Estimation of FEM Model Parameters Using Data Assimilation and Its Application to an Electrical Machine

Alexandre Bacchus (1, 2, 3) , Abdelmounaouim Tounzi (2) , Jean-Philippe Argaud (4) , Bertrand Bouriquet (4) , Mélisande Biet (2) , Ludovic Macaire (1, 3) , Y. Le Menach (2)

1 CRIStAL - Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189
2 L2EP - Laboratoire d’Électrotechnique et d’Électronique de Puissance - ULR 2697
3 LAGIS-SI
4 EDF R&D - EDF R&D

The goal of this paper is to identify the parameter set of a given electrical machine. The identification method is based on data assimilation coupled with FEM model. Data assimilation method is an optimization approach that limits the space of candidate parameter sets by centering it around those of the ideal machine. An application to an electrical machine is presented based on the analysis of flux sensor signals. The methodology is validated by using twin experiments that consider simulated data and tested considering measurements extracted from the ”real” alternator.

Mots clés: Data assimilation; electrical machine; measurement probes; numerical simulation

Domaines: Science non linéaire [physics]; Informatique [cs]

Alexandre Bacchus, Abdelmounaouim Tounzi, Jean-Philippe Argaud, Bertrand Bouriquet, Mélisande Biet, et al.. Estimation of FEM Model Parameters Using Data Assimilation and Its Application to an Electrical Machine. IEEE Transactions on Magnetics, 2015, PP (99), pp.1 - 4. ⟨10.1109/TMAG.2015.2495265⟩. ⟨hal-01242419⟩