抄録
This paper proposes an image-based robust hovering control for multirotor aerial robot under the effects of uncertainties in the vehicle dynamics which contain external disturbances, nonlinear dynamics, coupling, and parametric uncertainties. The proposed image-based control capable to control the pose (position and orientation) of the aerial robot relative to the fixed-target points, using visual features extracted from the image. The proposed controller scheme consists of two parts: spherical image-based visual servoing (IBVS) controller and a robust flight controller based on robust compensating technique for velocity and attitude control. The robust compensator based on second order robust filter is added to improve the attitude tracking performance when the aerial robot leaves the nominal conditions due to presence of uncertainties. Compared to other methods, the proposed method is robust against uncertainties, does not need to keep features in the field of view, and effective in the environment which pose estimation of target is difficult. The simulation results prove the effectiveness and robustness of the proposed controller and show the high potential in the outdoor flight condition.
