Enhanced LQR control for unmanned helicopter in hover

Real time adaptability is of central importance for the control of unmanned helicopter flying under different circumstances. In this paper, an active model is employed to handle the time varying uncertainties involved in the helicopter dynamics during flight. In the scheme, a normal LQR control designed from a simplified model at hovering is enhanced by means of unscented-Kalman-filter (UKF) based estimation, which tries to online capture the error between the simplified model and the full dynamics. This is intended to achieve adaptive performance without the need of adjusting the controller modes or parameters along with the changing dynamics of helicopter. Simulations with respect to a model helicopter are conducted to verify both the UKF-based estimation and the enhanced LQR control. Results are also demonstrated with the normal LQR control with the active model enhancement.