Observer-embedded L-2-gain control

This paper demonstrates that the H-infinity controller of Doyle ct al. [1] can be reinterpreted as an observer-embedded L-2-gain controller. The embed ded observer is similar to the conventional Luenberger observer except havi ng an additional state-dependent calibration term that compensates fur exog enous inputs and model uncertainties. Based on tills fact, we derive mixed Linear Matrix Inequalities, Hamiltonian Matrix, and Linear Parameterisation to provide solutions to feasible observer-embedded L-2-gain controllers. ( C) 2001 Elsevier Science Ltd. All rights reserved.