(logo by Janette Forte)
This is the official site of the UT Austin Villa 3D Simulation team from the Department of Computer Science at the University of Texas at Austin.
This web page provides supplementary material to the following paper:
Using Dynamic Rewards to Learn a Fully Holonomic Bipedal Walk
Patrick MacAlpine and Peter Stone
the AAMAS Adaptive Learning
Agents (ALA) Workshop in Valenica, Spain, June 2012.
The full paper can be found here
This page provides details on the optimization of walk parameters for
a fully holonomic walk (omnidirectional walk which can walk at the
same maximum velocity in all directions). The walk was learned using
optimization and the CMA-ES policy search algorithm. Rewards for
walking speed in the three cardinal directions of forward, backwards,
and sideways were dynamically reweighted from one generation to the
next during learning in order to encourage equal walking speeds in all directions.
Fully Holonomic walk able to move in all directions with equal maximum
Download video: ogv, mp4
For any questions, please
contact Patrick MacAlpine.