Open AccessResearch on EEMD for vibration extreme feature preserving of rotating machinery
Author(s) -
Sanhua Li,
Chunhua Tian,
Chen Wang,
Zhihao Ma
Publication year - 2020
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/715/1/012104
Subject(s) - hilbert–huang transform , rotor (electric) , vibration , noise (video) , feature (linguistics) , feature extraction , mode (computer interface) , pattern recognition (psychology) , artificial intelligence , interference (communication) , computer science , signal (programming language) , control theory (sociology) , engineering , acoustics , computer vision , physics , mechanical engineering , filter (signal processing) , computer network , linguistics , philosophy , channel (broadcasting) , control (management) , image (mathematics) , programming language , operating system
Ensemble Empirical Mode Decomposition (EEMD) has been widely used due to the more advanced processing of mode mixing than Empirical Mode Decomposition (EMD). However, EEMD still suffers mode mixing in the vibration analysis of rotating machinery. It proposes Extremum Feature Preserving Ensemble Empirical Mode Decomposition (EFPEEMD) for the feature extraction of rotor systems in rotating machinery. In the method, a fluctuant variation (FV) index is developed to quantitatively recognize the amplitude of added noise on the basis of the empirical rules in EEMD. The results of a misalignment vibration signal of a rotor system and an inner race defect signal of a rolling bearing show that, the primary vibration modes can be extracted rapidly and independently based on the elimination of the non-stationary and non-linear interference. Moreover, the fluctuant variation index removes the originally manual and empirical steps, thus enhancing the accuracy and automation of EEMD on the basis of preserving the extreme information.