Denoising in Steel Structural Impulsive Vibration Signal Based on Independent Component Analysis

Abstract:

The analysis of structure vibration signals is influenced by noise mixed in the signals. Independent component analysis (ICA) method is introduced to denoise the vibration signals in this paper. The representative algorithms: FastICA and JADE are told in detail. The algorithms are applied to separate steel structural vibration signals. The denoising performances in impulsive vibration signals generated by steel structure demonstrate the effectiveness and good robustness of ICA method.

• Periodical: Advanced Materials Research (Volumes 219-220)
• Pages: 1337-1341
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.219-220.1337
• Online since: March 2011
• Authors: Jun Hong Cao, Zhuo Bin Wei
• Keywords: De-Noising, Impulsive Vibration, Independent Component Analysis (ICA), Steel Structure

References

[1] Zhao X Y, Li H N. A improved denoising mathod based on wavelet analysis and its application in structure health monitoring [J]. Journal of vibration and shock, 2007, 26(10):137-139. (In Chinese).

[2] Xia W J, FU X J. De-noising in Intensive Noises Vibration Signal Based on Independent Component Analysis (ICA) [J]. Turbine technology, 2006, 48(2):121-123. (In Chinese).

[3] Jiao W D, Yang S X, Wu Z T. Study of Noise Removal Technique Based on Independent Component Analysis [J]. Journal of Zhejiang University (Engineering Science), 2004, 38(7):872-876. (In Chinese).

[4] Hyvärinen A, Karhunen J, Oja E. Independent Component Analysis [M]. John Wiley, New York, (2001).

[5] Cichocki　A, Amari　S. Adaptive Blind Signal and Image Processing [M]. John Wiley, New York, (2003).

[6] Hyvärinen　A, Ramkumar P, Parkkonen　L, Hari R. Independent component analysis of short-time Fourier transforms for spontaneous EEG/MEG analysis [J]. NeuroImage, 2010, 49(1):257-271.

DOI: 10.1016/j.neuroimage.2009.08.028

[7] ZuoMing　J, Lin　Jing, Fan Xianfeng. Feature separation using ICA for a one-dimensional time series and its application in fault detection [J]. Journal of Sound and Vibration, 2005, 287(3):614-624.

DOI: 10.1016/j.jsv.2005.02.005

[8] Choi　Seungjin, Cichocki　Andrzej, Park Hyung-Min, Lee Soo-Young. Blind Source Separation and Independent Component Analysis: A Review [J]. Neural Information Processing, 2005, 6(1):1-57.

[9] Tse Peter W, Zhang J Y, Wang X J. Blind source separation and blind equalization algorithms for mechanical signal separation and identification [J]. Journal of Vibration and Control, 2006, 12(4):395-423.

DOI: 10.1177/1077546306063248

[10] Yang Y, Yuan H Q, Zhao L Z. Application of Independent Component Analysis in Noise Reduction of Structural Vibration Signal [J]. Journal of Huazhong Univesity of Science and Technolony (Urban Science Edition), 2008, 25(3):234-237. (In Chinese).

[11] Hyvärinen A. Fast and robust fixed-point algorithms for independent component analysis [J]. IEEE Trans.Neural Networks, 1999, 10(3):626-634.

DOI: 10.1109/72.761722

[400] 800](a).

[1200] 1600][2000]-40 -20.

20][40](mv).

[400] 800](b).

20][40](ms) (mv) Observed Signal.

[400] 800][1200][1600][2000]-10 -5.

5][10](mv).

[400] 800][1200][1600][2000]-20 -10.

10][20](ms) (mv) (a) (b) Separated Signal Fig.3 Observed signal of impulse and sin-wave Fig.4 Separated signal of impulse and sin-wave.

[400] 800][1200][1600][2000]-40 -20.

20][40](ms) (mv) Observed Signal (b) (a).

10](mv).

5][10](ms) (mv) Separated Signal (a) (b) Fig.5 Observed signal of impulse and random Fig.6 Separated signal of impulse and random.