Learning Optimal Robot Ball Catching Trajectories Directly from the Model-based Trajectory Loss

Arne Hasselbring, Udo Frese, Udo Frese, Thomas Röfer, Thomas Röfer

2022

Abstract

This paper is concerned with learning to compute optimal robot trajectories for a given parametrized task. We propose to train a neural network directly with the model-based loss function that defines the optimization goal for the trajectories. This is opposed to computing optimal trajectories and learning from that data and opposed to using reinforcement learning. As the resulting optimization problem is very ill-conditioned, we propose a preconditioner based on the inverse Hessian of the part of the loss related to the robot dynamics. We also propose how to integrate this into a commonly used dataflow-based auto-differentiation framework (TensorFlow). Thus it keeps the framework’s generality regarding the definition of losses, layers, and dataflow. We show a simulation case study of a robot arm catching a flying ball and keeping it in the torus shaped bat. The method can also optimize “voluntary task parameters”, here the starting configuration of the robot.

Paper Citation

in Harvard Style

Hasselbring A., Frese U. and Röfer T. (2022). Learning Optimal Robot Ball Catching Trajectories Directly from the Model-based Trajectory Loss. In Proceedings of the 19th International Conference on Informatics in Control, Automation and Robotics - Volume 1: ICINCO, ISBN 978-989-758-585-2, pages 201-208. DOI: 10.5220/0011279000003271

in Bibtex Style

@conference{icinco22,
author={Arne Hasselbring and Udo Frese and Thomas Röfer},
title={Learning Optimal Robot Ball Catching Trajectories Directly from the Model-based Trajectory Loss},
booktitle={Proceedings of the 19th International Conference on Informatics in Control, Automation and Robotics - Volume 1: ICINCO,},
year={2022},
pages={201-208},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0011279000003271},
isbn={978-989-758-585-2},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 19th International Conference on Informatics in Control, Automation and Robotics - Volume 1: ICINCO,
TI - Learning Optimal Robot Ball Catching Trajectories Directly from the Model-based Trajectory Loss
SN - 978-989-758-585-2
AU - Hasselbring A.
AU - Frese U.
AU - Röfer T.
PY - 2022
SP - 201
EP - 208
DO - 10.5220/0011279000003271