Research on Small Sample Rolling Bearing Fault Diagnosis Method Based on Mixed Signal Processing Technology DOI Open Access

Peibo Yu,

Jianjie Zhang, Baobao Zhang

и другие.

Symmetry, Год журнала: 2024, Номер 16(9), С. 1178 - 1178

Опубликована: Сен. 9, 2024

The diagnosis of bearing faults is a crucial aspect ensuring the optimal functioning mechanical equipment. However, in practice, use small samples and variable operating conditions may result suboptimal generalization performance, reduced accuracy, overfitting for these methods. To address this challenge, study proposes fault method based on symmetric two-stream convolutional neural network (CNN). employs hybrid signal processing techniques to issue limited data. parallel (CNN) analysis Initially, data are transformed into time–frequency maps through utilization short-time Fourier transform (STFT) simultaneous compressed wavelet (SCWT). Subsequently, two sets one-dimensional vectors generated by reconstructing high-resolution features faulty using Feature splicing fusion then performed generate information assist classification. experimental results demonstrate that proposed mixed-signal effective small-sample datasets, verify feasibility generality CNN-support vector machine (SVM) model under conditions.

Язык: Английский

Effectively reduce transient vibration of 2D wing with bi-stable metamaterial DOI
Bing Hu, Xin Fang, Jihong Wen

и другие.

International Journal of Mechanical Sciences, Год журнала: 2024, Номер 272, С. 109172 - 109172

Опубликована: Март 8, 2024

Язык: Английский

Процитировано

17

Dynamic responses of shear thickening fluid-filled lattice structures DOI Creative Commons

Qifang Hu,

Guoxing Lu, Kwong Ming Tse

и другие.

International Journal of Impact Engineering, Год журнала: 2024, Номер 189, С. 104954 - 104954

Опубликована: Март 29, 2024

Lattice structures that possess exceptional energy absorption capabilities show promise across various fields. However, their tendency to exhibit relatively lower stiffness due rotation or bending during impacts poses a challenge. Shear thickening fluid with the unique property of shear effect can be an ideal filler overcome limitation lattice structure. In this paper, composites combined STF new were fabricated investigate dynamic response using experimental and numerical means found incorporating into these not only preserved inherent mechanical characteristics but also led substantial improvements on performance. The displayed noteworthy boost in stiffness, increasing significantly from 11.02 85.56 times. Additionally, there was improvement absorption, ranging 2.78 5.51 These findings indicate holds more enhancing structure's compared its capacity. Moreover, strain rate, weight fraction STF, cell size structure considered further impact behaviour STF-filled structures. Our increase following filling is correlated rate decreases when increased. critical predicted, around 104 s−1, which indicates onset shear-thickening STF. Furthermore, it observed higher fractions responsive changes rate. initially properties experienced most significant after being filled optimal outcomes provide valuable insights for design practical applications.

Язык: Английский

Процитировано

12

Nonlinear free vibrations of Timoshenko–Ehrenfest beams using finite element analysis and direct scheme DOI
Nasser Firouzi, Stefano Lenci, Marco Amabili

и другие.

Nonlinear Dynamics, Год журнала: 2024, Номер 112(9), С. 7199 - 7213

Опубликована: Март 12, 2024

Язык: Английский

Процитировано

12

Mechanical properties of the novel spider-web bionic hierarchical honeycomb under out-of-plane crushing DOI
Zhenzhen Cai, Xiaolin Deng

Applied Physics A, Год журнала: 2024, Номер 130(1)

Опубликована: Янв. 1, 2024

Язык: Английский

Процитировано

11

Dynamic stability of the Mindlin-Reissner plate using a time-modulated axial force DOI Creative Commons
Nasser Firouzi, Fadi Dohnal

Mechanics Based Design of Structures and Machines, Год журнала: 2024, Номер unknown, С. 1 - 18

Опубликована: Июнь 14, 2024

A time-periodic axial force acting on a continuous structure can be destabilized by parametric excitation. Moreover, it is capable of increasing the equivalent damping and shortening transient vibrations. In this work, concept extended, dynamic stability Mindlin-Reissner plate analyzed. The FE formulation for four-node element derived employing variational approach including action an force. system evaluated both numerically analytically, using Floquet theory first-order approximation based averaging method, respectively. influence length, thickness, aspect ratio highlighted. It demonstrated that instability tongue at combination resonance frequency summation type shifted toward higher amplitudes width decreasing length. Alternatively, in vicinity difference shows antiresonance which maximized exactly opposite tuning design. This guideline important when design under harmonic considered.

Язык: Английский

Процитировано

6

Assessment and control of the mine tremor disaster induced by the energy accumulation and dispersion of thick-hard roofs DOI Creative Commons
Bin Yu,

Mingxian Peng,

Yang Tai

и другие.

International Journal of Mining Science and Technology, Год журнала: 2024, Номер 34(7), С. 925 - 941

Опубликована: Июль 1, 2024

Язык: Английский

Процитировано

4

Dynamic stability analysis of tapered rotating beams with 2D functionally graded materials: A comparative study of Floquet theory approaches DOI
Moustafa S. Taima, T.A. El-Sayed, Michael I. Friswell

и другие.

Thin-Walled Structures, Год журнала: 2025, Номер unknown, С. 113257 - 113257

Опубликована: Апрель 1, 2025

Язык: Английский

Процитировано

0

Numerical investigation on effect of different parameters on nonlinear vibration response of fully geometrically exact Timoshenko beams DOI
Nasser Firouzi, Ahmed S. M. Alzaidi,

Hamid Nezaminia

и другие.

Applied Physics A, Год журнала: 2024, Номер 130(9)

Опубликована: Авг. 9, 2024

Язык: Английский

Процитировано

3

Research on Small Sample Rolling Bearing Fault Diagnosis Method Based on Mixed Signal Processing Technology DOI Open Access

Peibo Yu,

Jianjie Zhang, Baobao Zhang

и другие.

Symmetry, Год журнала: 2024, Номер 16(9), С. 1178 - 1178

Опубликована: Сен. 9, 2024

The diagnosis of bearing faults is a crucial aspect ensuring the optimal functioning mechanical equipment. However, in practice, use small samples and variable operating conditions may result suboptimal generalization performance, reduced accuracy, overfitting for these methods. To address this challenge, study proposes fault method based on symmetric two-stream convolutional neural network (CNN). employs hybrid signal processing techniques to issue limited data. parallel (CNN) analysis Initially, data are transformed into time–frequency maps through utilization short-time Fourier transform (STFT) simultaneous compressed wavelet (SCWT). Subsequently, two sets one-dimensional vectors generated by reconstructing high-resolution features faulty using Feature splicing fusion then performed generate information assist classification. experimental results demonstrate that proposed mixed-signal effective small-sample datasets, verify feasibility generality CNN-support vector machine (SVM) model under conditions.

Язык: Английский

Процитировано

3