Active Modeling Based Control for Aerial Manipulator Systems

Expanding an unmanned aerial vehicle (UAV) with an equipped robotic arm increases the ability of the UAV to actively interact with the environment. During the process of active tasks, precise motion control under model uncertainties and external disturbances is a challenging issue. To solve this issue, an active modeling based control scheme is proposed in this paper. In this scheme, the aerial manipulator system (AMS) is modeled with consideration of various uncertainties represented by the model error and the process noise. The model error caused by the gravity and movement of the manipulator is estimated by the unscented kalman filter (UKF). The estimation is further integrated into a nominal controller to achieve high-precision motion control. Detailed experimental results on the self-bulit testbed illustrate the effectiveness and good performance of the proposed method.