Robust observer based model predictive control of a 3-DOF helicopter system

Abstract

Helicopter systems are characterized by highly nonlinear dynamics, multiple operating regions, and significant interaction among state variables. In this paper, an observer based model predictive control (MPC) scheme with successive linearization is presented, for a 3 degree of freedom (DOF) helicopter system. All control simulations were performed under the conditions of noisy measurements. To illustrate the advantage by using unscented Kalman filter (UKF) as the observer, the performance of UKF based MPC is compared with those of MPC algorithms using linear filters and extended Kalman filter (EKF). The simulation results have shown that for this application the UKF-based MPC has superior performance, in terms of the disturbance rejection and set-point tracking. © 2013 Springer Science+Business Media Dordrecht.