Blind Separation of Excavator Noise Signals in Frequency Domain

Abstract:

In order to identify noise sources of an excavator in non-library environment, a complex-valued algorithm in frequency domain was applied. Firstly, an acoustic camera was used to acquire excavator’s noise signals, which were convolutive mixtures in time domain interfered by echo. Secondly, signals in time domain transformed into frequency domain by FT, turned to be complex-valued mixtures. Then, independent components of noise signals were obtained through separation of complex-valued mixtures using complex-valued algorithm based on independent component analysis. Finally, according to noise of diesel with muffler was mainly consist of surface noise, the relationship between principal frequencies and structrual parts was founded by comparing frequency-amplitude spectra and modal analysis in Ansys. Research shows that complex-valued algorithm based on fast fixed-point independent component analysis can effectively separate noise signals from an excavator in time domain, and noise sources can be well ascertained by comparing the modal analysis with blind separation components.

• Periodical: Applied Mechanics and Materials (Volumes 105-107)
• Pages: 723-728
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.105-107.723
• Online since: September 2011
• Authors: Li Da Liao, Qing Hua He, Zhong Lin Hu
• Keywords: Excavator, Independent Component Analysis (ICA), Modal Analysis, Noise

References

[1] J.H. Zhang, Y.L. Yu and B. Han: Identification of diesel front sound source based on continuous wavelet transform, Chinese Journal of Mechanical Engineering, Vol. 17(2004) No. 2, p.268.

DOI: 10.3901/cjme.2004.02.268

[2] Z.Y. Hao and J. Han: Identification of diesel front sound source based on contnuous wavelet transform, Journal of Zhejiang University SCIENCE, Vol. 5(2004) No. 9, p.1069.

DOI: 10.1631/jzus.2004.1069

[3] S.J. Rennie, P. Aarabi and B.J. Frey: IEEE Transactions on Audio, Variational probabilistic speech separation using microphone arrays, Vol. 15(2007) No. 1, p.135.

DOI: 10.1109/tasl.2006.876865

[4] K. Torkkola: ICA'99, Blind separation for audio signals—are we there yet? Vol. 1 (2000) No. 1, p.321.

[5] J.K. Tugnait: Adaptive blind separation of convolutive mixtures of independent linear signals, Signal Processing, Vol. 1 (1999) No. 73, p.153.

DOI: 10.1016/s0165-1684(98)00189-3

[6] X. Wu, J.J. He and S.J. Jin. Blind separation of speech signals based on wavelet transform and independent component analysis, Transactions of Tianjin University, Vol. 16(2010)No. 2, p.123.

DOI: 10.1007/s12209-010-0022-5

[7] H. Park: Shekhar mixtures, Neurocomputing, Vol. 69(2006)No. 16, p. (2065).

[8] H. Aapo: IEEE Trans. on Neural Network(USA, 1999), Fast and robust fixed-point algorithm for independent component analysis, Vol. 10(1999)No. 3, p.626.

DOI: 10.1109/72.761722

[9] A.D. Back and A.C. Tsoi: Blind deconvolution of signals using a complex recurrent network, Neural Networks for Signal Processing, Vol. 1(1994) No. 4, p.565.

DOI: 10.1109/nnsp.1994.366009

[10] P. Smaragdis: Blind separation of convolved mixtures in the frequency domain, Neuro computing, Vol. 2(1998)No. 22, p.21.

DOI: 10.1016/s0925-2312(98)00047-2

[11] S. Fiori: Blind separation of circularly distributed sources by neural extended APEX algorithm, Neurocomputing. Vol. 6(2000)No. 34, p.239.

DOI: 10.1016/s0925-2312(00)00161-2

[12] H. Aapo and O. Erkki: Independent component analysis: algorithms and applications, Neural Networks, Vol. 13(2000)No. 4, p.411.

[13] H. Aapo, J. Karhunen, and E. Oja: Independent Component Analysis(A Wiley-Interscience Publication, USA 2001).

[14] E. Morean and O. Macchi: EUSIPCO'94 (Edinburgh, Scotland, 1994), Complex self-adaptive algorithms for source separation based on higher order contrasts. Vol. 1, p.1157.