Reliable guaranteed cost control for singularly perturbed uncertain systems

This paper studies the reliable guaranteed cost control of singularly perturbed systems with parameter uncertainties. To avoid the stiffness and for the sake of parallel computation, the full-order algebraic Riccati equation is decomposed into two reduced-order pure slow and pure fast algebraic Riccati equations. The designed control laws guarantee the robust stability and known quadratic performance bounds of the closed-loop systems despite any actuator failures within a selected subset of actuators. Examples are given to demonstrate the design algorithms addressed in this paper.