1. Rolling element bearings are the critical parts of every rotating machinery and their failure is one main reasons machine downtime even breakdown. The significance early detection cannot be overstated, as it plays a crucial role in maintaining proper functioning equipment, enhancing production efficiency, ensuring safety. Among them, envelope analysis most effective widely used approach, working according to principle linear filtering process signals remove undesirable components....

Abstract In addressing the problem of low prediction accuracy in remaining useful life (RUL) rolling bearings, caused by noise interference and insufficient extraction sensitive features deep learning models, this paper presents a method based on signal reconstruction dual-channel network fusion. First, issue extracting weak from bearing vibration signals, an optimized combination variational mode decomposition (VMD) Teager–Kaiser energy operator (TKEO) for is proposed. TKEO used to track...

Under normal circumstances, bearings generally run under variable loading conditions. such conditions, the vibration signals of bearing malfunctions are often nonstationary signals, which difficult to process effectively. In order accurately and effectively diagnose failure types damage degree load an intelligent diagnostic model based on variational mode decomposition (VMD) quantum chaotic fruit fly optimization algorithm (QCFOA) a multiclassification relevance vector machine (VRVM) is...

摘要: 低速重载机械出现早期故障时,振动信号中体现故障特征的微冲击成分具有稀疏性。根据振动动力学模型建立的故障信号过完备冗余字典,能实现对振动信号的稀疏逼近。分段正交匹配追踪(Stagewise orthogonal matching pursuit, StOMP)在正交匹配追踪(Orthogonal OMP)的基础上,采用框架的思想对信号进行稀疏分解,不但克服了OMP方法导致的过匹配现象,也提高了算法的收敛速度,但在计算残差在子原子库的表示时计算量很大。基于相干累积量的StOMP方法根据故障信号过完备字典中各原子的相关性,分析了内置相干累积量和外置相干累积量的关系,并通过故障信号在字典中的内、外置相干累积量的值快速确定原子的位移因子和频率因子,进而为StOMP方法提供更为高效的子原子库选取策略,最后结合原子淘汰算法对影响不大的原子进行筛选,最终选出最能稀疏表示信号的一组原子。

To address the challenges faced in prediction of rolling bearing life, where temporal signals are affected by noise, making fault feature extraction difficult and resulting low accuracy, a method based on optimal time–frequency spectra DenseNet-ALSTM network is proposed. Firstly, signal reconstruction introduced to enhance vibration signals. This involves using CEEMDAN deconvolution combined with Teager energy operator for reconstruction, aiming denoise highlight impacts. Subsequently, snake...

Aiming at non-stationary signals with complex components, the performance of a variational mode decomposition (VMD) algorithm is seriously affected by key parameters such as number modes K, quadratic penalty parameter α and update step τ. In order to solve this problem, an adaptive empirical (EVMD) method based on binary tree model proposed in paper, which can not only effectively problem VMD selection, but also reduce computational complexity searching optimal using intelligent optimization...

Aiming at the problems of early weak fault feature extraction bearings in rotating machinery, an improved stochastic resonance (SR) is proposed combined with advantage SR to enhance characteristic signals noise energy. Firstly, according characteristics large parameters actual signal, amplitude transform coefficient and frequency are introduced convert parameter signal into small which can be processed by SR, relationship second-order introduced. Secondly, a comprehensive evaluation index...

In the incipient fault vibration signals of rolling bearings, weak features are easily submerged in strong background noise and difficult to be extracted. The sparse decomposition method can perform well extraction features, but low signal-to-noise ratio (SNR) would cause excessive decomposition. To enhance maintain time–frequency structure impulses, a novel feature bearing based on signal reconstruction is proposed. Firstly, Teager energy operator (TEO) used obtain envelope impulse...

Summary In this abstract, we discuss some fundamental issues of seismic envelope inversion (EI). Firstly, give the relationship between beat tone and inversion. Then, use a combined model test to prove reliability low-frequency information recovered by We believe that was powerful tool recover low-wavenumber components velocity model. However, EI method can be trapped into local minima as well, when it comes observed data with higher dominant frequency band. To address this, propose...

Underwater acoustic sensor networks (UASNs), which adopt communication, have been widely used underwater, such as in oceans and rivers, for monitoring applications. However, the UASNs designed urban lake are facing challenges seldom regarded. Based on analysis of features lakes requirements, a UASN system is this article. The node deployment strategy discussed to balance density transmission bandwidth. Then hybrid topology proposed meet nodes water surface construct 2-D ring topology,...

During the data transmission in underwater sensor networks, affected by complex environment, especially when network loads become large, collisions will be intensified, resulting decrease performance. Therefore, how to forward sink nodes with high reliability is a key issue. Link quality one of important factors that affect performance communication routing. Poor link lead excessive retransmissions, leading low effective packets reception rate. Hence, many routing protocols based on are...

Under variable load conditions, the bearing vibration signal is non-stationary, which renders ineffective techniques used for fault diagnosis under constant running conditions. A model of variational mode decomposition (VMD) and multi-classification correlation vector machine (MRVM) based on chaotic quantum particle swarm optimization (CQPSO) proposed. First, number intrinsic function (IMF) penalty parameter VMD optimized by QPSO algorithm to search optimal combination value two parameters....

The consistency anomaly of multimodal data during transmission is a challenge faced by social networks, intelligent monitoring, and other applications. These existing research methods are based on supervised training learning, which require large amount high computing power. To overcome the problem, detection algorithm sparse subspace clustering proposed in this paper. First, to solve problem that performance algorithms usually highly correlated with anomalous has limited application,...

Normal vibration signal of rotating machinery is low-frequency signal, yet high-frequency component would exist when having fault, especially the fault caused by mechanical shock. But shock pulse very weak compared to component. Demodulated resonance technique can filter and keep Then generate wave in which amplified. The signals from gearbox are coupled each other make diagnosis difficult, with multiple sources decoupling detect independent measuring point. First, construct relative...

In order to meet consumer's huge demands for cosmetic match, this project plans develop a match application --- Beauty, based on support vector machine. according pictures including different shapes of face, can adopt Principle Component Analysis (PCA) optimize image pixel, get the lower dimension feature machine's input, and then cross-validation grid-search select penalty parameter c kernel g train model. Finally, facial are tested respectively by making use these models. And consumers...