An MPC-based Strategy Design for an Autonomous Aerial Robot

In this paper, a strategy design problem is investigated for an aerial robot which is required to autonomously select and interact with ground targets so as to complete a specified task in a complex envirioment with moving obstacles. First, the kinematic models of the aerial robot, ground targets and obstacles. Then, an MPC-based strategy design is proposed in which the current state of the aerial robot, the state prediction of the targets and obstacle avoiding are considered. Since the motion periods of targets are different which leads to incorrect estimations, an IMM filter is given to obtain more precise future state and motion period of the targets. Finally, simulation and experimental results show the effectiveness of the proposed method.