Title :
Electric motors analysis by means of their characterization using stochastic processing and artificial neural networks of second order
Author :
Ferrera, José Josias Aviles ; Manzano, Mario Alberto Ibarra ; Ojeda, Dora Luz Almanza ; Gutiérrez, Andrés Hernández
Author_Institution :
Fac. de Ingenieria Mecanica, Univ. de Guanajuato
Abstract :
In this article a characterization method of electric motors is presented. This method uses stochastic processing and artificial neural networks of second order to determine if a motor has some failures by means of the characterization of the tests made to the motor and the signals analysis obtained from it. First the signal is filtered to eliminate the noise using a stochastic filter and an artificial neural network to characterize and classify the signal in order to use it in the future as a comparison method for the analysis of the motor. The method and results obtained from the test made to the motor are presented.
Keywords :
electric machine analysis computing; electric motors; failure analysis; filtering theory; induction motors; machine testing; neural nets; signal classification; stochastic processes; electric motors signal analysis; failure analysis; induction motor; noise elimination; second order artificial neural networks; signal classification; stochastic filter; Artificial neural networks; Electric motors; Electrical equipment industry; Equations; Frequency; Neurons; Signal analysis; Stochastic processes; Stochastic resonance; Testing;
Conference_Titel :
Electronics and Photonics, 2006. MEP 2006. Multiconference on
Conference_Location :
Guanajuato
Print_ISBN :
1-4244-0627-7
Electronic_ISBN :
1-4244-0628-5
DOI :
10.1109/MEP.2006.335670
