Feature Extraction Technique And The Application In Detection Of Electric Motor Faults

How to identify motor incipient failure has become one research hotspot of the engineering area. Here, the key is to extract the fault feature. It is this paper of great significance from the theoretical and practical points of view. What focuses on is feature extraction technique and the application in detection of electric motor faults. Bilinear time-frequency analysis together with adaptive filter techniques are used to analyze and detect motor bar broken fault stator current. Through this method it extracts the feature frequency components from the time-frequency spectrum. It verifies that the method is feasibility and effective. Another method for detecting faults in motors based on the multiple signal classification(MUSIC) algorithm is presented. The MUSIC algorithm is higher in resolution of frequency and more accurate in fault detection compared with FFT analysis method. Simulation results show that the fault characteristic components can be obtained accurately through the methods presented in this paper, and that the feasibility of the method is confirmed. So it has advantages in detection faults under the complicated background.