For imbalanced bearing fault diagnosis, GAN (Generative Adversarial Networks) is a common data augmentation approach. Nevertheless, current GAN-based methods cannot update generator from time-frequency domain simultaneously, downgrading the authenticity of signal time-frequency character. In this paper, FTGAN (Fourier-like Transform GAN), a novel GAN method, is proposed by introducing Fourier-like transformer (FLT) based on autoencoder (AE) to improve synthetic data quality. FLT approximates discrete Fourier transform (DFT) by the neural network, learning a universal map from time to frequency domain during training. FTGAN with FLT can decouple input into time-frequency domain, fitting the distribution of time and frequency of data simultaneously. Multi-domain distribution is manipulated in FTGAN without introducing additional signal transformation means. Furthermore, TRTS and TSTR analysis of one-dimensional data is introduced to evaluate data quality. Real and synthetic data are applied as training or test sets of diagnostic classifiers by turns so that data quality can be analysed through diagnosis results. Experiment results show that the proposed method can generate bearing fault signals closer to real data in the time and frequency domain, effectively improving the performance under an imbalanced dataset.

Industrial oscillation recordings are often corrupted by underlying nonstationary trend and noisy artifacts, which can occlude features of interest and complicate subsequent oscillation detection and diagnosis. However, there are considerably fewer techniques available in the literature for systematically removing both trend and noise terms from the oscillation measurements. To cater for the general detrending and denoising needs of oscillatory signals, this paper proposes an integrated framework featuring the following steps: (i) The ensemble empirical mode decomposition is first adopted to decompose the industrial single-loop data into several Intrinsic Mode Functions (IMFs). (ii) To eliminate the trend term, a surrogate-based nonstationarity testing algorithm is implemented to automatically identify and remove the requisite IMFs. (iii) By applying canonical correlation analysis on the remained IMFs, the noise-dependent components can be further isolated, which finally yields the detrended and denoised oscillation data. We conducted performance comparison study through extensive simulations and industrial examples. The results demonstrate that the proposed work is a promising tool for industrial oscillation data preprocessing.

Vigor identification in sweet corn seeds is important for seed germination, crop yield, and quality. In this study, hyperspectral image (HSI) technology integrated with germination tests was applied for feature association analysis and germination performance prediction of sweet corn seeds. In this study, 89 sweet corn seeds (73 for training and the other 16 for testing) were studied and hyperspectral imaging at the spectral range of 400–1000 nm was applied as a nondestructive and accurate technique to identify seed vigor. The root length and seedling length which represent the seed vigor were measured, and principal component regression (PCR), partial least squares (PLS), and kernel principal component regression (KPCR) were used to establish the regression relationship between the hyperspectral feature of seeds and the germination results. Specifically, the relevant characteristic band associated with seed vigor based on the highest correlation coefficient (HCC) was constructed for optimal wavelength selection. The hyperspectral data features were selected by genetic algorithm (GA), successive projections algorithm (SPA), and HCC. The results indicated that the hyperspectral data features obtained based on the HCC method have better prediction results on the seedling length and root length than SPA and GA. By comparing the regression results of KPCR, PCR, and PLS, it can be concluded that the hyperspectral method can predict the root length with a correlation coefficient of 0.7805. The prediction results of different feature selection and regression algorithms for the seedling length were up to 0.6074. The results indicated that, based on hyperspectral technology, the prediction of seedling root length was better than that of seed length.

Oscillation detection has been a hot research topic in industries due to the high incidence of oscillation loops and their negative impact on plant profitability. Although numerous automatic detection techniques have been proposed, most of them can only address part of the practical difficulties. An oscillation is heuristically defined as a visually apparent periodic variation. However, manual visual inspection is labor-intensive and prone to missed detection. Convolutional neural networks (CNNs), inspired by animal visual systems, have been raised with powerful feature extraction capabilities. In this work, an exploration of the typical CNN models for visual oscillation detection is performed. Specifically, we tested MobileNet-V1, ShuffleNet-V2, EfficientNet-B0, and GhostNet models, and found that such a visual framework is well-suited for oscillation detection. The feasibility and validity of this framework are verified utilizing extensive numerical and industrial cases. Compared with state-of-the-art oscillation detectors, the suggested framework is more straightforward and more robust to noise and mean-nonstationarity. In addition, this framework generalizes well and is capable of handling features that are not present in the training data, such as multiple oscillations and outliers.

The detection and diagnosis of oscillation are of great importance to maintain the control performance of the process plant. Even though some algorithms based on time-frequency analysis for multiple oscillations diagnosis have been reported in the literature, their performance is severely constrained by the mode-mixing and mode-splitting problems. The presence of these issues leads not only to incorrectly detected number of oscillations, but also to incorrect identification of the oscillation sources. Motivated by the above challenge, this work proposes a novel adaptive time-frequency tool: median complementary ensemble empirical mode decomposition (MCEEMD) to construct a robust and automated framework for oscillation detection and diagnosis. In addition to detecting and diagnosing various single/multiple oscillations, the proposed method can distinguish oscillations with linear and nonlinear causes. Moreover, benefiting from the robustness of MCEEMD to mode mixing and mode splitting, the framework offers better accuracy and stability compared to state-of-the-art methods. A series of simulations and industrial cases demonstrate the effectiveness and superiority of the proposed framework.

Methods for the accurate prediction of icing loads in overhead transmission lines have become an important research topic for electrical power systems as they are necessary for ensuring the safety and stability of power-grid operations. Current machine learning models for the prediction of icing loads on transmission lines are afflicted by the following issues: insufficient prediction accuracy, high randomity in the selection of kernel functions and model parameters, and a lack of generalizability. To address these issues, we propose a field data–driven online prediction model for icing loads on transmission lines. First, the effects of the type of kernel function used in the support vector regression algorithm on the prediction accuracy of the model were analyzed using micrometeorological data and icing data collected by on-site monitoring systems. The particle swarm optimization algorithm was then used to optimize and determine the model parameters such as penalty coefficients. An offline support vector regression prediction model was thus constructed. Using the accurate online support vector regression algorithm, the weighting coefficients of the samples were dynamically adjusted to satisfy the Karush–Kuhn–Tucker conditions, which allowed online updates to be made to the regression function and prediction model. Finally, a simulation analysis was performed using actual icing incidents that occurred in a transmission line of the Yunnan Power Grid, which demonstrated that our model can make online predictions for the icing load on transmission lines in actual applications. Our model proved to be superior to conventional icing-load prediction models with regard to the single-step and multi-step prediction accuracies and generalizability. Hence, our prediction model will improve the decision-making processes regarding the deicing and maintenance of power transmission and transformation systems.