H∞ FILTER DESIGN FOR LINEAR PARAMETER-VARYING SYSTEMS USING PARAMETER-DEPENDENT LYAPUNOV FUNCTIONS

This paper considers full-order gain-scheduled H∞ filters for Linear Parameter-Varying (LPV) systems and proposes a method for their design using Parameter-Dependent Lyapunov Functions (PDLFs) via parametrically affine Linear Matrix Inequalities (LMIs). The state-space matrices of LPV systems and PDLFs are both parameter-dependent in negative as well as positive power series of parameters. Although our method has great computational complexity compared to existing methods, it encompasses some of them and is more effective than those, which is demonstrated with a numerical example.