Search

Data-driven fault diagnosis techniques are frequently applied to ensure the reliability and safety of industrial systems. However, as a common challenge, the class-imbalance problem reduces the performance of data-driven methods due to the lack of data information. We propose a weighted modified conditional variational auto-encoder (WM-CVAE) as a novel data augmentation technique to tackle the issue. The modified structure can alleviate the existing Kullback–Leibler (KL) divergence vanishing by an adaptive loss. Meanwhile, kernel mean matching (KMM) is proposed on weight computation to reduce the negative effect of dissimilar generated samples. Constructing the WM-CVAE data augmentation framework can effectively improve the data quality and learning capability in class-imbalance fault diagnosis. To validate the proposed WM-CVAE model, three real-world industrial datasets are used as study objects, and the random forest is used as the base learner in the fault classification tasks. The diagnostic results demonstrate that the proposed WM-CVAE data augmentation framework can improve learning results in class-imbalance fault diagnosis. • A general adaptive loss-based learning algorithm to solve the KL divergence vanishing problem widely existing in the original CVAE generators. • An adaptive weight case is proposed as an example to be analyzed and validated on real-world class-imbalance fault diagnosis datasets. • Kernel mean matching is applied to realize weight assignment by metric learning. • Research results show the WM-CVAE data augmentation approach is excellent for class-imbalance fault diagnosis tasks. [ABSTRACT FROM AUTHOR]

Accurate prediction of the remaining useful life (RUL) of lithium-ion battery is of great significance for the reliability of electronic equipment. In the conventional approaches, there are notable challenges in the RUL prediction for a single battery lacking historical data. To predict the battery's RUL under the condition of unlabeled small sample data and to describe the uncertainty of the parameter estimation in the degradation model, a novel adaptive approach based on Kalman filter and expectation maximum with Rauch–Tung–Striebel (KF-EM-RTS) is proposed to predict the battery's RUL. Specifically, without RUL labels and offline training, an online KF adaptive-update model based on the Wiener process is proposed for a single battery, in which the uncertainty of parameter estimation is described. Furthermore, the unknown model parameters can be adaptively estimated using EM-RTS to overcome the constraints of strong Markov characteristics, the convergence of which is proved. The real-world battery dataset provided by NASA Ames research center is applied to verify the proposed RUL prediction approach. Experimental results show that the proposed approach outperforms the existing conventional data-driven approaches for predicting the battery's RUL. • The KF-EM-RTS is proposed to predict the battery's RUL under the situation of unlabeled small sample data. • The unknown parameters in the dynamic degradation model can be adaptively estimated. • The uncertainty of the parameter estimation is described by a probability density function. • The optimal unknown model parameters are obtained with solid theoretical foundation. [ABSTRACT FROM AUTHOR]

Prediction of remaining useful life (RUL) is of vital significance in the prognostics health management (PHM) tasks. To deal with the reverse time series and to reflect the difference in RUL prediction results at different time instances, this paper proposes a novel bidirectional gated recurrent unit with temporal self-attention mechanism (BiGRU-TSAM) to predict RUL. Specifically, a novel approach is proposed where each of the considered time instance is assigned a self-learned weight according to the degree of significance. Furthermore, the parameter update process of the TSAM is obtained with solid theoretical foundation, and as a sign of interpretability, it is shown that the assigned weights can remain consistency over several independent training processes. On this basis, the BiGRU-TSAM is applied to predict RUL online. An aircraft turbofan engine dataset and a milling dataset are applied to verify the proposed RUL prediction approach. The experimental results show the superiority of the proposed approach over the existing ones based on machine learning and deep learning. • It is proposed to predict RUL with the aid of a novel BiGRU-TSAM network. • Each of the considered time instance is assigned a self-learned weight. • The parameter update process of the TSAM layer is obtained. • The assigned weights can remain consistency over several independent training processes. [ABSTRACT FROM AUTHOR]

