Improved Diagnosis of Induction Motor's Rotor Faults using the Papoulis Window

This paper discusses the rotor faults diagnosis in induction motors through stator current analysis based on the estimation of the Power Spectral Density (PSD) using the Periodogram technique. This approach is the most used technique in industry thanks to its simplicity of programming and its relatively low computation time compared to high resolution signal processing methods. However, this method has some disadvantages such as a frequency resolution depending on acquisition time and an almost impossible identification of low harmonics (case of incipient faults). This second disadvantage is due to the side lobes effect related to the selected weighting window, such as Blackman, Hamming, Hanning or Kaiser. To reduce this effect, a new weighting window named Papoulis window is proposed in this paper. To highlight the performance of this new window in rotor faults diagnosis of induction motors, several simulation tests are achieved and discussed.