Abstract

In this work we propose an online trajectory planner for humanoid walking. It is based on the observer trajectory planning problem where the moving observer (humanoid) should maneuver optimally to better estimate the position of the target (door). Thus, the uncertainty of the door location within the humanoid’s field of view is considered. In particular, we propose a terminal stochastic controller that recursively applies the Unscented Kalman Filter (UKF) over a time horizon for evaluating a set of objective functions. The aim is to minimize the estimation error of the door location while maintaining it inside the humanoid’s field of view, and avoiding collisions between the humanoid and the door frame. The output is the sequence of walking velocity references. In addition, we compute the humanoid’s odometric localization based on the UKF to provide the instantaneous humanoid location, which is an input to the trajectory planner. We validate the stochastic controller with a set of experiments in a scenario with doors using a real humanoid NAO, equipped with an RGB-D sensor mounted on its head.11The source code in ROS C++ is available at: https://sites.google.com/site/gustavoarechavaleta/sochum.

Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call

Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.