Title :
The Feature Extraction Based on Nonlinear Partial Orthogonal Analysis with B-Spline Function for Signal of Vibrational Sensors
Author :
Liu Gui-Guo ; Yue Bin ; Huang Dao ; Huang Hai-Yan
Author_Institution :
State Eng. Res. Center of Ind. Autom., ECUST, Shanghai
Abstract :
The new method of feature extraction based on nonlinear partial orthogonal analysis with B-spline function for signal of vibrational sensors is proposed in the paper. Firstly, the nonlinear signal is mapped into linear space; then, the signal is analyzed in the linear space; finally, a set of eigenvalue vectors are obtained with nonlinear partial orthogonal analysis. The vectors reflect a pattern of the signal steadily. The experiment for vibrational signals of normal and abnormal rolling bearings shows that the pattern of every signal is extracted with the method steadily and effectively.
Keywords :
eigenvalues and eigenfunctions; feature extraction; rolling bearings; sensors; signal processing; splines (mathematics); vibrations; abnormal rolling bearings; b-spline function; eigenvalue vectors; feature extraction; nonlinear partial orthogonal analysis; nonlinear signal; normal rolling bearings; vibrational sensors; Automation; Computer industry; Eigenvalues and eigenfunctions; Electronics industry; Fault diagnosis; Feature extraction; Intelligent sensors; Paper technology; Signal analysis; Spline; B-spline function mapping; fault diagnosis; feature extraction; nonlinear partial orthogonal analysis;
Conference_Titel :
Intelligent Computation Technology and Automation (ICICTA), 2008 International Conference on
Conference_Location :
Hunan
Print_ISBN :
978-0-7695-3357-5
DOI :
10.1109/ICICTA.2008.209