Please use this identifier to cite or link to this item: http://dx.doi.org/10.14279/depositonce-6208
Main Title: Transfer learning of gaits on a quadrupedal robot
Author(s): Degrave, Jonas
Burm, Michael
Kindermans, Pieter-Jan
Dambre, Joni
Wyffels, Francis
Type: Article
Language Code: en
Abstract: Learning new gaits for compliant robots is a challenging multi-dimensional optimization task. Furthermore, to ensure optimal performance, the optimization process must be repeated for every variation in the environment, for example for every change in inclination of the terrain. This is unfortunately not possible using current approaches, since the time required for the optimization is simply too high. Hence, a sub-optimal gait is often used. The goal in this manuscript is to reduce the learning time of a particle swarm algorithm, such that the robot's gaits can be optimized over a wide variety of terrains. To facilitate this, we use transfer learning by sharing knowledge about gaits between the different environments. Our findings indicate that using transfer learning new robust gaits can be discovered faster compared to traditional methods that learn a gait for each environment independently.
URI: https://depositonce.tu-berlin.de//handle/11303/6845
http://dx.doi.org/10.14279/depositonce-6208
Issue Date: 2015
Date Available: 23-Oct-2017
DDC Class: 590 Tiere (Zoologie)
150 Psychologie
Subject(s): Quadrupedal robot
gait optimization
transfer learning
particle swarm optimization
evolutionary algorithms
Sponsor/Funder: EC/FP7/248311/EU/Adaptive Modular Architecture for Rich Motor Skills/AMARSi
Usage rights: Terms of German Copyright Law
Journal Title: Adaptive behavior
Publisher: Sage Publications
Publisher Place: London [u.a.]
Volume: 23
Issue: 2
Publisher DOI: 10.1177/1059712314563620
Page Start: 69
Page End: 82
EISSN: 1741-2633
ISSN: 1059-7123
Notes: Dieser Beitrag ist mit Zustimmung des Rechteinhabers aufgrund einer (DFG geförderten) Allianz- bzw. Nationallizenz frei zugänglich.
This publication is with permission of the rights owner freely accessible due to an Alliance licence and a national licence (funded by the DFG, German Research Foundation) respectively.
Appears in Collections:Fachgebiet Maschinelles Lernen » Publications

Files in This Item:
File SizeFormat 
1059712314563620.pdf2.3 MBAdobe PDFView/Open


Items in DepositOnce are protected by copyright, with all rights reserved, unless otherwise indicated.