Unknown domain shift caused by the unavailability of target domain during training phase degrades the performance of intelligent fault diagnosis models in practical applications. Domain generalization (DG)-based methods have recently emerged to alleviate the influence of domain shift and improve the generalization ability of models toward invisible working conditions. However, most existing studies are conducted on multiple fully labeled source domains. Meanwhile, domain-specific information related to the variations of working conditions is often neglected during model training. Therefore, in order to realize reliable generalization fault diagnosis based on partially labeled source domains, this article proposes a contrast-assisted domain-specificity-removal network (CDSRN) to extract transferable features from domain-specificity-removal perspective. Concretely, a domain-specific feature removal branch is designed to disentangle domain-invariant features and domain-specific features, thus excavating generalized information only in domain-invariance dimension. Simultaneously, proxy-contrastive representation enhancement module is embedded to facilitate the fault class-discriminative and domain-discriminative feature learning, thereby assisting the model in further improvement of generalization capability. Experimental studies confirm the effectiveness and competitiveness of the proposed CDSRN in semi-supervised generalization fault diagnosis.

Graph-based machine fault diagnosis methods are successfully used in extracting relationship information. However, the heavy computational burden of K-nearest neighbor graph (KNNG) has limited its application. To overcome it, a fast graph construction-driven rotating machine fault diagnosis method using an edge predictor is proposed in this article. The edge predictor, pretrained on an edge connection prediction task, is designed to learn how to get a distance matrix from an initial KNNG (IKNNG). Subsequently, numerous samples are directly input to the edge predictor, obtaining the generated distance matrix and enabling fast KNNG construction. Compared to the traditional KNNG construction method, this approach outputs directly without calculating the distance matrix, significantly reducing the computational burden. The experimental results show that the performance of the proposed method is as well as existing graph data-driven methods. Furthermore, theoretical analysis reveals that the quality of the constructed KNNG is similar to the KNNG obtained by traditional distance matrix calculations, but with a significantly reduced computational load.

Rotating machinery often runs under variable working conditions, which results that the working condition of testing samples is unknown for the diagnosis model. The performance of the existed diagnosis methods trained by the samples under the known working condition will be deteriorated when they are used to diagnose the machine under an unknown working condition. The core for solving this issue is to eliminate the influence of working conditions. Inspired by this idea, a mode-decoupling autoencoder (MDAE) with two autoencoders, namely, fault-related mode (FRM) autoencoder and working condition mode (WCM) autoencoder is proposed for machinery fault diagnosis under unknown working conditions. An optimization object with reconstruction loss term, elimination loss term and classification loss term, is custom-tailored for the MDAE to ensure that the FRM autoencoder extracts the FRM and eliminates the WCM as best it can. As a result, the embedding feature extracted by the FRM autoencoder can be directly input into the classifier for the machinery fault diagnosis under unknown working conditions. Experimental results validate the superiority of MDAE in machinery fault diagnosis under unknown working conditions. Moreover, a detailed discussion is performed on the effects of model setting and the interpretability of mode decoupling of MDAE, that is, the stability of MDAE is well at a certain range and the merit of MDAE is given that the WCM autoencoder can drive the trained FRM autoencoder to eliminate the WCM guided by the knowledge of the normal samples.

Graph data-driven methods have gradually attracted attention in transfer learning-based machine fault diagnosis. However, there are still some limitations. First, feature space deviation exists in the mapping of relationship information in the source and target domains during the graph construction, bringing negative transfer and limiting constructed graph quality. Second, interpretability of relationship information during graph construction for machine fault diagnosis is lacking. In this article, a dynamic graph-driven rotating machine fault diagnosis method via adaptively updating cross-domain relationship information is proposed. A dynamic transfer graph (DTG) construction framework is developed to keep the relationship information mapping in the cross-domain consistent. Meanwhile, an improved classification loss, which consists of multiscale cross-entropy loss and multiscale domain adaptation loss, is designed to construct high-quality DTG. In addition, the working mechanism of relationship information in DTG is revealed by exploring the changes of intraclass edges, interclass edges, and cross-domain edge connections in the graphs. Experimental results demonstrate its effectiveness.

Graph data-driven machine fault diagnosis methods make success using sufficient data recently. However, in the actual industry, there are rare failure data in historical data, leading to insufficient graph representation ability and reducing diagnosis performance. In this article, a generalized graph contrastive learning (GCL) framework for few-shot machine fault diagnosis is proposed. First, spectrum features of vibration data-based samples are used to calculate Euclidean distance matrix for constructing K-nearest neighborhood graph (KNNG), whereKadjacent neighbors of each sample are connected. Avoiding excess calculation cost for graph construction, positive and negative KNNGs are constructed by changing parameterK. To make full use of few-shot samples, an unsupervised GCL subtask is set for pretraining graph deep learning model. Further, the unsupervised pretrained model is semisupervised trained using original KNNGs for outputting unlabeled nodes’ labels. The proposed method achieves 99.83%, 99.56% in bearing and gearbox dataset, respectively, and the proposed GCL framework works for different graph neural networks.

