Application of generalized demodulation in bearing fault diagnosis of inverter fed induction motors

When bearing faults occur in induction motor fed with variable frequency power, some fault related characteristic components appear in the stator current. But these components are often submerged by the strong fundamental component of the power supply during the dynamic speed regulation process. In this paper, an approach based on generalized demodulation of rolling bearing fault diagnosis is proposed, and is used in the outer bearing fault. For the constant accelerating operation model of induction motor, after measuring the single-phase current signal, the empirical mode decomposition (EMD) method is used to identify the operation mode and the generalized demodulation is employed in time-frequency transform. Simulation results show that the proposed method can successfully extract the fault frequency and highlight the fault characteristic component from fundamental components.