Fusion of Vibration and Current signals for Improved Multi Class Fault Diagnosis of Three Phase Induction Motor

Immense advancement in Artificial Intelligence and Machine Leaning, data driven condition monitoring and fault diagnosis of three phase Induction Motor (IM) are gaining popularity as an integral part of industrial informatics and automation. Bearing faults are most commonly occurring mechanical defects in three phase induction motor and can easily be detected by motor frame vibration. On the other hand, it has been observed that the electrical faults have direct impact on stator current as compared to mechanical blemishes and they are well detected through the stator current signature analysis. With the scenario of variable speed induction motor drives with power electronics converters, stator winding is susceptible to frequent failures due to adverse impact of various harmonics contents. This paper reports a novel investigation of multi class fault diagnosis of IM with fusion of motor frame vibration and stator current signal to achieve improved classification accuracy both for the electrical and mechanical faults. Random Forest (RF) based machine Learning framework has been used to demonstrate the efficacy of the proposed fusion method.