Abstract
A new effective approach for fault detection and diagnosis (FDD) technique for on-line condition monitoring (CM) of stator winding faults due to insulation failure in industrial induction machine is presented in this paper. FDD techniques are broadly classified into three categories; signature-extraction, model and knowledge-based approach. Recent advancement in computing performance and sensor technology permit the use of advanced systems to achieve this objective. Computer-aided advanced diagnosis system uses pattern recognition which is mainly comprises of feature extraction, feature selection and feature classification respectively. Introduction of wavelets in signature-based feature extraction process along with feature classification using knowledge-based approach is used for a reliable automatic identification of stator winding faults. Based on mathematical modeling for stator winding inter-turn fault; a simulink model is designed to validate mathematically obtained results and it is seen that results are in good agreement with the experimental results.

This publication has 12 references indexed in Scilit: