Condition Monitoring of Induction Motors: Use of Artificial Neural Networks in fault diagnosis, by analysis of data collected from vibration and current sensors

Induction motors are widely used in present day industries, on account of its ruggedness,low cost and easy speed control.Condition monitoring of induction motors-the process by which certain parameters are continuously observed and analyzed for early fault detection,has become a vital part of machine maintenance.In this work,Artificial Neural Networks(ANNs)have been used for condition monitoring of 3-phase induction motors.RBF and FFBP neural nets were used for comparative analysis.Data collected from vibration and current sensors were processed and fed as inputs to the ANN. Continuous Wavelet Transform(CWT) and Park's Transform were used for processing. Faults such as bearing fault, broken rotor bar defect and stator winding unbalance faults have been dealt with. A multi-class approach using ANN has been attempted. This book looks into the process by which neural nets may be used to detect faults, which should help researchers,practicing engineers involved in industrial processes using induction motors and anyone else who is interested in using measured parameters in machines, to predict the occurrence and nature of fault.