| Throughout development of fault diagnosis, in order to obtain good diagnosis result, processing of fault data is essential, and for sake of good analysis effect, method is important too. Based on wavelet analysis theory, and aiming at fault data character extraction in induction motor mixed fault diagnosis, this essay dedicates on implication of wavelet analysis in this field, and gain good effect in experiment. First we discuss the selected topic and background of this paper, and then go into the correlative fundamental theory of wavelet transform. We mainly focus on signals of induction motor with eccentricity fault and mixed fault of eccentricity plus rub and related algorithm. It mainly includes following aspects:In this paper, firstly implication of wavelet analysis in signal harmonic component extraction is presented in detail, and on this foundation signal given frequency band component extraction method is proposed.Then we investigate frequency aliasing phenomenon, reason why frequency aliasing phenomenon is easy happened when wavelet is implied in field of motor fault diagnosis, and frequency bands where frequency aliasing phenomenon is easer happened, and put forward approved Morlet wavelet analysis method which can effectively alleviates frequency aliasing phenomenon.Third we set up vibration signal detected system, design experiments to obtain three groups of fault signals: eccentricity fault, mixed fault of eccentricity plus mild rub, mixed fault of eccentricity plus moderate rub, and process these data using analysis method proposed in this paper. Character parameter extracted this way can clearly distinguish these two patterns of fault.Finally we investigate mixed program method with MATLAB and LabVIEW, and accomplish software exploitation of motor fault diagnosis detection system. The system fulfills following functions: signal de-noising, signals analysis using wavelet and extraction of character parameters; display analysis result and give out tocsin.
|
PDF Full Text Request