Industrial time series prediction (ITSP) is critical to the predictive maintenance system of modern industry. However, time-varying conditions and complex industrial processes cause the distribution drift of industrial time series, raising the difficulty of prediction. This article proposes an ITSP model considering distribution information, namely MDEformer. First, the multivariate dynamic embedding (MDE) is designed to provide the property of the channel-binding dynamic distribution awareness. Specifically, a dynamic mode transition and selection module is adopted to exploit dynamic distribution features of time series, and the bidirectional dynamic residual connection integrates dynamic distribution information into embedding vectors to filter distribution change interference. Then, the vanilla Transformer encoder is used to achieve multivariate prediction. Finally, a generative pretraining and fine-tuning strategy is used to enhance the generalization ability in real production scenarios. Extensive results on a real-world zinc smelting dataset illustrate the superiority of MDEformer.

Industrial time series prediction (ITSP) is an indispensable part of predictive control in modern industry. Recently, supervised deep learning-based methods have provided solutions with sufficient annotated data. However, there is massive unlabeled data with complex temporal features in modern industrial production, resulting in poor performance of these methods. To address this problem, a self-supervised generative pre-trained model with a learnable mask network (SSGPM-LMN) is proposed in this paper. First, the multivariate time series are made into patches channel-independently. Then, these patches are fed into a Transformer encoder with the learnable mask-reconstruction paradigm, drawing mask indices with high temporal features by calculating the cosine similarity in low-dimensional feature space to better learn general representations. Furthermore, a two-step fine-tuning strategy, including linear probing and full fine-tuning, is adopted for various downstream scenarios. Finally, extensive experimental results on case studies of ITSP and transfer learning indicate that our SSGPM-LMN achieves superior performance.

At present, data-driven fault diagnosis methods have made excellent achievements. In industrial scenarios, it is difficult to obtain sufficient amount of fault data, which means intelligent fault diagnosis is often faced with imbalanced data problem. Moreover, the label noise is usually brought due to manual recording errors so as to seriously affect the diagnosis performance. To address these problems, this paper proposed a safe-domain generative adversarial network with Transformer (SDGAN). A safe domain selecting method is used to remove the noisy samples and construct a pure dataset which poses no risk to the training process of GAN. Therefore, GAN is able to generate high-quality minority samples to balance the original dataset. In addition, the Vision Transformer (ViT) is also applied as a classifier to recognize the global information for each fault sample and achieve high diagnostic accuracy. The experimental results show that SDGAN achieves great diagnosis performance on various imbalanced ratios and noise ratios cases. Furthermore, SDGAN outperforms other baseline methods on imbalanced fault diagnosis with label noise, which indicates that the SDGAN can effectively solve real-world industrial problems.

Mechanical fault diagnosis is crucial to ensure the safe operations of equipment in intelligent manufacturing systems. Deep learning-based methods have been recently developed for fault diagnosis due to their advantages in feature representation. However, most of these methods fail to learn relations between samples and thus perform poorly without sufficient labeled data. In this article, we propose a new few-shot learning method named dual graph neural network (DGNNet) with residual blocks to address fault diagnosis problems with limited data. First, the residual module learns the feature of samples with image data transferred from original signals. Second, two complete graphs built on the sample features are used to extract the instance-level and distribution-level relations between samples. In particular, an alternate update policy between the instance and distribution graphs integrates the multilevel relations to propagate the label information of a few labeled samples to unlabeled samples. This technique leverages labeled and unlabeled samples to identify unseen faults, encouraging DGNNet competency in fault diagnosis tasks with very few labeled samples. Extensive results on various datasets show that DGNNet achieves excellent performance in supervised fault diagnosis tasks and outperforms baselines by a great margin in semisupervised cases.

Product backorder is a common problem in supply chain management systems. It is essential for entrepreneurs to predict the likelihood of backorder accurately to minimize a company’s losses. However, existing methods are hard to achieve satisfactory results since the number of backorders and non-backorders are extremely imbalanced. Besides, the backorder data’s attributes are complex to oversample them effectively. To address these problems, a novel hybrid sampling method is proposed to help predict extremely imbalanced backorder. The Randomized Undersampling (RUS) and a Conditional Wasserstein Generative Adversarial Network (CWGAN) are innovatively introduced into backorder prediction. First, RUS is used to reduce the majority non-backorder samples. Second, CWGAN is served as an oversampling technique to generate high-quality backorder samples. It utilizes unique structures in the generator and the discriminator to effectively model both numerical and categorical variables. Finally, the training dataset is balanced, and the Random Forest Classifier (RFC) is adopted to make backordering prediction. In the experiments of Kaggle’s dataset ‘Can you predict product backorder?’, our proposed method is superior to all benchmark methods in terms of standard evaluation metrics. The results show that our proposed product backorder prediction model is effective.

In order to improve fault forecasting model accuracy of back propagation neural network (BPNN), an improved prediction method of optimized BPNN based on Multilevel Genetic Algorithm (MGA) was proposed. We design new chromosome with multilevel structure, improve the encoding mode, fitness function and genetic operator. Which can optimizes the initial values of weights, thresholds and the structure of BPNN synchronously. Enhancing the ability of nonlinear learning and generalization of BPNN. Case study of continuous casting equipment verified that the proposed model with higher prediction accuracy is better than classical BPNN and GA-BPNN prediction method for fault prediction.