On The Modeling And Estimation Of Nonlinear Instantaneous Frequency And Amplitude Of Non-Stationary Signals

Meryem Jabloun (1) , Nadine Martin (1)

1 GIPSA-SA-IGA - GIPSA - Signal et Automatique pour le Diagnostic et la Surveillance

Advanced signal processing methods are helpful in the context of diagnostic and fault detection of induction machines. They provide accurate information about highly nonlinear systems, which need complex physical models. We recall and compare some available methods in modeling and/or detection of nonlinear and nonstationary signals in the time and frequency domain. We particularly outline a new method we developed, which is based on a local modeling of the instantaneous amplitude and frequency of the signal. This method reaches significant performance especially when signals are embedded in strong noise and present a high variation of the instantaneous frequency.

Mots clés: Spectrogram; Non-stationary signal; nonlinear amplitude and frequency modulations; polynomial phase signal; time-frequency distribution; Spectrogram.

Domaines: Traitement du signal et de l'image [eess.SP]; Traitement du signal et de l'image [eess.SP]

Meryem Jabloun, Nadine Martin. On The Modeling And Estimation Of Nonlinear Instantaneous Frequency And Amplitude Of Non-Stationary Signals. IAR 2007 - 5th IAR Workshop on Advanced Control and Diagnosis, Nov 2007, Grenoble, France. ⟨hal-00200314⟩