Recently, research on multisensor fault diagnosis under noisy signals has gained significant attention. Due to various degrees of external interference and differences in sensor precision, the signal quality across different channels is inconsistent. These discrepancies are often neglected by fault diagnosis models and are also difficult to capture accurately. To solve this problem, this article introduces a multiscale channel attention-driven graph dynamic fusion network for mechanical fault diagnosis. It can mine the differences in importance among channels at multiple scales and calculate the channel attention weights to enhance the node feature representation. Additionally, a graph dynamic fusion framework for multisource features is proposed to process the subgraphs in parallel, which achieves a deep-level feature fusion and enables dynamic adjustments to the fusion process based on real-time model output. With the proposed graph dynamic reconstruction module, the reliability of the feature fusion process is further improved. In the experimental part, three noise distribution scenarios were simulated to validate the robustness of the proposed method on an axial flow pump and a gearbox. The comparative analysis with various state-of-the-art models and traditional deep learning models confirms the effectiveness and superiority of the proposed method.

Rotating machinery often runs under large-speed-fluctuation (LSF) conditions, which results in severe data distribution domain shift for intelligent fault diagnosis methods. However, this challenge is rarely discussed in current studies. Hence, in this article, motivated by the active imagination of a human being, a new tool named actively imaginative data augmentation (AIDA) is constructed to solve machinery intelligent diagnosis under LSF conditions. The two adversarial training steps, namely, knowledge learning and sample imagining, are included in AIDA. In knowledge learning, a deep model is trained to learn the classification knowledge. In sample imagining, the parameters of the deep model are fixed and samples are generated via inversely training the model. As a result, diversified samples and an intelligent deep model adapting to the LSF condition are obtained by alternately carrying out the two steps. Moreover, a detailed discussion is given to interpret the process of actively imagining samples in the proposed AIDA in which some measures are designed, and the feature visualization is conducted. Experimental results show the effectiveness and superiority of AIDA in machinery diagnosis under LSF conditions, and the good performance of AIDA is due to the diversified dataset generated by changing the degrees and directions of each sample imagining.

By learning effective information from unlabeled nodes, node-level graph data-driven diagnosis methods perform better than graph-level methods. However, features of unlabeled nodes, indirectly involved in graph feature learning, are not fully utilized. To overcome aforementioned limitations, a semisupervised machine fault diagnosis fusing unsupervised graph contrastive learning (GCL) is proposed. A new GCL framework, where positive and negative graphs are generated by calculating Pearson correlation coefficient, is fused into the graph transformer network (GTN). Furthermore, a new combined loss, including a supervised cross-entropy loss and a new unsupervised GCL loss, is designed for GTN training. Contrastive learning of positive and negative graphs is guided by the unsupervised GCL loss. While the semisupervised graph feature learning for original graphs is mainly driven by the supervised cross-entropy loss, where the GTN for graph feature learning shares parameters. Experimental results on public and real datasets show the proposed method achieves a competitive performance.

Modern detectors in remote-sensing images follow the pipeline that feature maps extracted from ConvNets are shared between classification and regression tasks. However, there exist obvious conflicting demands in multiorientation object detection of remote-sensing images (RSOD) that classification is insensitive to orientations, while regression is quite sensitive. In addition, previous works cannot promise the reliability of rotation-invariant or rotation-equivariant features with only qualitative or intuitive analysis. To address these issues, we propose an encoder-encoder architecture, called rotated feature network (RFN), which produces rotation-sensitive feature maps (RS) for regression and rotation-invariant feature maps (RI) for classification. Specifically, the encoder unit assigns weights for rotated feature maps. The decoder unit extracts RS and RI by performing resuming operators on rotated and reweighed feature maps, respectively. To make the rotation-invariant characteristics more reliable, a metric is adopted to quantitatively evaluate the rotation-invariance by adding a constraint item in the loss, yielding a promising detection performance. Compared with the state-of-the-art methods, our method can achieve a significant improvement in NWPU very high resolution (VHR)-10 and RSOD data sets. The proposed RFN is further evaluated on the scene classification in remote-sensing images, demonstrating its good generalization ability. It can be integrated into an existing framework, leading to better performance with only a slight increase in test time and model complexity.

A novel adaptive sliding mode control algorithm is derived to deal with synchronized joint position tracking control of robot manipulators. The proposed algorithm does not require the precise dynamic model, and is very practical. The cross-coupled technology is incorporated into the adaptive sliding mode control architecture through feedback of joint position errors and synchronization errors. Its robustness is verified by the Lyapunov stability theory. Simulation results obtained from a 3-link non-linear planer robot manipulator demonstrate the effectiveness of the approach under various disturbances.

A novel adaptive sliding mode control (ASMC) algorithm is derived to deal with seam tracking control problem of welding robotic manipulator, during the process of large-scale structure component welding. The controllers robustness is verified by the Lyapunov stability theory, and the analytical results show that the proposed algorithm enables better high-precision tracking performance with chattering-free than classic sliding mode control (SMC) algorithm.