Artificial Neural Networks for Automatic Classification of Induction Machine Faults

Authors

Abla Bouguerne
Department of Electrical Engineering, Echahid Cheikh Larbi Tebessi University, Algeria
Aziz Boukadoum
Department of Electrical Engineering, Echahid Cheikh Larbi Tebessi University, Algeria
Mohamed Salah Djebbar
Department of Electrical Engineering, Echahid Cheikh Larbi Tebessi University, Algeria

Synopsis

In this paper, new classification method of induction motor faults is proposed that permit to give a best solution to the classification problem. This method based on Time–frequency representation for classification of the current waveforms. It is composed of two steps: feature extraction and rule decision. Artificial neural networks (ANN) are used for decision criterion. The diagnosis allows the detection of bearing fault, stator fault and the rotor fault. The flexibility of this method allows an accurate classification independent from the level of load. This method is validated on a 5.5-kW induction motor test bench.

ICAECE2023
Published
February 5, 2024