Diagnostics of Induction Motor Based on Spectral Analysis of Stator Current with Application of Backpropagation Neural Network

Abstract

Nowadays PC computers make possible the signal characteristics by calculation. The paper presents an automatic computerized system for the diagnosis of the rotor bars of the induction motor by applying spectral analysis and backpropagation neural network. Software to recognize the current of induction motor was implemented. System of current recognition is based on a study of the frequency spectrum of stator current. The studies were conducted for two conditions of induction motor. The results of the numerical experiments are presented and discussed in the paper. The researches show that the system can be useful for protection of the engines and metallurgical equipment.