Non-Uniform Time Resampling for Diagnosing Broken Rotor Bars in Inverter-Fed Induction Motors

Fault detection in inverter-fed induction motors (IMs) is an actual industrial need. Many line-fed machines are being replaced by inverter-fed drives for improving control during startup and also for saving energy. Broken rotor bars (BRBs) in IMs is one of the most difficult faults to be detected, particularly when the motor is fed by an inverter in a soft startup. The difficulty of detecting BRBs is that the characteristic fault-related frequencies are very close to the fundamental frequency, and the amplitude of the fundamental is significantly higher than the fault-related frequency components. This paper proposes an effective method that allows the detection of the BRB fault in inverter-fed IMs during a soft startup transient based on a non-uniform resampling algorithm. The proposed algorithm transforms the nonstationary fundamental frequency into a stationary component by non-uniform resampling, whereas the fault-related components are considerably separated from the fundamental one, making easier to follow their evolution during the startup transient. Simulation and experimental results demonstrate the effectiveness of the proposed method to detect the fault.