Self-scheduled robust decoupling control with [formula omitted] performance of hypersonic vehicles

Abstract

This paper is concerned with self-scheduled robust decoupling control for hypersonic vehicles subject to both parameter uncertainty and external disturbance. The overall design procedure includes three loops to be closed in sequence, where the robust decoupling of the middle-loop system is most important and challenging. The decoupling problem of middle loop is firstly reformulated as the robust H∞ control of a general uncertain linear parameter-varying (LPV) error system, in which the parameters are the command signals. Then, through transforming the robust H∞ problem of the uncertain LPV error system into a certain one, the tensor-product (TP) model transformation technique can be directly employed to obtain a convex polytopic form, which greatly decreases the computing load of the controller design. Hence, the controller is obtained by solving limited number of linear matrix inequalities (LMIs) corresponding to the vertex systems. The resulting controller is time-varying and self-scheduled. Moreover, the decoupling is guaranteed against parameter uncertainty, whereas approximate decoupling can only be achieved when an external disturbance exists. Nevertheless, the degree of approximate decoupling can be enhanced by choosing a small performance index γ. The efficacy of the proposed method is supported by simulations.