Probabilistic Gradient Ascent with Applications to Bipedal Robot Locomotion
Paper Details
- Title: Probabilistic Gradient Ascent with Applications to Bipedal Robot Locomotion
- Authors:
- David Budden
- Josiah Walker
- Madison Flannery
- Alexandre Mendes
- Keywords:
- Robotics
- Robotic Soccer
- Optimisation
- Bipedal Locomotion
- Reinforcement Learning
Abstract
Bipedal robotic locomotion is an emerging field within the multi-billion dollar robotics industry, with global initiatives (such as RoboCup, FIRA and the DARPA Robotics Challenge) striving toward the development of robots able to complete complex physical tasks within a human-engineered environment. This paper details the redevelopment of an omnidirectional walk engine for the DARwIn-OP,
with an improved online optimisation framework developed for 13 of its internal parameters. Applying two well-known optimisation algorithms within this framework yields significant improvement in walk speed and stability. A new non-convex optimisation algorithm (Probabilistic Gradient Ascent) is derived from a reinforcement learning framework and applied to the same task, yielding an average speed improvement of 50.4% and setting a new maximum speed benchmark of 34.1 cm/s.
Publication Details
- Status: Accepted
- Date Accepted: 17/010/2013
- Conference: Australasian Conference on Robotics and Automation (ACRA) 2013
How to Cite (BibTeX)
{budden2013probabilistic,
title={Probabilistic gradient ascent with applications to bipedal robot locomotion},
author={Budden, D. and Walker, J. and Flannery, M. and Mendes, A.},
booktitle={Australasian Conference on Robotics and Automation (ACRA)},
year={2013}
}
Supplementary Materials
- Original manuscript available here
- The following video provides a demonstration of the experimental results