For imbalanced bearing fault diagnosis, generative adversarial networks (GANs) are a common data augmentation (DA) approach. Nevertheless, current GAN-based methods cannot update the generator from time–frequency domain simultaneously, downgrading authenticity of signal character. In this article, Fourier-like transform GAN (FTGAN), novel method, is proposed by introducing transformer (FLT) based on autoencoder (AE) to improve synthetic quality. FLT approximates discrete Fourier (DFT) neural...

The multivariate empirical mode decomposition (MEMD) has been pioneered recently for adaptively processing of multichannel data. Despite its high efficiency on time-frequency analysis nonlinear and nonstationary signals, computational load over-decomposition have restricted wider applications MEMD. To address these challenges, a fast MEMD (FMEMD) algorithm is proposed featured by the following contributions: 1) A novel concept, pseudo direction-independent intrinsic function (IMIMF) which...

Computational imaging provides comprehensive and reliable information about human tissue for medical diagnosis treatment, with image fusion as one of the most important technologies in field. Empirical mode decomposition (EMD), a promising model processing, has been used some methods. However, varying number decomposed layers leads to problems using EMD fusion. In this article, we propose method images incorporating L2 -norm-based features, match/salience/fuzzy-weighted measure, 2-D...

Oscillations that propagate throughout a plant have severe impacts on an industrial process. However, it is still open problem to design time-frequency analyzing tool can characterize the amplitude, frequency, and trend information nonstationary plant-wide oscillations. Motivated by this difficulty, novel algorithm, multivariate intrinsic time-scale decomposition (MITD), proposed. At first, sifting process added into standard (ITD), ensuring each of decomposed product be monocomponent. Then,...

A novel detector based on improved variational mode decomposition (VMD) is proposed to detect the nonlinearity-induced oscillations. Despite its high adaptivity and frequency resolution, effectiveness of VMD highly depends parameters, including number <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$K$ </tex-math></inline-formula> , initial center notation="LaTeX">$\omega _{init}$ penalty coefficient...

Industrial signal measurement and processing are increasingly being deployed in control performance assessment, particularly oscillation monitoring applications. In this paper, we present a novel detector, which mainly leverages the recently developed fast multivariate empirical mode decomposition (FMEMD). Our use of FMEMD is motivated by following facts: (i) considerably fewer techniques now available for both single-loop plant-wide oscillations; (ii) presence noise intermittency can...

The detection of plant-wide oscillation in an industrial process is great significance. Recently, indirect multivariate intrinsic time-scale decomposition (MITD) (IMITD) has been pioneered for the adaptive processing multi-loop data, which restricted by problem projection sensitivity. To solve challenge, direct MITD (DMITD) algorithm proposed and featured following contributions: 1) Three novel concepts including extremum, baseline-node, baseline-operator are defined purpose developing...

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 techniques have proposed, most them can only address part practical difficulties. An is heuristically defined as visually apparent periodic variation. However, manual visual inspection labor-intensive prone missed detection. Convolutional neural networks (CNNs), inspired by animal systems, raised with...

The conventional semi-supervised extreme learning machine (SS-ELM) algorithm can provide a solution to the lack of labeled samples in wind turbine blade icing fault detection, but its performance is limited by irrationality spherical nearest neighbor graph (SNNG) calculation strategy. To solve this problem, novel ellipsoidal (ESS-ELM) proposed paper and applied detection. In study, we creatively propose (ENNG) strategy that considers distribution information construct ESS-ELM algorithm....

Density peak clustering (DPC) can identify cluster centers quickly, without any prior knowledge. It is supposed that the have a high density and large distance. However, some real datasets hierarchical structure, which will result in local having but smaller DPC flat algorithm searches for globally, considering differences. To address this issue, Multi-granularity (MG-DPC) based on Variational mode decomposition (VMD) proposed. MG-DPC find global coarse-grained space, as well fine-grained...

The detection and diagnosis of oscillation are great importance to maintain the control performance process plant. Even though some algorithms based on time-frequency analysis for multiple oscillations have been reported in literature, their is severely constrained by mode-mixing mode-splitting problems. presence these issues leads not only incorrectly detected number oscillations, but also incorrect identification sources. Motivated above challenge, this work proposes a novel adaptive tool:...