Fault Diagnosis of Induction Motor Based on wavelet packet analysis and negative selection algorithm

摘要

For the spectrum characteristics of the motor vibration,the vibration fault diagnosis method of motor rotor rubbing based on wavelet packet transform and real-valued negative selection algorithm is put forward.Using wavelet packet analysis to energy analysis of rotor rubbing and extracting the fault feature vectors,then by real-valued negative selection algorithm to identify the normal and failure mode eigenvectors.The experimental results show that with this method all the rotor rubbing faults can be detected comprehensive and rapidly.This method has high feasibility of the wavelet packet analysis and real-valued negative selection algorithm in the rotor rubbing fault diagnosis.