An integrated approach to the design of an aircraft gain scheduled controller

Abstract

This thesis addresses the problem of integrated design of the aircraft plant parameters and of the corresponding feedback controller. The plant parameters are typically the sizes of the control surfaces or other aerodynamical surfaces of the aircraft. The approach is to rewrite the aircraft dynamic requirements as linear matrix inequalities (LMI's) and to optimize a linear cost function associated with aircraft plant parameters, while meeting the LMI constraints. An algorithm using Matlab and LMI-Lab has been developed. This algorithm has been used for integrated plant/controller design of an F-4 aircraft at five different flight conditions. The result consists of a set of controllers one for each flight condition, and a single solution for the optimal sizes of the F-4 stabilator and spoiler control surfaces.