Deep Convolutional Transfer Learning Network: A New Method for Intelligent Fault Diagnosis of Machines With Unlabeled Data

Guo, Liang; Lei, Yaguo; Xing, Saibo; Yan, Tao; Li, Naipeng

doi:10.1109/tie.2018.2877090

Cited by 887 publications

(343 citation statements)

References 23 publications

Supporting

Mentioning

314

Contrasting

Unclassified

Order By: Relevance

“…The architecture is in according with ADDA where two different feature extractors with partially tied weights are used. In [31], Guo et al proposed a deep convolutional transfer learning network (DCTLN) for fault diagnosis on unlabeled data. In this method, a feature extractor and a health condition classifier are employed to learn class discriminative features, while a domain classifier and MMD based distribution discrepancy metrics are used to guide the feature extractor to learn domain invariant features.…”

Section: Unsupervised Domain Adaptationmentioning

confidence: 99%

Triplet Loss Guided Adversarial Domain Adaptation for Bearing Fault Diagnosis

Wang

Liu

2020

Sensors

View full text Add to dashboard Cite

Recently, deep learning methods are becomingincreasingly popular in the field of fault diagnosis and achieve great success. However, since the rotation speeds and load conditions of rotating machines are subject to change during operations, the distribution of labeled training dataset for intelligent fault diagnosis model is different from the distribution of unlabeled testing dataset, where domain shift occurs. The performance of the fault diagnosis may significantly degrade due to this domain shift problem. Unsupervised domain adaptation has been proposed to alleviate this problem by aligning the distribution between labeled source domain and unlabeled target domain. In this paper, we propose triplet loss guided adversarial domain adaptation method (TLADA) for bearing fault diagnosis by jointly aligning the data-level and class-level distribution. Data-level alignment is achieved using Wasserstein distance-based adversarial approach, and the discrepancy of distributions in feature space is further minimized at class level by the triplet loss. Unlike other center loss-based class-level alignment approaches, which hasto compute the class centers for each class and minimize the distance of same class center from different domain, the proposed TLADA method concatenates 2 mini-batches from source and target domain into a single mini-batch and imposes triplet loss to the whole mini-batch ignoring the domains. Therefore, the overhead of updating the class center is eliminated. The effectiveness of the proposed method is validated on CWRU dataset and Paderborn dataset through extensive transfer fault diagnosis experiments.

show abstract

Section: Unsupervised Domain Adaptationmentioning

confidence: 99%

Triplet Loss Guided Adversarial Domain Adaptation for Bearing Fault Diagnosis

Wang

Liu

2020

Sensors

View full text Add to dashboard Cite

show abstract

“…Recently, transfer learning, mainly the domain adaptation branch, has been applied to various fields and achieved excellent results. It has also gained extensive attention in the field of fault monitoring and diagnosis [19][20][21][22]. In powerful deep structure applications, preprocess is not required.…”

Section: Introductionmentioning

confidence: 99%

Generative Transfer Learning for Intelligent Fault Diagnosis of the Wind Turbine Gearbox

Guo

Zhang

et al. 2020

Sensors

View full text Add to dashboard Cite

Intelligent fault diagnosis algorithms based on machine learning and deep learning techniques have been widely used in industrial applications and have obtained much attention as well as achievements. In real industrial applications, working loads of machines are always changing. Hence, directly applying the traditional algorithms will cause significant degradation of performance with changing conditions. In this paper, a novel domain adaptation method, named generative transfer learning (GTL), is proposed to tackle this problem. First, raw datasets were transformed to time–frequency domain based on short-time Fourier transformation. A domain discriminator was then built to distinguish whether the data came from the source or the target domain. A target domain classification model was finally acquired by the feature extractor and the classifier. Experiments were carried out for the fault diagnosis of a wind turbine gearbox. The t-distributed stochastic neighbor embedding technique was used to visualize the output features for checking the effectiveness of the proposed algorithm in feature extraction. The results showed that the proposed GTL could improve classification rates under various working loads. Compared with other domain adaptation algorithms, the proposed method exhibited not only higher accuracy but faster convergence speed as well.

show abstract

“…The deep learning algorithm cannot effectively learn the fault data features due to the lack of diversity of fault sample data and the limited sample set. Therefore, it is difficult to obtain an accurate diagnosis model for status monitoring and fault diagnosis, which seriously restricts the development of data-driven fault diagnosis method in the industrial field [7,8].…”

Section: Introductionmentioning

confidence: 99%

Modeling of Coal Mill System Used for Fault Simulation

Boyu

Zeng

et al. 2020

Energies

View full text Add to dashboard Cite

Monitoring and diagnosis of coal mill systems are critical to the security operation of power plants. The traditional data-driven fault diagnosis methods often result in low fault recognition rate or even misjudgment due to the imbalance between fault data samples and normal data samples. In order to obtain massive fault sample data effectively, based on the analysis of primary air system, grinding mechanism and energy conversion process, a dynamic model of the coal mill system which can be used for fault simulation is established. Then, according to the mechanism of various faults, three types of faults (i.e., coal interruption, coal blockage and coal self-ignition) are simulated through the modification of model parameters. The simulation shows that the dynamic characteristic of the model is consistent with the actual object, the relative error of each output variable is less than 2.53%, and the total average relative error of all outputs is about 1.2%. The model has enough accuracy and adaptability for fault simulation, and the problem of massive fault samples acquisition can be effectively solved by the proposed method.

show abstract

Deep Convolutional Transfer Learning Network: A New Method for Intelligent Fault Diagnosis of Machines With Unlabeled Data

Cited by 887 publications

References 23 publications

Triplet Loss Guided Adversarial Domain Adaptation for Bearing Fault Diagnosis

Triplet Loss Guided Adversarial Domain Adaptation for Bearing Fault Diagnosis

Generative Transfer Learning for Intelligent Fault Diagnosis of the Wind Turbine Gearbox

Modeling of Coal Mill System Used for Fault Simulation

Contact Info

Product

Resources

About