Sampling-based receding horizon collision-free control for a class of Micro Aerial Vehicles

A novel sampling-based receding horizon control strategy that guarantees collision-free navigation for a class of aerial robots is the topic of this paper. The proposed approach combines the concepts of receding horizon control and sampling-based navigation strategies in order to derive a model-based control framework, which respects input and state constraints, and achieves avoidance of any known obstacle while remaining computationally lightweight even for systems of high-order and complex, convex or non-convex obstacles and long prediction horizons. The control law is applied for the case of a multirotor Micro Aerial Vehicle that optionally also employs its capacity to direct its thrust via a rotors´ tilting mechanism. Extensive simulation studies indicate that high performance collision-free navigation is achieved and reasonably long prediction horizons can be handled while remaining applicable for on-board deployment.