Fault diagnosis of motor based on wavelet analysis and antibody memory clonal algorithm

A new approach based on a wavelet analysis and antibody clonal algorithm is proposed for stator, rotor, bearings eccentric, and air gap eccentric fault diagnosis of motor. Stator current monitoring data are pre-processed and decomposed by wavelet technique. The energy of the band that extracted by wavelet analysis is normalized as eigenvectors of the fault diagnosis. Antigens are the fault eigenvectors, and antibodies are clustering center established by immune algorithm. Then antibody memory clonal algorithm begins to identify fault classification.The experiment and application results show that the wavelet analysis and memory clonal algorithm may classify the working condition of motor, and it has higher accuracy and enrichment rate.