In real industries, there often exist application scenarios where the target domain holds fault categories never observed in the source domain, which is an open-set domain adaptation (DA) diagnosis issue. Existing DA diagnosis methods under the assumption of sharing identical label space across domains fail to work. What is more, labeled samples can be collected from different sources, where multisource information fusion is rarely considered. To handle this issue, a multisource open-set DA diagnosis approach is developed. Specifically, multisource domain data of different operation conditions sharing partial classes are adopted to take advantage of fault information. Then, an open-set DA network is constructed to mitigate the domain gap across domains. Finally, a weighting learning strategy is introduced to adaptively weigh the importance on feature distribution alignment between known class and unknown class samples. Extensive experiments suggest that the proposed approach can substantially boost the performance of open-set diagnosis issues and outperform existing diagnosis approaches.

A Multi-Source Weighted Deep Transfer Network for Open-Set Fault Diagnosis of Rotary Machinery

Various domain adaptation (DA) methods have been proposed to address distribution discrepancy and knowledge transfer between the source and target domains. However, many DA models focus on matching the marginal distributions of two domains and cannot satisfy fault-diagnosed-task requirements. To enhance the ability of DA, a new DA mechanism, called deep joint distribution alignment (DJDA), is proposed to simultaneously reduce the discrepancy in marginal and conditional distributions between two domains. A new statistical metric that can align the means and covariances of two domains is designed to match the marginal distributions of the source and target domains. To align the class conditional distributions, a Gaussian mixture model is used to obtain the distribution of each category in the target domain. Then, the conditional distributions of the source domain are computed via maximum-likelihood estimation, and information entropy and Wasserstein distance are employed to reduce class conditional distribution discrepancy between the two domains. With joint distribution alignment, DJDA can achieve domain confusion to the highest degree. DJDA is applied to the fault transfer diagnosis of a wind turbine gearbox and cross-bearing with unlabeled target-domain samples. Experimental results verify that DJDA outperforms other typical DA models.

Deep Joint Distribution Alignment: A Novel Enhanced-Domain Adaptation Mechanism for Fault Transfer Diagnosis

Most of domain adaptation (DA) methods just focus on the case that the label spaces of two domains are identical. However, it is more valuable for studying partial DA in practical engineering. Whereupon, a novel multiscale weight-selection adversarial network (MWSAN) is proposed to enhance the effect of partial DA. MWSAN is constructed via the designed multiscale domain adversarial network (MDAN) and the multiscale weight-selection mechanism of instance and class. MDAN is innovatively built to avoid the overfitting of a single classifier and strengthen the domain confusion. For the weight-selection of instance, the instance weights are directly obtained by the probability output of MDAN. For the weight-selection of class, the possible Gaussian distributions of target-domain samples is estimated by Gaussian mixture model, and the Gaussian distribution of each class of source-domain samples is computed by maximum likelihood estimation, then the Wasserstein distance is employed to computed the class weights. With the multiscale backbone network and weighting mechanism, MWSAN can achieve the partial DA to the larger degree. The proposed method is applied to the partial transfer fault diagnosis of planetary gearboxes with the unlabeled target-domain samples, and the experimental results indicate that MWSAN outperforms other typical DA methods. 

Partial Transfer Fault Diagnosis by Multiscale Weight-Selection Adversarial Network

Most of domain adaptation (DA) methods just focus on the case that the label spaces of two domains are identical. However, it is more valuable for studying partial DA in practical engineering. Whereupon, a novel multiscale weight-selection adversarial network (MWSAN) is proposed to enhance the effect of partial DA. MWSAN is constructed via the designed multiscale domain adversarial network (MDAN) and the multiscale weight-selection mechanism of instance and class. MDAN is innovatively built to avoid the overfitting of a single classifier and strengthen the domain confusion. For the weight-selection of instance, the instance weights are directly obtained by the probability output of MDAN. For the weight-selection of class, the possible Gaussian distributions of target-domain samples is estimated by Gaussian mixture model, and the Gaussian distribution of each class of source-domain samples is computed by maximum likelihood estimation, then the Wasserstein distance is employed to computed the class weights. With the multiscale backbone network and weighting mechanism, MWSAN can achieve the partial DA to the larger degree. The proposed method is applied to the partial transfer fault diagnosis of planetary gearboxes with the unlabeled target-domain samples, and the experimental results indicate that MWSAN outperforms other typical DA methods.

Feature distance-based deep prototype network for few-shot fault diagnosis under open-set domain adaptation scenario

Many deep-learning methods have been developed for fault diagnosis. However, due to the difficulty of collecting and labeling machine fault data, the datasets in some practical applications are relatively much smaller than the other big data benchmarks. In addition, the fault data come from different machines. Therefore, on some occasions, fault diagnosis is a multidomain problem with small data, where satisfactory transfer performance is difficult to obtain and has been rarely explored from the few-shot learning viewpoint. Different from the existing deep transfer learning solutions, a novel transfer relation network (TRN), combining a few-shot learning mechanism and transfer learning, is developed in this study. Specifically, the fault diagnosis problem has been treated as a similarity metric-learning problem instead of solely feature weighted classification. A feature net and a relation net have been, respectively, constructed for feature extraction and relation computation. The Siamese structure has been borrowed to extract the features of the source and the target domain samples with shared weights. Multikernel maximum mean discrepancy (MK-MMD) is employed on several higher layers with different tradeoff parameters to enable an efficient domain feature transfer considering different feature properties. To implement efficient diagnosis based on small data, an episode-based few-shot training strategy is adopted to train TRN. Average pooling has been adopted to suppress the noise influence from the vibration sequence which turns out to be important for the success of time sequence-based fault diagnosis. Transfer experiments on four datasets have verified the superior performance of TRN. A significant improvement of classification accuracy has been made compared with the state-of-the-art methods on the adopted datasets.

Huiyang Hu

Papers

Transfer Relation Network for Fault Diagnosis of Rotating Machinery With Small Data.