IQC-based gain-scheduled control of linear periodic systems using output feedback

Abstract

In this thesis, output feedback control of linear periodic systems is studied. Themain approach is usign a gain-scheduled controller, where the synthesis of the controlleris based on linear parameter-varying control theory and an integral quadratic constraint (IQC) result for linear periodic systems. Existence conditions for a stabilizing controllerare given in terms of finite-dimensional linear matrix inequalities (LMI's). For illustration of the application of the theory, the tracking control problem for a two-link arm system with a periodically excited base is studied. This problem is turned into an L2-gain minimization problem, which, using the aforementioned theories, canbe cast as a minimization problem over a convex set defined by finite-dimensional LMI's. Using weighting functions, different controllers are designed for tracking different reference signals, in particular a unit step and a sinusoid. Nonlinear simulations are made for the closed-loop systems. The results show that the IQC result can besuccessfully applied to the output feedback control of linear periodic systems.