Rolling bearing fault characteristic extraction method based on dictionary learning

The invention discloses a rolling bearing fault characteristic extraction method based on dictionary learning, and belongs to the field of signal processing. The method comprises the following steps that: obtaining a vibration signal from a sensor installed on a bearing to be shown as s, and marking...

Full description

Saved in:
Bibliographic Details
Main Authors ZHOU HAN, SUN CHUAN, CHAI YI, HAN FENGBIN, YIN HONGPENG
Format Patent
LanguageChinese
English
Published 21.12.2018
Subjects
CALCULATING
COMPUTING
COUNTING
HANDLING RECORD CARRIERS
MEASURING
PHYSICS
PRESENTATION OF DATA
RECOGNITION OF DATA
RECORD CARRIERS
TESTING
TESTING STATIC OR DYNAMIC BALANCE OF MACHINES ORSTRUCTURES
TESTING STRUCTURES OR APPARATUS NOT OTHERWISE PROVIDED FOR
Online AccessGet full text

Cover

More Information
Summary:The invention discloses a rolling bearing fault characteristic extraction method based on dictionary learning, and belongs to the field of signal processing. The method comprises the following steps that: obtaining a vibration signal from a sensor installed on a bearing to be shown as s, and marking sampling frequency as fs; adopting a harmonic separation filter, removing harmonic components whichare contained in the s and are generated by the vibration of other mechanisms to obtain a processing signal y; carrying out Shift-Invariant K-SVD (Singular Value Decomposition) dictionary learning onthe signal y to obtain sparse representation and an optimal dictionary, and according to the sparse representation and the dictionary, reconstructing and recovering a pure fault signal x; and carrying out envelope spectrum transformation on the signal x to obtain the frequency characteristics of a bearing fault vibration signal. By use of the method, a sparse representation method is adopted, andearly weak fault character
Bibliography:Application Number: CN201811027613