Performance comparison based selecting an optimal method for fault detection and diagnosis methods for induction motor

Abstract

One of the basic and common electrical machines is Induction Motors. Induction motors are highly used in the industry due to more technical and economic reasons. Because of different operating conditions, the induction motors are facing various stresses. These stresses provide different faults and failures in the induction motors. So that it is essential to monitor the conditions of the machine during operating conditions avoid disastrous faults. There are various kinds of fault monitoring methods which are classified as signal processing based, modelling based and artificial intelligence technique-based methods. In signal processing, signals are processed and faults are identified in the machine. In model-based methods, correct models are designed for faulty machine to diagnose the faults. One of the problems is designing accurate faulty model is difficult. Artificial intelligence methods can provide improved accuracy in fault detection. Neural Networks, Recurrent Neural Network, Fuzzy Neural Network, Radial Basis Function Neural Network and ANFIS are some of the soft computing methods used for fault detection in induction motors. These methods can able to locate the fault in the circuit in the stator winding. newline