Motions obtaining of multi-degree-freedom underwater robot by using reinforcement learning algorithms

Youkun Han, Hajime Kimura

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Citations (Scopus)

Abstract

This paper deals with motions obtaining of an underwater robot arm which have multi-degree of freedom by using reinforcement learning algorithms. A natural gradient Actor-Critic algorithm which uses Eligibility Traces is applied to the robot arm. In this algorithm, motion planning problems are modeled as finite state Markov decision processes. The robot arm is developed to have 4 joints, each joint consists 1 servo motor. The experiment results show the robot arm successfully learning to swim by feasible learning steps.

Original languageEnglish
Title of host publicationTENCON 2010 - 2010 IEEE Region 10 Conference
Pages1498-1502
Number of pages5
DOIs
Publication statusPublished - Dec 1 2010
Event2010 IEEE Region 10 Conference, TENCON 2010 - Fukuoka, Japan
Duration: Nov 21 2010Nov 24 2010

Other

Other2010 IEEE Region 10 Conference, TENCON 2010
CountryJapan
CityFukuoka
Period11/21/1011/24/10

Fingerprint

Reinforcement learning
Learning algorithms
Robots
Motion planning
Experiments

All Science Journal Classification (ASJC) codes

  • Computer Science Applications
  • Electrical and Electronic Engineering

Cite this

Motions obtaining of multi-degree-freedom underwater robot by using reinforcement learning algorithms. / Han, Youkun; Kimura, Hajime.

TENCON 2010 - 2010 IEEE Region 10 Conference. 2010. p. 1498-1502 5686136.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Han, Y & Kimura, H 2010, Motions obtaining of multi-degree-freedom underwater robot by using reinforcement learning algorithms. in TENCON 2010 - 2010 IEEE Region 10 Conference., 5686136, pp. 1498-1502, 2010 IEEE Region 10 Conference, TENCON 2010, Fukuoka, Japan, 11/21/10. https://doi.org/10.1109/TENCON.2010.5686136
Han, Youkun ; Kimura, Hajime. / Motions obtaining of multi-degree-freedom underwater robot by using reinforcement learning algorithms. TENCON 2010 - 2010 IEEE Region 10 Conference. 2010. pp. 1498-1502
@inproceedings{7ef4b509f6744b07bd97782e5e90cbf1,
title = "Motions obtaining of multi-degree-freedom underwater robot by using reinforcement learning algorithms",
abstract = "This paper deals with motions obtaining of an underwater robot arm which have multi-degree of freedom by using reinforcement learning algorithms. A natural gradient Actor-Critic algorithm which uses Eligibility Traces is applied to the robot arm. In this algorithm, motion planning problems are modeled as finite state Markov decision processes. The robot arm is developed to have 4 joints, each joint consists 1 servo motor. The experiment results show the robot arm successfully learning to swim by feasible learning steps.",
author = "Youkun Han and Hajime Kimura",
year = "2010",
month = "12",
day = "1",
doi = "10.1109/TENCON.2010.5686136",
language = "English",
isbn = "9781424468904",
pages = "1498--1502",
booktitle = "TENCON 2010 - 2010 IEEE Region 10 Conference",

}

TY - GEN

T1 - Motions obtaining of multi-degree-freedom underwater robot by using reinforcement learning algorithms

AU - Han, Youkun

AU - Kimura, Hajime

PY - 2010/12/1

Y1 - 2010/12/1

N2 - This paper deals with motions obtaining of an underwater robot arm which have multi-degree of freedom by using reinforcement learning algorithms. A natural gradient Actor-Critic algorithm which uses Eligibility Traces is applied to the robot arm. In this algorithm, motion planning problems are modeled as finite state Markov decision processes. The robot arm is developed to have 4 joints, each joint consists 1 servo motor. The experiment results show the robot arm successfully learning to swim by feasible learning steps.

AB - This paper deals with motions obtaining of an underwater robot arm which have multi-degree of freedom by using reinforcement learning algorithms. A natural gradient Actor-Critic algorithm which uses Eligibility Traces is applied to the robot arm. In this algorithm, motion planning problems are modeled as finite state Markov decision processes. The robot arm is developed to have 4 joints, each joint consists 1 servo motor. The experiment results show the robot arm successfully learning to swim by feasible learning steps.

UR - http://www.scopus.com/inward/record.url?scp=79951630018&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79951630018&partnerID=8YFLogxK

U2 - 10.1109/TENCON.2010.5686136

DO - 10.1109/TENCON.2010.5686136

M3 - Conference contribution

AN - SCOPUS:79951630018

SN - 9781424468904

SP - 1498

EP - 1502

BT - TENCON 2010 - 2010 IEEE Region 10 Conference

ER -