Methodology for filtering and tracking frequency-changing components during motor start-up

Induction Motors (IM) are widely used in industrial applications, and they represent the 80% of the consumed energy in the industry. IM faults are unavoidable and cause several problems such as production time delays, money looses, etc. For this reason computationally efficient methodologies for IM fault detection are necessary. Many efforts have been carried out in this field. However, easy-to-implement methodologies still requires attention. In this work a FFT-based methodology for the analysis of non-stationary current signals of IM is proposed. The feasibility of the proposed methodology for being used as basis for the IM condition monitoring is demonstrated by means of simulation and experimentation. Obtained results demonstrated that the methodology can filter and track frequency-changing components during the start-up of IM with one broken bar.