Real-time implementation of the sliding DFT applied to on-line's broken bars diagnostic

Electric motors have a very important role for safe and efficient running in many industrial or power plant. Early detection of abnormalities in the motors contributes to avoid expensive failures. It has been shown that rotor faults can be detected thanks to the spectral analysis of the stator current using FFT/DFT algorithms. They are usually performed and analysed offline after data acquisition. The proposed method allows to obtain the DFT immediately after the arrival of each new sample. This technique is suitable for the online running spectral analysis of the stator current in order to detect the components due to rotor broken bars while the machine is operating. Experimental results show the efficiency of the proposed method for an online diagnostic.