Negentropy and gradient iteration based fast independent component analysis for multiple random fault sources blind identification and separation

Abstract

In order to separate multiple random fault sources only from mixed vibration measurement response signals of mechanical system, Negentropy and Gradient iteration based fast independent component analysis (FastICA) is applied for blind signal separation (BSS). After finding the association between independent components (ICs) matrix and multiple random fault sources, multiple random fault sources identification problem is turned into ICA of the stationary random vibration response signals of mechanical system. This method uses negative entropy maximization as criterion of independence, Gradient iteration as optimization method to extract random fault sources one by one. Simulation experiment results verified that this method could identify and separate multiple random fault sources only from mixed vibration measurement response signals of mechanical system correctly and effectively.

Keywords

Blind signal separation fast independent component analysis multiple random fault sources blind identification and separation vibration measurement response signals negentropy gradient iteration

Get full access to this article

View all access options for this article.

References

McFadden

P.D.

and Smith

J.D.

, Vibration monitoring of rolling element bearings by the high-frequency resonance technique-a review [J], Tribology International17(1) (1984), 3-10.

, Yan

, Tian

, Yuan

, Peng

and Li

, Blind vibration component separation and nonlinear feature extraction applied to the nonstationary vibration signals for the gearbox multi-fault diagnosis [J], Measurement46(1) (2013), 259-271.

Alexander

, Learning methods for mechanical vibration analysis and health monitoring [M], Delft University of Technology Press, 1998, 15-16.

Guo

, Na

, Li

and Fung

R.-F.

, Envelope extraction based dimension reduction for independent component analysis in fault diagnosis of rolling element bearing[J], Journal Of Sound And Vibration333(3) (2014), 2983-2994.

Samworth

R.J.

and Yuan

, Independent component analysis via nonparametric maximum likelihood estimation [J], The Annals of Statistics40(6) (2012), 2973-3002.

Langlois

, Chartier

and Gosselin

, An introduction to independent component analysis: InfoMax and FastICA algorithms [J], Tutorials in Quantitative Methods for Psychology6(1) (2010), 31-38.

Raimondo

, Kamienkowski

J.E.

, Sigman

et al., CUDAICA: GPU optimization of infomax-ICA EEG analysis [J], Computational Intelligence and Neuroscience (2012), 2.

Zhou

, Yang

, Xie

et al., Online blind source separation using incremental nonnegative matrix factorization with volume constraint[J], Neural Networks, IEEE Transactions on22(4) (2011), 550-560.

, Jin

, Lou

et al., An optimized EASI algorithm [J], Signal Processing89(3) (2009), 333-338.

10.

Adali

, Anderson

and Fu

G.S.

, Diversity in independent component and vector analyses: Identifiability, algorithms, and applications in medical imaging [J], Signal Processing Magazine, IEEE31(3) (2014), 18-33.

11.

Chen

, Hung

, Komori

et al., Robust Independent Component Analysis via Minimum-Divergence estimation [J], Selected Topics in Signal Processing, IEEE Journal of7(4) (2013), 614-624.