Parameter-dependent robust control for a rotorcraft UAV

Abstract

This work presented in this publication is part of a project that aims to design and implement controllers that make it easier to operate helicopter UAVs. We have chosen for a goal-oriented control concept for the design of an inner-loop controller that provides velocity tracking control. In this paper, modelling and control design techinques are presented that focus on robust and parameter-dependent control. Numerical linearization of a nonlinear helicopter model for a number of points in the ight regime that consists of forward an backward ight yields a set of linear models. A root-locus plot of the linear helicopter models reveals a strong variation of the system dynamics. The variation of the stability and control derivatives is captured in two different models: an element-wise undertainty model and a parameter-dependent model. Both models are represented as LFTs and used in a control synthesis architecture. The two uncertainty models are evaluated by means of the local closed-loop performance of the resulting H1-controllers. The parameter-dependent model yields a controller which has better closed-loop performance indicating that the model is less conservative than the element-wise uncertainty model. This model is also used to design - and LPV-controllers. Unfortunately, even these more advanced control design techniques are unable to cope with a signicant crosscoupling effect that is linked with the dependency on forward speed. Improvement in performance is expected when parameter rates are included in the LPV synthesis.