Generación de trayectorias de marcha para un robot humanoide a partir de captura de movimiento

Abstract

In this research paper, we propose a method to adapt the gaits trajectories of a human being to a humanoid robot. To achieve this goal, first we made an experiments with with two different motion capture systems to obtain the joint trajectories of human walking. The trajectories obtained by motion capture serve as input to a trajectory generator off-line, called dynamic filter that takes into account the kinematic and dynamic constraints necessary to prevent the robot fall while walking. To validate the trajectories, a simulator was used for the robot Bioloid Premium Kit based on the V-Rep environment. After, these were implemented on the real robot. Even though the motion generation is done offline, the dynamics filter is a good option for making in automated way the trajectories generation.