Title :
Vibro-acoustic fault detection and diagnosis in hybrid electric vehicle
Author :
Salim, G. ; Ouadie, Bennouna ; Ghaleb, Hoblos
Author_Institution :
Autom. & Syst. Lab., Inst. de Rech. en Syst. Electroniques Embarques, Rouen, France
Abstract :
The aim of this paper is to study the effects of faults on vibration and noise spectrum of the Permanent Magnet Synchronous Motor (PMSM) of a hybrid vehicle. In this work, a new approach is used by adding the acoustic noise. This vibro-acoustic modeling allows discovering the influence of faults on spectrum features. Then, a feed forward neural network based on Levenberg-Marquardt training is used for classification. Finally, all the technique is implemented on the PMSM of a hybrid vehicle.
Keywords :
acoustic noise; fault diagnosis; feedforward neural nets; hybrid electric vehicles; permanent magnet motors; power engineering computing; synchronous motors; Levenberg-Marquardt training; PMSM; acoustic noise; feed forward neural network; hybrid electric vehicle; hybrid vehicle; permanent magnet synchronous motor; vibro-acoustic fault detection; vibro-acoustic fault diagnosis; Acoustics; Circuit faults; Fault detection; Force; Permanent magnet motors; Stators; Vibrations; Fault detection and diagnosis; PMSM; analytical model; electromagnetic noise; neural network; vibrations;
Conference_Titel :
Power Engineering, Energy and Electrical Drives (POWERENG), 2013 Fourth International Conference on
Conference_Location :
Istanbul
DOI :
10.1109/PowerEng.2013.6635625
