Harmonics-to-noise ratio (HNR) is an important health index of rotating machine, which has been applied in blind deconvolution (BD) method to realize periodic impulse detection. However, most fault impulses are not strictly periodic, but pseudo-cyclostationary, will affect the performance HNR characterization some extent. This limits its applications. Therefore, this paper, a novel BD method, maximum squared envelope spectrum harmonic-to-interference (MSESHIRD), proposed more effectively...

Sparse representations based on convolutional sparse dictionary learning (CSDL) provides an excellent framework for extracting fault impulse response caused by bearing faults. In order to achieve fast learning, most CSDL-based diagnosis techniques recommend truncating the original data. However, choice of truncation length is very difficult. An improper will lead problems pattern rupture and uneven sparsity distribution. By contrast, if data are not truncated, these occur. this result in...

Since minimum correlated generalized Lp/Lq deconvolution (MCG-Lp/Lq-D) cannot simultaneously have a good sensing ability to repetitive fault impulses and strong robustness noise when processing low SNR signals, new method, improved (MICG-Lp/Lq-D), is proposed in this article. The method designed based on novel norm, the (ICG-Lp/Lq) norm. ICG-Lp/Lq norm not only has low-order statistics form that similar (CG-Lp/Lq) used MCG-Lp/Lq-D, but also high shift-order signal correlation construction...

Signal decomposition techniques generally have two limitations. First, for adaptive correlated kurtogram (ACK) and variational mode (VMD), etc., the definition of target has no direct relationship with features fault impacts, i.e., is not guided by health indices used to characterize bearing faults, resulting in impacts cannot be completely separated from vibration data. Second, it hard accurately determine stop parameter (the number final output modes), frequent over- or...

In recent years, the rolling bearing fault diagnosis technique based on deep learning (DL) provides a more intelligent and reliable way for safe operation of mechanical systems. However, this still exists problems high model complexity poor generalization ability in application. To solve above problem, novel periodic cyclic sparse network with entire domain adaptation (PcsNet-EDA) transfer is proposed article. The design pattern makes weight matrices convolutional layer fully connected...

Sparse representation has been widely used in bearing fault impact detection, which can find the that best matches waveform from pre-defined dictionary and recover impulse waveform. However, current of sparse efficiency algorithm need to be improved. In order accurately detect original signal, a convolutional coding using pathfinder algorithm-optimized orthogonal matching pursuit with asymmetric Gaussian chirplet model (CSC-OAGCM) is proposed this paper. A new time-frequency atom prototype,...

The health status of bearing is directly related to the safe operation rotating machinery. Bearing fault detection technology great significance reduce or eliminate safety accidents. Singular value decomposition (SVD), as an effective low-rank approximation tool data matrix, widely used in detection. construction trajectory matrix and selection singular are two important factors that affect performance SVD-based diagnosis methods. In this paper, a new similar SVD, shifted rank-1...

Convolutional sparse representation based on local orthogonal matching pursuit (SR-LocOMP) can recover wheelset-bearing fault impulses without being affected by random slippage and plays an important role in diagnosis. However, the performance of SR-LocOMP at a low signal-to-noise ratio (SNR) is not satisfied. In addition, when analyzed signals contain compound faults, feature extraction ability will be greatly affected. To further improve SR-LocOMP, this article proposes improved algorithm,...

Rotating machinery is an important and easily damaged component in large-scale equipment. Under the coupling action of system components, occurrence rate compound faults very high, which seriously endangers equipment safety. The vibration signals rotating machine include operation information, periodic impacts, environmental noise, even accidental impacts. To effectively extract multi-fault features from fault signals, a multi-period pulse detection indicator called harmonic spectrum...

Since the cyclostationarity in vibration signals is key to judge rotating machine health state, spectral harmonics-to-interference ratio (SHIR) has been used construct single node feature learning network (SHIR-based blind deconvolution, BD-SHIR) realize condition monitoring of machine. However, BD-SHIR still several obvious limitations, including need for prior fault information and tendency fall into local optimal solutions which will affect its state performance. Therefore, this paper...

Fast kurtogram (FK) has been proven to be an effective tool for resonance frequency band detection, which is widely used in bearing fault diagnosis. However, FK not robust impulsive noise, and its segmentation rule fixed, leads over-decomposition or under-decomposition of the signal decomposition results. Therefore, adaptive cyclic content ratiogram proposed this paper. Firstly, based on energy distribution vibration signals different components, spectral performed adaptively, multiple...

Minimum correlated generalized Lp / Lq deconvolution (MCG‐ ‐D) is an important tool to detect periodic impulses in vibration mixture. It proved be a more stable technique than maximum kurtosis (MCKD) recover the fault impulse under strong noise conditions. However, MCG‐ ‐D still has limitations. One of necessary conditions for success provide precise period fault. An imprecise prior will lead performance degradation or even failure method. Therefore, this paper, with adaptive estimation...

Transfer learning (TL) has made great progress in intelligent fault diagnosis of bearing. However,due to the harsh working conditions bearings engineering practice, addition bearing vibration signals, sensors also inevitably collect noise leading performance degradation TL. In this paper, a novel model based on TL is proposed solve problem. Different from previous studies, an adaptive input length instead fixed established diagnose with different parameters. A signal processing method...