With the frequent occurrence of safety incidents in cyber–physical systems (CPSs), great significance has been attached to the study of defense schemes against cyber–physical attacks. In this paper, an integrated data-driven defense scheme is proposed, which can sensitively detect data integrity attacks such as false data injection (FDI) attacks, denial-of-service (DoS) attacks, and replay attacks, and ensures secure transmission against eavesdropping attacks. Specifically, a novel deep learning model is designed so that both the online detection task and the encryption/decryption task can be completed under the same framework. The main idea is inspired by denoising auto-encoders whereas necessary changes are made to adapt to the challenges in the context of CPS attacks, and in light of this, the proposed approach is called modified denoising auto-encoder (MDAE). Unlike supervised classifier-based detectors, the proposed detector can retain sensitivity to unknown attacks because it is trained to learn the normal operation behavior. Moreover, to improve the detectability of the DoS and replay attacks on all data, the check code is designed. Encrypting the transmitted data through nonlinear mapping is achieved using the same MDAE, which prevents the attackers from recording useful information. Benefiting from the fact that the dimension of the variables is reduced after encryption, the transmission traffic can be saved. Simulation results on the measurement data instances generated by the IEEE 118-bus system validate the encryption effects and detection accuracy of the proposed scheme and show the superiority by comparison study. • A modified denoising autoencoder-based detector with check code is designed. • Detecting different types of data integrity attacks with unknown patterns. • An encryption scheme to prevent attackers from getting useful information. • Reducing network transmission traffic and errors caused by noise. [ABSTRACT FROM AUTHOR]

Domain adaptation has been widely applied in data-driven fault diagnosis tasks to address the domain shift problem between source and target data. However, conventional domain adaptation methods require both domains to be known, which is not always feasible due to privacy concerns and big data transmission. To overcome this limitation, a dedicated method called source-free domain adaptation (SFDA) has been developed to ensure reliable performance without relying on source data during target model adaptation. SFDA can achieve accurate classification tasks under domain shift problems and source data-free scenarios. We propose a generalized source model with manifold Mixup data augmentation and label smoothing techniques to avoid overfitting during the source model training. Based on this model, a novel self-training framework is proposed to implement the domain adaptation task and achieve accurate prediction performance. The experimental results from three real-world datasets demonstrate the effectiveness of the proposed approach. • A novel source data-free domain adaptation approach to enable fault diagnosis in industrial systems. • A well-designed manifold Mixup module is internal to the model and trained jointly during source training. • A self-training strategy is applied with pseudo-labels and uncertainty reduction in target model adaptation. • Experimental results show the proposed method is excellent for fault diagnosis tasks under domain shift conditions. [ABSTRACT FROM AUTHOR]

Most supervised learning-based approaches follow the assumptions that offline data and online data must obey a similar distribution, which is difficult to satisfy in realistic remaining useful life (RUL) prediction. To solve the problem, domain adaptation (DA) learning-oriented transfer learning (TL) was proposed. Nevertheless, only adopting a conventional global DA approach may confuse the fine-grained features between subdomains represented by different degenerate stages. Consequently, a novel variational auto-encoder-long–short-term memory network-local weighted deep sub-domain adaptation network (VLSTM-LWSAN) is proposed for RUL prediction. Specifically, the input data are compressed into the interpretable latent space, from which the fine-grained features between subdomains are local alignment through local weighted deep sub-domain adaptation network. In this sense, the discrepancy between the unlabeled target domain and the source domain is decreased. The proposed VLSTM-LWSAN is verified by an aircraft turbofan engine dataset. The research results represent that the VLSTM-LWSAN outperforms some deep learning approaches without transfer learning and conventional transfer learning approaches. • A VLSTM is used to compress the input data into an interpretable latent space. • An LWSAN is designed to capture the fine-grained information in degradation process. • A visual analysis of the interpretability latent space is conducted to study LWSAN. • Research results show the VLSTM-LWSAN-based RUL prediction approach is excellent. [ABSTRACT FROM AUTHOR]