Meaningful Learning Processes of Service Robots for Tracking Trajectories through Virtual Environments.

Abstract

This paper is devoted to the study and control of an aerial manipulator robot (AMR) to perform tracking tasks autonomously, in order to apply LQR linear system control algorithms and application methods in a 3D virtual environment. Prior to obtaining a linearized kinematic model of the robotic systems, which allows to perform missions that require both navigation and manipulation capabilities in partially structured areas or environments. Through the use of the advanced control algorithm, a virtualized environment was developed in a 3D simulator for educational processes as a form of testing, which allows to evaluate the movement and evolution of the control errors, both for verification and visualization of the RMA behavior. Finally, the stability and robustness of the proposed RMA control is tested and experimentally analyzed using the DJI Matrice 600 Pro UAV tethered to an anthropomorphic 3DOF robotic arm. Therefore, these results are exposed and discussed to validate the proposed controller and ensure its correct operation.