Wavelet-network for classification of induction machine faults using optimal time-frequency representations

• Published in: 2011 7th International Conference on Electrical and Electronics Engineering (ELECO) pp. I-358 - I-362
• Main Authors: Boukra, T., Lebaroud, A., Medoued, A.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.12.2011
• Subjects: Artificial neural networks; Classification algorithms; Feature extraction; Kernel; Rotors; Stators; Training
• ISBN: 1467301604; 9781467301602

This paper presents a new diagnosis method for classifying current waveform events that are related to a variety of induction machine faults. The method is composed of two sequential processes: feature extraction and classification. The essence of the feature extraction is to project a faulty machine signal onto a low dimension time-frequency representation (TFR), which is deliberately designed for maximizing the separability between classes. A distinct TFR is designed for each class. The performance of fault classification is presented using two types of classifiers namely the Wavelet Neural Network (WNN) and the classical Artificial Neural Network (ANN) with Levenberg Marquardt algorithm. The flexibility of this method allows an accurate classification independently from the level of load. This method has been validated on a 5.5-kW induction motor test bench.