Open AccessFeature extraction of non-linearity and low noise-signal ratio signals and its application to engine fault diagnosis
Author(s) -
Fengli Wang,
Yannian Cai
Publication year - 2019
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1423/1/012025
Subject(s) - hilbert–huang transform , diesel engine , fractal dimension , cylinder head , vibration , noise (video) , acoustics , fault (geology) , nonlinear system , leakage (economics) , fractal , feature extraction , signal (programming language) , engineering , computer science , pattern recognition (psychology) , control theory (sociology) , automotive engineering , mathematics , artificial intelligence , physics , mathematical analysis , internal combustion engine , electrical engineering , geology , filter (signal processing) , image (mathematics) , macroeconomics , control (management) , quantum mechanics , programming language , seismology , economics
Since the vibration signals of engine cylinder head come from different interference sources, it is difficult to detect weak but useful signals in the background of noise. Aiming at nonlinear and low noise characteristics of engine vibration signals, a method based on integrated empirical mode decomposition (EMD) and fractal dimension is proposed. Firstly, the integrated EMD method is used to decompose the nonlinear low noise signal into a set of intrinsic mode functions from high to low. Then the morphological fractal dimension is applied to the intrinsic mode function (IMFs). The fractal dimension of the characteristic IMF is calculated. The vibration signals of a diesel engine exhaust valve under normal and leakage conditions are analysed. The results show that this method can extract fault features of diesel engine exhaust valve.