Estimation of nascent stage bearing faults of induction motor by stator current signature using adaptive signal processing

In this paper a technique is proposed to estimate the bearing faults of induction motor using sensor less monitoring. Sensor less monitoring is more suitable than conventional vibrating meters. It is very difficult to distinguish the bearing component from its current signature in nascent stage. Therefore in this work bearing fault components are estimated by cancelling the nonbearing fault components in a stator current signature. All the stator current components are treated as noise and cancelled by a wiener filter in real time fashion. Bearing faults with wide range of fault frequencies are estimated using noise cancellation. The results of different bearing faults have been affirmed by this method. Finally a threshold limit is given to intimate the severity of the faults.