Artificial neural network-based induction motor fault classifier using continuous wavelet transform

Abstract

Induction motors are used in industrial, commercial and residential applications because they have considerable merits over other types of electric motors. These motors are used in various operating stresses that give rise to faults. Most recurrent faults in induction motors are bearing faults, stator interturn faults and cracked rotor bars. Early detection of induction motor faults is crucial for their reliable and economical operation. This could be done by motor monitoring, incipient fault detection and diagnosis. In many situations, failure of critically loaded machine can shut down an entire industry process. This growing demand for high-quality and low-cost production has increased the need for automated manufacturing systems with effective monitoring and control capabilities. Condition monitoring and fault diagnosis of an induction motor are of great importance in the production line. It can reduce the cost of maintenance and risk of unexpected failures by allowing the early detection of catastrophic failures. This work documents experimental results for multiple fault detection in induction motors using signal-processing and artificial neural network-based approaches. Motor line currents recorded under various fault conditions were analyzed using continuous wavelet transform. A feedforward neural network was used for fault characterization based on fault features extracted using continuous wavelet transform.