On the monitoring of the defects of squirrel cage induction motors

When the electric motor is an important element in industrial process in terms of safety and efficiency, we must make an early detection. The earlier the incipient fault is detected the easier remediable faults will be cheaper. Monitoring is made through the spectral analysis of the stator current. Usually the fast Fourier transform is used. However its real achievement is waning in regards of on-line methods like the discrete Fourier transform or sliding Hartley transform. This technique is advisable in order to have an updated spectrum at each sampling time. This is the best method to show up the sidebands in the stator current when an incipient fault occurs meanwhile the motor is operating. Another way is to make a time-frequency analysis. In this case, we can discern the instant where the fault appears always through the spectral analysis of the stator current of the induction motor. Experimental results show the efficiency of the presented method.