15th Triennial World Congress of the International Federation of Automatic Control
  Barcelona, 21–26 July 2002 
ROBUST H-INFINITY REDUCED ORDER FILTERING FOR UNCERTAIN BILINEAR SYSTEMS
H. Souley Ali, M. Zasadzinski, H. Rafaralah y and M. Darouac h
CRAN, IUT de Longwy, Université Henri Poincaré – Nancy I
186, rue de Lorraine, 54400 Cosnes et Romain, FRANCE
e-mail: {souley,mzasad}@iut-longwy.uhp-nancy.fr

This paper investigates H reduced order unbiased filtering problems for a nominal bilinear system and a bilinear system affected by norm-bounded structured uncertainties in all the system matrices. First, the unbiasedness condition is derived. Second, a change of variable is introduced on the inputs of the system to reduce the conservatism inherent to the filter stability requirement and to treat the product of the inputs by the disturbances. Then the solution is expressed in terms of LMI by transforming the problem into a robust state feedback in the nominal case and a robust static output feedback in the uncertain case.
Keywords: Bilincar systems, Robust estimation, H-infinity optimization, Uncertainty
Session slot T-Fr-A08: Nonlinear estimation/Area code 2c : Non-linear Systems