Induction Motor Fault Identification using Support Vector Machine

Different induction motor applications have subjected this machine to different conditions that will affect its effective operation. Various types of faults are caused by the environment in which induction motors are installed, which eventually affect the entire operation of the industrial plant, these machines operate. This study aims to develop a more effective method for identifying and classifying faults in three-phase induction motors using Support Vector Machines. The method of Artificial Intelligence (AI), which incorporates the Support Vector Machines (SVM) algorithm, is used to classify nine conditions of fault in induction motors, namely the unbalanced supply voltages, over-voltage, under-voltage, blocked rotor, line-to-line short circuit, line-to-earth short circuit, single phasing, mechanical overload, and a healthy motor. A SIMULINK simulation of an induction motor will be created to simulate the faults independently, and the results will be trained to the SVM model. Using an Orange tool, the SVM method was used in conjunction with the one-versus-all multiclass approach and the Radial Basis Function (RBF) kernel.