Multidimensional denoising of rolling element bearings with compound fault based on tensor factorization

DOI

10.2991/macmc-17.2018.68How to use a DOI?

Keywords

Tensor factorization; fault diagnosis; high order singular value decomposition; denoising

Abstract

this article presents a multidimensional denoising technique of rolling element bearings based on tensor factorization which can model a signal in the high dimensional space so as to solve multi-channel signal filtering. The vibration signal is formulated as a 4-way tensor, temporal signal, frequency and 2 data channels. Tensor model is then factorized via truncated high order singular value decomposition. L-curve criterion is adopted to find the truncation parameters used in tensor factorization. The proposed approach is applied to remove noise of bearing vibration signal on test-rigs. Experimental results showed that the proposed method can well remove noise and keep fine structures of the signal as much as possible. This tensor based multidimensional signal filtering will broaden view for dealing with heterogeneous and multiaspect data in an age of big data.

Copyright

© 2018, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Chaofan Hu
AU  - Yanxue Wang
PY  - 2018/01
DA  - 2018/01
TI  - Multidimensional denoising of rolling element bearings with compound fault based on tensor factorization
BT  - Proceedings of the 2017 4th International Conference on Machinery, Materials and Computer (MACMC 2017)
PB  - Atlantis Press
SP  - 350
EP  - 356
SN  - 2352-5401
UR  - https://doi.org/10.2991/macmc-17.2018.68
DO  - 10.2991/macmc-17.2018.68
ID  - Hu2018/01
ER  -