optimal probability weights, which can guide to filter out the irrelevant source examples in domain adaptation. Two experiments show that the proposed GT-DAN is superior to the existing methods in the partial DA diagnosis performance.

show abstract

A novel multi-adversarial cross-domain neural network for bearing fault diagnosis

Cited by 25 publications

References 46 publications

Using Data From Similar Systems for Data-Driven Condition Diagnosis and Prognosis of Engineering Systems: A Review and an Outline of Future Research Challenges

Using Data From Similar Systems for Data-Driven Condition Diagnosis and Prognosis of Engineering Systems: A Review and an Outline of Future Research Challenges

An intelligent diagnosis method of rolling bearing based on multi-scale residual shrinkage convolutional neural network

A game theory enhanced domain adaptation network for mechanical fault diagnosis

Contact Info

Product

Resources

About