Sparse Representation of Vibration Signals Using Trained Dictionary

Jun Feng Guo; Xiao Hui Zheng; Xing Chun Wei; Rui Cheng Feng

doi:10.4028/www.scientific.net/AMM.574.690

Paper Titles

Generalized Hybrid Dislocated Function Projective Synchronization of Chaotic Systems with Time Delay
p.672

The Model of Optimal Firing Distribution for Naval Gun Firing on Shore
p.679

Feature Extraction Method for Rolling Bearing’s Week Fault Based on Kalman Filter and FSK
p.684

Sparse Representation of Vibration Signals Using Trained Dictionary
p.690

Cutting Institutions Running State Estimation Strategy Based on Wavelet Packet Analysis Technology
p.696

A Structural Reliability Analysis Approach for Uncertain Structures via a PSO-DE Hybrid Algorithm
p.702

Continue Wavelet Transform's Analysis in Zhuji Station's Rainfall Sequence Transformation
p.708

Condensed Nearest Neighbor Decision Rule Input Weight Sequential Feed-Forward Neural Networks
p.712

HomeApplied Mechanics and MaterialsApplied Mechanics and Materials Vol. 574Sparse Representation of Vibration Signals Using...

Sparse Representation of Vibration Signals Using Trained Dictionary

Abstract:

The sparsity of a signal is critical for compression and high compression performance is obtained by utterly sparse signal. But real signals are not sparse commonly, so sparse transformation is considered. In addition, the sparse degree of coefficients is mainly determined by the quality of transform base. Therefore, this paper constructs a transform base (trained dictionary) by K-SVD algorithm for rolling element bearing vibration signals and uses the Orthogonal Matching Pursuit (OMP) algorithm to conduct sparse representation and simulation. Results show that the trained dictionary can be more fitted with the features of signals, the residual components are smaller and the reconstruct similarity is higher compared to the untrained dictionaries, obtaining better representation.

You might also be interested in these eBooks

Recent Research on Mechanical Engineering, Mechatronics and Automation

View Preview

Info:

Periodical:

Applied Mechanics and Materials (Volume 574)

Pages:

690-695

DOI:

https://doi.org/10.4028/www.scientific.net/AMM.574.690

Citation:

Cite this paper

Online since:

July 2014

Authors:

Jun Feng Guo*, Xiao Hui Zheng, Xing Chun Wei, Rui Cheng Feng

Keywords:

Bearing Vibration Signals, K-SVD Algorithm, Sparse Decomposition, Trained Dictionary

Export:

RIS, BibTeX

Price:

Permissions CCC:

Request Permissions

Permissions PLS:

Request Permissions

Сopyright:

Citation:

* - Corresponding Author

References

[1] S. Mallat and Z. Zhifeng, Matching pursuits with time-frequency dictionaries, IEEE Trans. Signal Process., vol. 41, no. 12, pp.3397-3415, (1993).

DOI: 10.1109/78.258082

Google Scholar

[2] M. Qingfeng and F. Xiaoqi, Characteristic waveform extraction of mechanical dynamic information using signal decomposition method, China, no. CN02114530. X, 2002-04-22.

Google Scholar

[3] M. Aharon, M. Elad, and A. M. Bruckstein, K-SVD: An algorithm for designing of over-complete dictionaries for sparse representation Technion-Israel Inst. of Technology, 2005, Tech. Ref.

DOI: 10.1109/tsp.2006.881199

Google Scholar

[4] B. Shan, W. Hao and Rui Zhao, Infrared image de-noising based on K-SVD over-complete dictionaries learning, Image and Signal Processing (CISP), 2012 5th International Congress, Chongqing, Sichuan, China, (2012).

DOI: 10.1109/cisp.2012.6469839

Google Scholar

[5] Y. Yang, Signal OMP sparse decomposition using particle swarm algorithm, Microcomputer Information, no. 12, pp.178-180, (2008).

Google Scholar

[6] http: /csegroups. case. edu/bearingdatacenter/pages/download-data-file, Case Western Reserve University Bearing Data Center, (2013).

Google Scholar

[7] Zhang G X, Time-Frequency Atom Decomposition with Quantum-Inspired Evolutionary Algorithms, Cirvuits, Systems and Signal Processing, vol. 29, no. 2, pp.209-233, (2009).

DOI: 10.1007/s00034-009-9142-3

Google Scholar