H∞ dynamic output feedback for LPV systems subject to inexactly measured scheduling parameters

This paper proposes a design procedure for reduced-order dynamic output feedback (DOF) gain-scheduling controllers with H-infinity guaranteed cost for linear parameter-varying (LPV) continuous-time systems, where the measurement of the scheduling parameters may be affected by uncertainties. Thanks to the flexibility of the proposed modeling, the LPV-DOF controllers can be implemented in terms of a selected set of parameters, which are supposed to be available for measurement in real time. The design conditions can cope with both additive and multiplicative noises, considered as time-varying uncertainties, affecting the measures. All parameters and uncertainties are modeled through the multi-simplex framework, i.e., the Cartesian product of simplexes. The problem is solved through a two-stage procedure based on linear matrix inequalities.