A Montecarlo reactive navigation algorithm for a dual arm aerial robot

Abstract

In this paper, we propose a reactive algorithm for a dual arm aerial robot during its navigation phase. This algorithm takes into account not only the aerial platform dynamics but also the constraints imposed by the dual-arm manipulator. The UAV is fully and continuously controlled using velocity commands but the arms are only allowed to move between a set of predefined configurations. The arms configurations are bounded in order to achieve a reduced solution space for collision checking. The reactive navigation algorithm is based on commands from a higher-level path planner and receives as input point-cloud sensor readings. This work is within the framework of the AEROARMS project, so we consider the dual-arm aerial platform developed in the project. Thus, a simple but realistic model for velocity dynamics of the mentioned UAV has been obtained by means of real data, and the proposed method has been tested in a SITL simulation framework.