Electric motor faults diagnosis using artificial neural networks

摘要: This paper presents electric motor fault diagnosis using two kinds of Artificial Neural Networks (ANN): feedforward networks and self organising maps (SOM). Major faults such as bearing faults, stator winding fault, unbalanced rotor broken bars are considered. The ANNs were trained tested measurement data from currents mechanical vibration signals. effects different network structures the training set sizes on performance discussed. study shows that ANN with a very simple internal structure can give satisfactory results, while SOMs classify type during steady state working conditions. experiment results also show is more promising scheme in this case where motors available.